PURLs

PRACTICE CHANGER
Use a validated tool to screen women of childbearing age for intimate partner violence (IPV) and follow up with any woman with a positive screen.¹

STRENGTH OF RECOMMENDATION
B: Based on a systematic review of 10 randomized controlled trials (RCTs), 11 prospective cohort and cross-sectional studies, and 13 diagnostic accuracy studies.

ILLUSTRATIVE CASE
A healthy 27-year-old woman schedules a visit to discuss birth control options. Should you screen her for IPV, and if so, what instrument should you use?

Each year in the United States, an estimated 5.3 million women ages 18 and older are affected by IPV, resulting in nearly 2 million injuries and more than $4 billion in direct medical and mental health costs.² In addition to the immediate effects, which include death as well as injuries from physical and sexual assault,² IPV has long-term consequences, such as chronic physical and mental illness and substance abuse.³

Too little evidence of benefit?
In 2011, the Institute of Medicine (IOM) recommended for the first time that all women of childbearing age be screened for IPV and identified IPV screening as one of a number of preventive services that are important to women’s health.⁴ The IOM’s recommendation is in line with positions held by the American Medical Association’s National Advisory Council on Violence and Abuse⁵ and the American College of Obstetrics and Gynecology.⁶ These recommendations differ from that of the US Preventive Services Task Force (USPSTF), which determined in 2004 that there was insufficient evidence for or against screening women for IPV.⁷ In issuing its I rating, the USPSTF cited a lack of studies evaluating the accuracy of screening tools for identifying IPV and a lack of evidence as to whether interventions lead to a reduction in harm.

The 2012 systematic review detailed below was undertaken on behalf of the USPSTF to assess the latest evidence and update its recommendation. The USPSTF and the Agency for Healthcare Research and Quality (AHRQ) determined the focus and scope of the review.

STUDY SUMMARY
USPSTF issues a B recommendation for IPV screening
Thirty-four studies of women who sought care in either primary care settings or emergency departments (EDs) but had no complaints related to IPV were included in the review, which addressed four key questions.

Question 1: Does screening women for current, past, or increased risk for IPV reduce exposure to IPV, morbidity, or mortality? No, according to one large RCT whose validity was compromised by high dropout rates. The researchers reviewed a multicenter RCT with 6,743 participants ages 18 to 64 to answer that question. (The study was deemed to be of fair quality because of the high percentage of dropouts from both the screened and unscreened groups.)

The women, recruited from primary care, acute care, and obstetrics and gynecology clinics in Canada, were randomly assigned to either screening with the Woman Abuse Screening Tool (WAST)—an eight-question, self-administered and validated tool—or no screening. Primary outcomes were exposure to abuse and quality of life in the 18 months after screening; secondary outcomes included both mental and physical ailments.

Those in the intervention group underwent screening before seeing their clinicians, who received the positive results before the patient encounter but were not told how, or whether, to respond. Women in both the screened and unscreened groups had access to IPV resources, including psychologists, social workers, crisis hotlines, sexual assault crisis centers, counseling services, and women’s shelters, as well as physician visits. In addition, all participants completed a validated Composite Abuse Scale, a broader (30-question) self-administered measure of IPV, at the end of the visit. Those with positive scores were followed for 18 months.

At follow-up, women in both the screened and unscreened groups had accessed additional health care services. Both groups also had reduced IPV, posttraumatic stress disorder, depression, and alcohol problems, as well as improved quality of life and mental health. There was no statistical difference in outcomes between the groups.

Question 2: How effective are the screening techniques? The efficacy of at least five tools has been demonstrated. Fifteen diagnostic accuracy studies, using cross-sectional and prospective data, evaluated a total of 13 screening instruments.

Five of the 13 screening tools—the face-to-face Hurt, Insult, Threaten, and Scream (HITS) tool, the self-administered Ongoing Violence Assessment Tool (OVAT), the face-to-face Slapped, Threatened and Throw (STaT) instrument, the self-administered Humiliation, Afraid, Rape, Kick (HARK) tool, and the WAST—were at least 80% sensitive and 50% specific in identifying IPV in asymptomatic women.

Question 3: How well do the interventions reduce exposure to IPV, morbidity, or mortality in women with positive screens? Interventions improve outcomes, according to several studies.

One good-quality RCT comparing prenatal behavioral counseling by psychologists or social workers with usual care found that the intervention led to decreased IPV up to 10 weeks postpartum and improved birth outcomes. These included a reduction in preterm births, increased mean gestational age, and decreased rates of very low birth weight, although the difference for very low birth weight was not statistically significant.

One fair-quality trial comparing home visitation by paraprofessionals with usual care for postpartum women led to lower rates of IPV for those in the home visitation group three years after the intervention.

Another study compared a counseling intervention with usual care for women who had reported recent IPV. The intervention led to a decrease in pregnancy coercion—being physically or verbally threatened with pregnancy or prevented from using contraception—and an increase in the likelihood of ending an unsafe relationship.

Two trials evaluating counseling versus wallet-sized referral cards and nurse management versus usual care during pregnancy showed improved outcomes in both the intervention and control groups, with no statistically significant difference between them.

Question 4: What are the adverse effects of screening for IPV and interventions to reduce harm? There are few—if any—adverse effects, according to three RCTs and several descriptive studies. The RCTs found no adverse effects of screening or IPV interventions. Descriptive studies showed low levels of harm among a wide range of study populations and a variety of methods. However, some women experienced loss of privacy, emotional distress, and concerns about further abuse.

WHAT’S NEW
B recommendation finalized
Given the relative safety of screening, the potential benefits of interventions for women who have positive screens, and the availability of accurate screening instruments, the USPSTF disseminated a draft recommendation that health care providers screen all women between ages 14 and 46 for IPV. (The recommendation was finalized in late January 8).

CAVEATS
Universal screening questions remain
While the findings from this systematic review led the USPSTF to upgrade its recommendation for IPV screening from a rating of I (insufficient evidence) to a rating of B (moderate to substantial benefit of screening), additional high-quality studies are needed to definitively reveal the benefit of screening.

The validity of the large multicenter RCT that found no benefit from IPV screening was compromised by high dropout rates and, potentially, by the fact that women in the control group had access to materials that increased IPV awareness. Overall, the trials included in this review ranged from fair to good quality and had relatively high and differential rates of loss to follow-up, enrollment of dissimilar groups, and concern for the Hawthorne effect (in which participants change their behavior simply as a result of being involved in a study).

What’s more, some trials used narrowly defined populations, which could limit applicability. And, while some earlier studies had found higher rates of IPV disclosure using self-administered instruments compared with face-to-face questioning, more research is needed to identify the optimal screening method.⁹

CHALLENGES TO IMPLEMENTATION
The right screen—and reliable follow-up
Five of the screening instruments used in studies included in this systematic review accurately identified women with past or present IPV. Three of these are suitable for use in primary care:

•

HARK, a self-administered screen (available on the Internet at www.ncbi.nlm.nih.gov/pmc/articles/PMC2034562/table/T1)

• HITS, a face-to-face screen

•

WAST, a self-administered screen

More information about the latter two screens is available at www.cdc.gov/ncipc/pub-res/images/ipvandsvscreening.pdf.

After deciding which instrument to use, primary care clinicians still must determine how to incorporate screening into a busy practice.

Finally, clinicians should not screen for IPV until reliable procedures and resources for follow-up of patients who screen positive have been identified. Resources are readily available through local and national hotline numbers. The number of the National Domestic Violence Hotline is (800) 799-SAFE.       

REFERENCES
1. Nelson HD, Bougatsos C, Blazina I. Screening women for intimate partner violence: a systematic review to update the US Preventive Services Task Force recommendation. Ann Intern Med. 2012;156:796-808.

2.

National Center for Injury Prevention and Control. Costs of intimate partner violence against women in the United States. March 2003. www.cdc.gov/violenceprevention/pdf/IPVBook-a.pdf. Accessed November 7, 2012.

3. Coker AL, Davis KE, Arias I, et al. Physical and mental health effects of intimate partner violence for men and women. Am J Prev Med. 2002;23:260-268.

4. Committee on Preventive Services for Women, IOM. Clinical preventive services for women: closing the gaps. July 2011. www.iom.edu/Reports/2011/Clinical-Preventive-Services-for-Women-Closing-the-Gaps.aspx. Accessed November 7, 2012.

5. AMA, National Advisory Council on Violence and Abuse. Policy compendium. April 2008. www.ama-assn.org/ama1/pub/upload/mm/386/vio_policy_comp.pdf. Accessed November 7 2012.

6. American College of Obstetricians and Gynecologists. Screening tools—domestic violence. www.acog.org/About_ACOG/ACOG_Departments/Violence_Against_Women/Screening_Tools__Domestic_Violence. Accessed November 7 2012.

7. US Preventive Services Task Force. Screening for family and intimate partner violence. 2004. www.uspreventiveservicestaskforce.org/3rduspstf/famviolence/famviolrs.htm. Accessed November 7, 2012.

8. US Preventive Services Task Force. Screening for intimate partner violence and abuse of elderly and vulnerable adults. www.uspreventiveservicestaskforce.org/uspstf12/ipvelder/ipvelderfinalrs.htm. Accessed January 29, 2013.

9. Kapur NA, Windish DM. Optimal methods to screen men and women for intimate partner violence. J Interpers Violence. 2011;26:2335-2352.

Acknowledgement
The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center For Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center For Research Resources or the National Institutes of Health.

Copyright © 2012. The Family Physicians Inquiries Network. All rights reserved.

Reprinted with permission from the Family Physicians Inquiries Network and The Journal of Family Practice. 2013;62:90-92.

PRACTICE CHANGER
Use a validated tool to screen women of childbearing age for intimate partner violence (IPV) and follow up with any woman with a positive screen.¹

STRENGTH OF RECOMMENDATION
B: Based on a systematic review of 10 randomized controlled trials (RCTs), 11 prospective cohort and cross-sectional studies, and 13 diagnostic accuracy studies.

ILLUSTRATIVE CASE
A healthy 27-year-old woman schedules a visit to discuss birth control options. Should you screen her for IPV, and if so, what instrument should you use?

Each year in the United States, an estimated 5.3 million women ages 18 and older are affected by IPV, resulting in nearly 2 million injuries and more than $4 billion in direct medical and mental health costs.² In addition to the immediate effects, which include death as well as injuries from physical and sexual assault,² IPV has long-term consequences, such as chronic physical and mental illness and substance abuse.³

Too little evidence of benefit?
In 2011, the Institute of Medicine (IOM) recommended for the first time that all women of childbearing age be screened for IPV and identified IPV screening as one of a number of preventive services that are important to women’s health.⁴ The IOM’s recommendation is in line with positions held by the American Medical Association’s National Advisory Council on Violence and Abuse⁵ and the American College of Obstetrics and Gynecology.⁶ These recommendations differ from that of the US Preventive Services Task Force (USPSTF), which determined in 2004 that there was insufficient evidence for or against screening women for IPV.⁷ In issuing its I rating, the USPSTF cited a lack of studies evaluating the accuracy of screening tools for identifying IPV and a lack of evidence as to whether interventions lead to a reduction in harm.

The 2012 systematic review detailed below was undertaken on behalf of the USPSTF to assess the latest evidence and update its recommendation. The USPSTF and the Agency for Healthcare Research and Quality (AHRQ) determined the focus and scope of the review.

STUDY SUMMARY
USPSTF issues a B recommendation for IPV screening
Thirty-four studies of women who sought care in either primary care settings or emergency departments (EDs) but had no complaints related to IPV were included in the review, which addressed four key questions.

Question 1: Does screening women for current, past, or increased risk for IPV reduce exposure to IPV, morbidity, or mortality? No, according to one large RCT whose validity was compromised by high dropout rates. The researchers reviewed a multicenter RCT with 6,743 participants ages 18 to 64 to answer that question. (The study was deemed to be of fair quality because of the high percentage of dropouts from both the screened and unscreened groups.)

The women, recruited from primary care, acute care, and obstetrics and gynecology clinics in Canada, were randomly assigned to either screening with the Woman Abuse Screening Tool (WAST)—an eight-question, self-administered and validated tool—or no screening. Primary outcomes were exposure to abuse and quality of life in the 18 months after screening; secondary outcomes included both mental and physical ailments.

Those in the intervention group underwent screening before seeing their clinicians, who received the positive results before the patient encounter but were not told how, or whether, to respond. Women in both the screened and unscreened groups had access to IPV resources, including psychologists, social workers, crisis hotlines, sexual assault crisis centers, counseling services, and women’s shelters, as well as physician visits. In addition, all participants completed a validated Composite Abuse Scale, a broader (30-question) self-administered measure of IPV, at the end of the visit. Those with positive scores were followed for 18 months.

At follow-up, women in both the screened and unscreened groups had accessed additional health care services. Both groups also had reduced IPV, posttraumatic stress disorder, depression, and alcohol problems, as well as improved quality of life and mental health. There was no statistical difference in outcomes between the groups.

Question 2: How effective are the screening techniques? The efficacy of at least five tools has been demonstrated. Fifteen diagnostic accuracy studies, using cross-sectional and prospective data, evaluated a total of 13 screening instruments.

Five of the 13 screening tools—the face-to-face Hurt, Insult, Threaten, and Scream (HITS) tool, the self-administered Ongoing Violence Assessment Tool (OVAT), the face-to-face Slapped, Threatened and Throw (STaT) instrument, the self-administered Humiliation, Afraid, Rape, Kick (HARK) tool, and the WAST—were at least 80% sensitive and 50% specific in identifying IPV in asymptomatic women.

Question 3: How well do the interventions reduce exposure to IPV, morbidity, or mortality in women with positive screens? Interventions improve outcomes, according to several studies.

One good-quality RCT comparing prenatal behavioral counseling by psychologists or social workers with usual care found that the intervention led to decreased IPV up to 10 weeks postpartum and improved birth outcomes. These included a reduction in preterm births, increased mean gestational age, and decreased rates of very low birth weight, although the difference for very low birth weight was not statistically significant.

One fair-quality trial comparing home visitation by paraprofessionals with usual care for postpartum women led to lower rates of IPV for those in the home visitation group three years after the intervention.

Another study compared a counseling intervention with usual care for women who had reported recent IPV. The intervention led to a decrease in pregnancy coercion—being physically or verbally threatened with pregnancy or prevented from using contraception—and an increase in the likelihood of ending an unsafe relationship.

Two trials evaluating counseling versus wallet-sized referral cards and nurse management versus usual care during pregnancy showed improved outcomes in both the intervention and control groups, with no statistically significant difference between them.

Question 4: What are the adverse effects of screening for IPV and interventions to reduce harm? There are few—if any—adverse effects, according to three RCTs and several descriptive studies. The RCTs found no adverse effects of screening or IPV interventions. Descriptive studies showed low levels of harm among a wide range of study populations and a variety of methods. However, some women experienced loss of privacy, emotional distress, and concerns about further abuse.

WHAT’S NEW
B recommendation finalized
Given the relative safety of screening, the potential benefits of interventions for women who have positive screens, and the availability of accurate screening instruments, the USPSTF disseminated a draft recommendation that health care providers screen all women between ages 14 and 46 for IPV. (The recommendation was finalized in late January 8).

CAVEATS
Universal screening questions remain
While the findings from this systematic review led the USPSTF to upgrade its recommendation for IPV screening from a rating of I (insufficient evidence) to a rating of B (moderate to substantial benefit of screening), additional high-quality studies are needed to definitively reveal the benefit of screening.

The validity of the large multicenter RCT that found no benefit from IPV screening was compromised by high dropout rates and, potentially, by the fact that women in the control group had access to materials that increased IPV awareness. Overall, the trials included in this review ranged from fair to good quality and had relatively high and differential rates of loss to follow-up, enrollment of dissimilar groups, and concern for the Hawthorne effect (in which participants change their behavior simply as a result of being involved in a study).

What’s more, some trials used narrowly defined populations, which could limit applicability. And, while some earlier studies had found higher rates of IPV disclosure using self-administered instruments compared with face-to-face questioning, more research is needed to identify the optimal screening method.⁹

CHALLENGES TO IMPLEMENTATION
The right screen—and reliable follow-up
Five of the screening instruments used in studies included in this systematic review accurately identified women with past or present IPV. Three of these are suitable for use in primary care:

•

HARK, a self-administered screen (available on the Internet at www.ncbi.nlm.nih.gov/pmc/articles/PMC2034562/table/T1)

• HITS, a face-to-face screen

•

WAST, a self-administered screen

More information about the latter two screens is available at www.cdc.gov/ncipc/pub-res/images/ipvandsvscreening.pdf.

After deciding which instrument to use, primary care clinicians still must determine how to incorporate screening into a busy practice.

Finally, clinicians should not screen for IPV until reliable procedures and resources for follow-up of patients who screen positive have been identified. Resources are readily available through local and national hotline numbers. The number of the National Domestic Violence Hotline is (800) 799-SAFE.       

REFERENCES
1. Nelson HD, Bougatsos C, Blazina I. Screening women for intimate partner violence: a systematic review to update the US Preventive Services Task Force recommendation. Ann Intern Med. 2012;156:796-808.

2.

National Center for Injury Prevention and Control. Costs of intimate partner violence against women in the United States. March 2003. www.cdc.gov/violenceprevention/pdf/IPVBook-a.pdf. Accessed November 7, 2012.

3. Coker AL, Davis KE, Arias I, et al. Physical and mental health effects of intimate partner violence for men and women. Am J Prev Med. 2002;23:260-268.

4. Committee on Preventive Services for Women, IOM. Clinical preventive services for women: closing the gaps. July 2011. www.iom.edu/Reports/2011/Clinical-Preventive-Services-for-Women-Closing-the-Gaps.aspx. Accessed November 7, 2012.

5. AMA, National Advisory Council on Violence and Abuse. Policy compendium. April 2008. www.ama-assn.org/ama1/pub/upload/mm/386/vio_policy_comp.pdf. Accessed November 7 2012.

6. American College of Obstetricians and Gynecologists. Screening tools—domestic violence. www.acog.org/About_ACOG/ACOG_Departments/Violence_Against_Women/Screening_Tools__Domestic_Violence. Accessed November 7 2012.

7. US Preventive Services Task Force. Screening for family and intimate partner violence. 2004. www.uspreventiveservicestaskforce.org/3rduspstf/famviolence/famviolrs.htm. Accessed November 7, 2012.

8. US Preventive Services Task Force. Screening for intimate partner violence and abuse of elderly and vulnerable adults. www.uspreventiveservicestaskforce.org/uspstf12/ipvelder/ipvelderfinalrs.htm. Accessed January 29, 2013.

9. Kapur NA, Windish DM. Optimal methods to screen men and women for intimate partner violence. J Interpers Violence. 2011;26:2335-2352.

Acknowledgement
The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center For Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center For Research Resources or the National Institutes of Health.

Copyright © 2012. The Family Physicians Inquiries Network. All rights reserved.

Reprinted with permission from the Family Physicians Inquiries Network and The Journal of Family Practice. 2013;62:90-92.

PRACTICE CHANGER
Use a validated tool to screen women of childbearing age for intimate partner violence (IPV) and follow up with any woman with a positive screen.¹

STRENGTH OF RECOMMENDATION
B: Based on a systematic review of 10 randomized controlled trials (RCTs), 11 prospective cohort and cross-sectional studies, and 13 diagnostic accuracy studies.

ILLUSTRATIVE CASE
A healthy 27-year-old woman schedules a visit to discuss birth control options. Should you screen her for IPV, and if so, what instrument should you use?

Each year in the United States, an estimated 5.3 million women ages 18 and older are affected by IPV, resulting in nearly 2 million injuries and more than $4 billion in direct medical and mental health costs.² In addition to the immediate effects, which include death as well as injuries from physical and sexual assault,² IPV has long-term consequences, such as chronic physical and mental illness and substance abuse.³

Too little evidence of benefit?
In 2011, the Institute of Medicine (IOM) recommended for the first time that all women of childbearing age be screened for IPV and identified IPV screening as one of a number of preventive services that are important to women’s health.⁴ The IOM’s recommendation is in line with positions held by the American Medical Association’s National Advisory Council on Violence and Abuse⁵ and the American College of Obstetrics and Gynecology.⁶ These recommendations differ from that of the US Preventive Services Task Force (USPSTF), which determined in 2004 that there was insufficient evidence for or against screening women for IPV.⁷ In issuing its I rating, the USPSTF cited a lack of studies evaluating the accuracy of screening tools for identifying IPV and a lack of evidence as to whether interventions lead to a reduction in harm.

The 2012 systematic review detailed below was undertaken on behalf of the USPSTF to assess the latest evidence and update its recommendation. The USPSTF and the Agency for Healthcare Research and Quality (AHRQ) determined the focus and scope of the review.

STUDY SUMMARY
USPSTF issues a B recommendation for IPV screening
Thirty-four studies of women who sought care in either primary care settings or emergency departments (EDs) but had no complaints related to IPV were included in the review, which addressed four key questions.

Question 1: Does screening women for current, past, or increased risk for IPV reduce exposure to IPV, morbidity, or mortality? No, according to one large RCT whose validity was compromised by high dropout rates. The researchers reviewed a multicenter RCT with 6,743 participants ages 18 to 64 to answer that question. (The study was deemed to be of fair quality because of the high percentage of dropouts from both the screened and unscreened groups.)

The women, recruited from primary care, acute care, and obstetrics and gynecology clinics in Canada, were randomly assigned to either screening with the Woman Abuse Screening Tool (WAST)—an eight-question, self-administered and validated tool—or no screening. Primary outcomes were exposure to abuse and quality of life in the 18 months after screening; secondary outcomes included both mental and physical ailments.

Those in the intervention group underwent screening before seeing their clinicians, who received the positive results before the patient encounter but were not told how, or whether, to respond. Women in both the screened and unscreened groups had access to IPV resources, including psychologists, social workers, crisis hotlines, sexual assault crisis centers, counseling services, and women’s shelters, as well as physician visits. In addition, all participants completed a validated Composite Abuse Scale, a broader (30-question) self-administered measure of IPV, at the end of the visit. Those with positive scores were followed for 18 months.

At follow-up, women in both the screened and unscreened groups had accessed additional health care services. Both groups also had reduced IPV, posttraumatic stress disorder, depression, and alcohol problems, as well as improved quality of life and mental health. There was no statistical difference in outcomes between the groups.

Question 2: How effective are the screening techniques? The efficacy of at least five tools has been demonstrated. Fifteen diagnostic accuracy studies, using cross-sectional and prospective data, evaluated a total of 13 screening instruments.

Five of the 13 screening tools—the face-to-face Hurt, Insult, Threaten, and Scream (HITS) tool, the self-administered Ongoing Violence Assessment Tool (OVAT), the face-to-face Slapped, Threatened and Throw (STaT) instrument, the self-administered Humiliation, Afraid, Rape, Kick (HARK) tool, and the WAST—were at least 80% sensitive and 50% specific in identifying IPV in asymptomatic women.

Question 3: How well do the interventions reduce exposure to IPV, morbidity, or mortality in women with positive screens? Interventions improve outcomes, according to several studies.

One good-quality RCT comparing prenatal behavioral counseling by psychologists or social workers with usual care found that the intervention led to decreased IPV up to 10 weeks postpartum and improved birth outcomes. These included a reduction in preterm births, increased mean gestational age, and decreased rates of very low birth weight, although the difference for very low birth weight was not statistically significant.

One fair-quality trial comparing home visitation by paraprofessionals with usual care for postpartum women led to lower rates of IPV for those in the home visitation group three years after the intervention.

Another study compared a counseling intervention with usual care for women who had reported recent IPV. The intervention led to a decrease in pregnancy coercion—being physically or verbally threatened with pregnancy or prevented from using contraception—and an increase in the likelihood of ending an unsafe relationship.

Two trials evaluating counseling versus wallet-sized referral cards and nurse management versus usual care during pregnancy showed improved outcomes in both the intervention and control groups, with no statistically significant difference between them.

Question 4: What are the adverse effects of screening for IPV and interventions to reduce harm? There are few—if any—adverse effects, according to three RCTs and several descriptive studies. The RCTs found no adverse effects of screening or IPV interventions. Descriptive studies showed low levels of harm among a wide range of study populations and a variety of methods. However, some women experienced loss of privacy, emotional distress, and concerns about further abuse.

WHAT’S NEW
B recommendation finalized
Given the relative safety of screening, the potential benefits of interventions for women who have positive screens, and the availability of accurate screening instruments, the USPSTF disseminated a draft recommendation that health care providers screen all women between ages 14 and 46 for IPV. (The recommendation was finalized in late January 8).

CAVEATS
Universal screening questions remain
While the findings from this systematic review led the USPSTF to upgrade its recommendation for IPV screening from a rating of I (insufficient evidence) to a rating of B (moderate to substantial benefit of screening), additional high-quality studies are needed to definitively reveal the benefit of screening.

The validity of the large multicenter RCT that found no benefit from IPV screening was compromised by high dropout rates and, potentially, by the fact that women in the control group had access to materials that increased IPV awareness. Overall, the trials included in this review ranged from fair to good quality and had relatively high and differential rates of loss to follow-up, enrollment of dissimilar groups, and concern for the Hawthorne effect (in which participants change their behavior simply as a result of being involved in a study).

What’s more, some trials used narrowly defined populations, which could limit applicability. And, while some earlier studies had found higher rates of IPV disclosure using self-administered instruments compared with face-to-face questioning, more research is needed to identify the optimal screening method.⁹

CHALLENGES TO IMPLEMENTATION
The right screen—and reliable follow-up
Five of the screening instruments used in studies included in this systematic review accurately identified women with past or present IPV. Three of these are suitable for use in primary care:

•

HARK, a self-administered screen (available on the Internet at www.ncbi.nlm.nih.gov/pmc/articles/PMC2034562/table/T1)

• HITS, a face-to-face screen

•

WAST, a self-administered screen

More information about the latter two screens is available at www.cdc.gov/ncipc/pub-res/images/ipvandsvscreening.pdf.

After deciding which instrument to use, primary care clinicians still must determine how to incorporate screening into a busy practice.

Finally, clinicians should not screen for IPV until reliable procedures and resources for follow-up of patients who screen positive have been identified. Resources are readily available through local and national hotline numbers. The number of the National Domestic Violence Hotline is (800) 799-SAFE.       

REFERENCES
1. Nelson HD, Bougatsos C, Blazina I. Screening women for intimate partner violence: a systematic review to update the US Preventive Services Task Force recommendation. Ann Intern Med. 2012;156:796-808.

2.

National Center for Injury Prevention and Control. Costs of intimate partner violence against women in the United States. March 2003. www.cdc.gov/violenceprevention/pdf/IPVBook-a.pdf. Accessed November 7, 2012.

3. Coker AL, Davis KE, Arias I, et al. Physical and mental health effects of intimate partner violence for men and women. Am J Prev Med. 2002;23:260-268.

4. Committee on Preventive Services for Women, IOM. Clinical preventive services for women: closing the gaps. July 2011. www.iom.edu/Reports/2011/Clinical-Preventive-Services-for-Women-Closing-the-Gaps.aspx. Accessed November 7, 2012.

5. AMA, National Advisory Council on Violence and Abuse. Policy compendium. April 2008. www.ama-assn.org/ama1/pub/upload/mm/386/vio_policy_comp.pdf. Accessed November 7 2012.

6. American College of Obstetricians and Gynecologists. Screening tools—domestic violence. www.acog.org/About_ACOG/ACOG_Departments/Violence_Against_Women/Screening_Tools__Domestic_Violence. Accessed November 7 2012.

7. US Preventive Services Task Force. Screening for family and intimate partner violence. 2004. www.uspreventiveservicestaskforce.org/3rduspstf/famviolence/famviolrs.htm. Accessed November 7, 2012.

8. US Preventive Services Task Force. Screening for intimate partner violence and abuse of elderly and vulnerable adults. www.uspreventiveservicestaskforce.org/uspstf12/ipvelder/ipvelderfinalrs.htm. Accessed January 29, 2013.

9. Kapur NA, Windish DM. Optimal methods to screen men and women for intimate partner violence. J Interpers Violence. 2011;26:2335-2352.

Acknowledgement
The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center For Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center For Research Resources or the National Institutes of Health.

Copyright © 2012. The Family Physicians Inquiries Network. All rights reserved.

Reprinted with permission from the Family Physicians Inquiries Network and The Journal of Family Practice. 2013;62:90-92.

ILLUSTRATIVE CASE

A mother brings in her 18-month-old son because he’s had a runny nose and low-grade fever for the past 4 days—and a cough that kept them both up last night. You diagnose a viral URI, and she requests a strong cough medicine so he (and she) can get a good night’s sleep. What can you recommend that is both safe and effective for a child of this age?

For primary care physicians, office visits for coughing kids with URIs are commonplace. In addition to the cost of such visits, Americans spend some $3.5 billion a year on over-the-counter (OTC) cough and cold remedies, and often give them to young children.

It’s not enough to tell parents what not to do
As physicians (and parents), we understand the desire to give a coughing child something to ease the symptoms. We also know that OTC cough and cold medications can lead to serious complications, and even death. Between 1983 and 2007, 118 pediatric deaths were attributed to the misuse of such preparations.² And, in a 3-year span (2005-2008), the American Association of Poison Control Centers received 64,658 calls for exposures to cough and cold remedies in children younger than 2 years of age, 28 of which resulted in a major adverse reaction or death.³

The US Food and Drug Administration recommends against the use of OTC cough and cold medications in children younger than 2 years,⁴ and the American Academy of Pediatrics has issued strict warnings about the use of OTC cough and cold preparations in children younger than 6 years.⁵ But warning parents of the dangers of giving them to young children without offering an alternative doesn’t satisfy anyone’s needs, and many parents continue to use these medications.

What about honey?
A study published in 2007 evaluated buckwheat honey and found it to be superior to no treatment and equal to honey-flavored dextromethorphan in reducing cough severity and improving sleep for children and their parents.⁶ Honey is known to have both antioxidant and antimicrobial properties—a possible scientific explanation for its effect. Before recommending honey for kids with URIs, however, more evidence of its efficacy was needed.

STUDY SUMMARY: Honey reduces cough frequency and severity

Cohen et al sought to determine whether honey, administered before bedtime, would decrease coughing in children between the ages of one and 5 years—and improve sleep for both the children and their caregivers.¹ They enrolled 300 children with a nocturnal cough of <7 days’ duration and a diagnosis of URI in a one-night study.

Children were excluded if they had any signs or symptoms of asthma, pneumonia, sinusitis, allergic rhinitis, or laryngotracheobronchitis, or if they had been given any cough remedy, including honey, the night before. Parents completed a 5-question survey, using a 7-point Likert scale to assess the child’s cough and both the child’s and parents’ sleep the previous night. Only children whose parents rated their child’s cough severity ≥3 in 2 of the 3 questions related to cough were included in the trial.

The study had a double-blind randomized design, with 4 treatment arms. Three groups received 10 g (about 1.5 tsp) of one of 3 types of honey: eucalyptus, citrus, or labiatae (derived from plants including sage, mint, and thyme); the fourth group received a placebo of silan date extract, which is similar to honey in color, texture, and taste.

Children in all 4 groups received the preparation 30 minutes before bedtime. Neither the parents nor the physicians or study coordinators knew which preparation the children received. The following day, research assistants telephoned the parent who had completed the initial survey and asked the same 5 questions. The primary outcome measure was the change in cough frequency from the night before to the night after treatment. Secondary measures included cough severity and the effect on sleep for both the child and the parent.

Of the 300 children initially enrolled, 270 (90%) completed the trial, with an even distribution among the groups. While there were improvements across all outcomes for both the treatment and placebo groups, the changes were statistically significant only in the treatment groups. There were no significant differences in efficacy noted among the 3 types of honey. Adverse effects of stomachache, nausea, or vomiting were noted by 4 parents in the treatment groups and one in the placebo group, a difference that was not statistically significant.

WHAT’S NEW?: We have more evidence of honey’s efficacy

For children older than one year with a viral URI, we can now recommend 1.5 tsp honey to be given prior to bedtime as a cough remedy. This may reduce the use of potentially harmful and often ineffective OTC cough and cold remedies.

CAVEATS: Honey is unsafe for the youngest children

An obvious limitation of this study was its brevity. Although one night of improved cough and sleep is important, a study that showed honey’s sustained benefit as a cough suppressant would be more convincing. What’s more, there are safety concerns that are age-related.

Honey is considered unsafe for children younger than one year because of the risk of botulism. And honey has the potential to increase dental caries if it is given nightly for a prolonged period of time.

We do not know whether all varieties of honey will have the same benefit, and the source of store-bought honey is not always identified. The authors of this study received funding from the Honey Board of Israel.

CHALLENGES TO IMPLEMENTATION: Parents may be reluctant to abandon OTCs

Changing the behavior of parents and other caregivers who are accustomed to treating children with OTC cough and cold remedies is likely to be an uphill battle. Because honey is readily available, however—often as close as the pantry—and perceived to be safe and nutritious, a recommendation from a trusted physician could go a long way toward its implementation.

Acknowledgement

The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

ILLUSTRATIVE CASE

A mother brings in her 18-month-old son because he’s had a runny nose and low-grade fever for the past 4 days—and a cough that kept them both up last night. You diagnose a viral URI, and she requests a strong cough medicine so he (and she) can get a good night’s sleep. What can you recommend that is both safe and effective for a child of this age?

For primary care physicians, office visits for coughing kids with URIs are commonplace. In addition to the cost of such visits, Americans spend some $3.5 billion a year on over-the-counter (OTC) cough and cold remedies, and often give them to young children.

It’s not enough to tell parents what not to do
As physicians (and parents), we understand the desire to give a coughing child something to ease the symptoms. We also know that OTC cough and cold medications can lead to serious complications, and even death. Between 1983 and 2007, 118 pediatric deaths were attributed to the misuse of such preparations.² And, in a 3-year span (2005-2008), the American Association of Poison Control Centers received 64,658 calls for exposures to cough and cold remedies in children younger than 2 years of age, 28 of which resulted in a major adverse reaction or death.³

The US Food and Drug Administration recommends against the use of OTC cough and cold medications in children younger than 2 years,⁴ and the American Academy of Pediatrics has issued strict warnings about the use of OTC cough and cold preparations in children younger than 6 years.⁵ But warning parents of the dangers of giving them to young children without offering an alternative doesn’t satisfy anyone’s needs, and many parents continue to use these medications.

What about honey?
A study published in 2007 evaluated buckwheat honey and found it to be superior to no treatment and equal to honey-flavored dextromethorphan in reducing cough severity and improving sleep for children and their parents.⁶ Honey is known to have both antioxidant and antimicrobial properties—a possible scientific explanation for its effect. Before recommending honey for kids with URIs, however, more evidence of its efficacy was needed.

STUDY SUMMARY: Honey reduces cough frequency and severity

Cohen et al sought to determine whether honey, administered before bedtime, would decrease coughing in children between the ages of one and 5 years—and improve sleep for both the children and their caregivers.¹ They enrolled 300 children with a nocturnal cough of <7 days’ duration and a diagnosis of URI in a one-night study.

Children were excluded if they had any signs or symptoms of asthma, pneumonia, sinusitis, allergic rhinitis, or laryngotracheobronchitis, or if they had been given any cough remedy, including honey, the night before. Parents completed a 5-question survey, using a 7-point Likert scale to assess the child’s cough and both the child’s and parents’ sleep the previous night. Only children whose parents rated their child’s cough severity ≥3 in 2 of the 3 questions related to cough were included in the trial.

The study had a double-blind randomized design, with 4 treatment arms. Three groups received 10 g (about 1.5 tsp) of one of 3 types of honey: eucalyptus, citrus, or labiatae (derived from plants including sage, mint, and thyme); the fourth group received a placebo of silan date extract, which is similar to honey in color, texture, and taste.

Children in all 4 groups received the preparation 30 minutes before bedtime. Neither the parents nor the physicians or study coordinators knew which preparation the children received. The following day, research assistants telephoned the parent who had completed the initial survey and asked the same 5 questions. The primary outcome measure was the change in cough frequency from the night before to the night after treatment. Secondary measures included cough severity and the effect on sleep for both the child and the parent.

Of the 300 children initially enrolled, 270 (90%) completed the trial, with an even distribution among the groups. While there were improvements across all outcomes for both the treatment and placebo groups, the changes were statistically significant only in the treatment groups. There were no significant differences in efficacy noted among the 3 types of honey. Adverse effects of stomachache, nausea, or vomiting were noted by 4 parents in the treatment groups and one in the placebo group, a difference that was not statistically significant.

WHAT’S NEW?: We have more evidence of honey’s efficacy

For children older than one year with a viral URI, we can now recommend 1.5 tsp honey to be given prior to bedtime as a cough remedy. This may reduce the use of potentially harmful and often ineffective OTC cough and cold remedies.

CAVEATS: Honey is unsafe for the youngest children

An obvious limitation of this study was its brevity. Although one night of improved cough and sleep is important, a study that showed honey’s sustained benefit as a cough suppressant would be more convincing. What’s more, there are safety concerns that are age-related.

Honey is considered unsafe for children younger than one year because of the risk of botulism. And honey has the potential to increase dental caries if it is given nightly for a prolonged period of time.

We do not know whether all varieties of honey will have the same benefit, and the source of store-bought honey is not always identified. The authors of this study received funding from the Honey Board of Israel.

CHALLENGES TO IMPLEMENTATION: Parents may be reluctant to abandon OTCs

Changing the behavior of parents and other caregivers who are accustomed to treating children with OTC cough and cold remedies is likely to be an uphill battle. Because honey is readily available, however—often as close as the pantry—and perceived to be safe and nutritious, a recommendation from a trusted physician could go a long way toward its implementation.

Acknowledgement

The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

ILLUSTRATIVE CASE

A mother brings in her 18-month-old son because he’s had a runny nose and low-grade fever for the past 4 days—and a cough that kept them both up last night. You diagnose a viral URI, and she requests a strong cough medicine so he (and she) can get a good night’s sleep. What can you recommend that is both safe and effective for a child of this age?

For primary care physicians, office visits for coughing kids with URIs are commonplace. In addition to the cost of such visits, Americans spend some $3.5 billion a year on over-the-counter (OTC) cough and cold remedies, and often give them to young children.

It’s not enough to tell parents what not to do
As physicians (and parents), we understand the desire to give a coughing child something to ease the symptoms. We also know that OTC cough and cold medications can lead to serious complications, and even death. Between 1983 and 2007, 118 pediatric deaths were attributed to the misuse of such preparations.² And, in a 3-year span (2005-2008), the American Association of Poison Control Centers received 64,658 calls for exposures to cough and cold remedies in children younger than 2 years of age, 28 of which resulted in a major adverse reaction or death.³

The US Food and Drug Administration recommends against the use of OTC cough and cold medications in children younger than 2 years,⁴ and the American Academy of Pediatrics has issued strict warnings about the use of OTC cough and cold preparations in children younger than 6 years.⁵ But warning parents of the dangers of giving them to young children without offering an alternative doesn’t satisfy anyone’s needs, and many parents continue to use these medications.

What about honey?
A study published in 2007 evaluated buckwheat honey and found it to be superior to no treatment and equal to honey-flavored dextromethorphan in reducing cough severity and improving sleep for children and their parents.⁶ Honey is known to have both antioxidant and antimicrobial properties—a possible scientific explanation for its effect. Before recommending honey for kids with URIs, however, more evidence of its efficacy was needed.

STUDY SUMMARY: Honey reduces cough frequency and severity

Cohen et al sought to determine whether honey, administered before bedtime, would decrease coughing in children between the ages of one and 5 years—and improve sleep for both the children and their caregivers.¹ They enrolled 300 children with a nocturnal cough of <7 days’ duration and a diagnosis of URI in a one-night study.

Children were excluded if they had any signs or symptoms of asthma, pneumonia, sinusitis, allergic rhinitis, or laryngotracheobronchitis, or if they had been given any cough remedy, including honey, the night before. Parents completed a 5-question survey, using a 7-point Likert scale to assess the child’s cough and both the child’s and parents’ sleep the previous night. Only children whose parents rated their child’s cough severity ≥3 in 2 of the 3 questions related to cough were included in the trial.

The study had a double-blind randomized design, with 4 treatment arms. Three groups received 10 g (about 1.5 tsp) of one of 3 types of honey: eucalyptus, citrus, or labiatae (derived from plants including sage, mint, and thyme); the fourth group received a placebo of silan date extract, which is similar to honey in color, texture, and taste.

Children in all 4 groups received the preparation 30 minutes before bedtime. Neither the parents nor the physicians or study coordinators knew which preparation the children received. The following day, research assistants telephoned the parent who had completed the initial survey and asked the same 5 questions. The primary outcome measure was the change in cough frequency from the night before to the night after treatment. Secondary measures included cough severity and the effect on sleep for both the child and the parent.

Of the 300 children initially enrolled, 270 (90%) completed the trial, with an even distribution among the groups. While there were improvements across all outcomes for both the treatment and placebo groups, the changes were statistically significant only in the treatment groups. There were no significant differences in efficacy noted among the 3 types of honey. Adverse effects of stomachache, nausea, or vomiting were noted by 4 parents in the treatment groups and one in the placebo group, a difference that was not statistically significant.

WHAT’S NEW?: We have more evidence of honey’s efficacy

For children older than one year with a viral URI, we can now recommend 1.5 tsp honey to be given prior to bedtime as a cough remedy. This may reduce the use of potentially harmful and often ineffective OTC cough and cold remedies.

CAVEATS: Honey is unsafe for the youngest children

An obvious limitation of this study was its brevity. Although one night of improved cough and sleep is important, a study that showed honey’s sustained benefit as a cough suppressant would be more convincing. What’s more, there are safety concerns that are age-related.

Honey is considered unsafe for children younger than one year because of the risk of botulism. And honey has the potential to increase dental caries if it is given nightly for a prolonged period of time.

We do not know whether all varieties of honey will have the same benefit, and the source of store-bought honey is not always identified. The authors of this study received funding from the Honey Board of Israel.

CHALLENGES TO IMPLEMENTATION: Parents may be reluctant to abandon OTCs

Changing the behavior of parents and other caregivers who are accustomed to treating children with OTC cough and cold remedies is likely to be an uphill battle. Because honey is readily available, however—often as close as the pantry—and perceived to be safe and nutritious, a recommendation from a trusted physician could go a long way toward its implementation.

Acknowledgement

The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

Practice Changer
Express your concern about long-term use of benzodiazepines in a letter—a simple intervention that patients often respond to.¹

Strength of recommendation
A: Based on a well-done meta-analysis with few clinical trials.

Illustrative Case
A 65-year-old patient has been taking lorazepam for insomnia for more than a year. You are concerned about her ongoing use and want to wean her from the medication. What strategies can you use to decrease, or eliminate, her use of the drug?

Benzodiazepines are commonly used medications, with an estimated 12-month prevalence of use of 8.6% in the United States.² While short-term use of these antianxiety medications can be effective, long-term use (> 3 months) is associated with significant risk.

Abuse linked to chronic use
Prescription drug abuse has recently become the nation’s leading cause of accidental death, overtaking motor vehicle accidents.³ And tranquilizers, including benzodiazepines, are the second most abused prescription medication, after pain relievers.⁴ In addition to dependence and withdrawal, chronic use of benzodiazepines is associated with daytime somnolence, blunted reflexes, memory loss, cognitive impairment, and an increased risk for falls and fractures—particularly in older patients.⁵
Reducing long-term use of benzodiazepines in a primary care setting is important but challenging. Until recently, most of the successful strategies reported were resource intensive and required multiple office visits.⁶

STUDY SUMMARY
Brief interventions effective
This study was a meta-analysis of randomized controlled trials in which “minimal interventions” were compared with usual care for their effectiveness in reducing or eliminating benzodiazepine use in primary care patients. A minimal intervention was defined as a letter, self-help information, or short consultation with a primary care provider. In each case, the message to the patient included (a) an expression of concern about the patient’s long-term use of the medication, (b) information about the potential adverse effects of the medication, and (c) advice on how to gradually reduce or stop use.

Three studies met the inclusion criteria for randomization, blinding, and analysis by intention-to-treat.^7-9 All three (n = 615) had a six-month follow-up period, a higher proportion of women (> 60%), and participants with a mean age > 60. Each study compared a letter with usual care; two of the three had a third arm that included both a letter and a short consultation.

Pooled results showed twice the reduction in benzodiazepine use in the intervention groups compared with the control groups (risk ratio [RR] = 2.04). The RR for cessation of benzodiazepine use was 2.3. The number needed to treat for a reduction or cessation of use was 12. The studies reported benzodiazepine reduction rates of 20% to 35% in the intervention groups versus 6% to 15% in the usual care groups.^7-9 There appeared to be no additional benefit to adding the brief consultation compared with the letter alone.

WHAT’S NEW
Easy-to-implement strategy
While many methods to reduce benzodiazepine use have been studied, most involved levels of skill and resources that are not feasible for widespread use. This study found that a letter, stating the risks of continued use of the medication and providing a weaning schedule and tips for handling withdrawal, can be effective in reducing chronic use in a small but significant part of the population.

CAVEATS
Withdrawal effects unaddressed
The study did not adequately address the adverse effects of withdrawal from benzodiazepines, with one of the studies reporting significantly worse qualitative (but not quantitative) withdrawal symptoms at six months.7 This is of particular concern, as withdrawal symptoms are associated with the potential for relapse and concomitant abuse of other drugs and alcohol. We recommend that primary care clinicians screen for substance abuse prior to the intervention and arrange for adequate follow-up.

All three studies in the meta-analysis lasted six months; no longer-term outcomes were reported. In addition, the study did not yield enough information to identify patients who would be most likely to respond to this brief intervention.

CHALLENGES TO

IMPLEMENTATION
Which patients to target?
Identifying patients who are appropriate candidates for this brief intervention and providing adequate monitoring for adverse effects of withdrawal are the main challenges of this practice changer. Nonetheless, chronic benzodiazepine use is of considerable concern, and we believe that this is a useful, and manageable, intervention.

REFERENCES
1. Mugunthan K, McGuire T, Glasziou P. Minimal interventions to decrease long-term use of benzodiazepines in primary care: a systematic review and meta-analysis. Br J Gen Pract. 2011;61:e573-e578.
2. Tyrer PJ. Benzodiazepines on trial. Br Med J. 1984;288:1101-1102.
3. CDC. Deaths: Leading causes for 2008. June 6, 2012. www.cdc.gov/nchs/data/nvsr/nvsr60/nvsr60_06.pdf. Accessed October 10, 2012.
4. National Institute on Drug Abuse. Topics in brief: Prescription drug abuse. www.drugabuse.gov/publications/topics-in-brief/prescription-drug-abuse. Accessed October 11, 2012.
5. Morin CM, Bastien C, Guay B, et al. Randomized clinical trial of supervised tapering and cognitive behavior therapy to facilitate benzodiazepine discontinuation in older adults with chronic insomnia. Am J Psychiatry. 2004;161:332-342.
6. Oude Voshaar RC, Couvee JE, van Balkorn AJ, et al. Strategies for discontinuing long-term benzodiazepine use: a meta-analysis. Br J Psychiatr. 2006;189: 213-220.
7. Bashir K, King M, Ashworth M. Controlled evaluation of brief intervention by general practitioners to reduce chronic use of benzodiazepines. Br J Gen Pract. 1994;44:408-412.
8. Cormack MA, Sweeney KG, Hughes-Jones H, et al. Evaluation of an easy, cost-effective strategy to cut benzodiazepine use in general practice. Br J Gen Pract. 1994;44:5-8
9. Heather NA, Bowie A, Ashton H, et al. Randomized controlled trial of two brief interventions against long-term benzodiazepine use: outcome of intervention. Addict Res Theory. 2004;12:141-145.

ACKNOWLEDGEMENT
The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.
Copyright © 2012. The Family Physicians Inquiries Network. All rights reserved.
Reprinted with permission from the Family Physicians Inquiries Network and The Journal of Family Practice. 2012;61(11):671-672.

Practice Changer
Express your concern about long-term use of benzodiazepines in a letter—a simple intervention that patients often respond to.¹

Strength of recommendation
A: Based on a well-done meta-analysis with few clinical trials.

Illustrative Case
A 65-year-old patient has been taking lorazepam for insomnia for more than a year. You are concerned about her ongoing use and want to wean her from the medication. What strategies can you use to decrease, or eliminate, her use of the drug?

Benzodiazepines are commonly used medications, with an estimated 12-month prevalence of use of 8.6% in the United States.² While short-term use of these antianxiety medications can be effective, long-term use (> 3 months) is associated with significant risk.

Abuse linked to chronic use
Prescription drug abuse has recently become the nation’s leading cause of accidental death, overtaking motor vehicle accidents.³ And tranquilizers, including benzodiazepines, are the second most abused prescription medication, after pain relievers.⁴ In addition to dependence and withdrawal, chronic use of benzodiazepines is associated with daytime somnolence, blunted reflexes, memory loss, cognitive impairment, and an increased risk for falls and fractures—particularly in older patients.⁵
Reducing long-term use of benzodiazepines in a primary care setting is important but challenging. Until recently, most of the successful strategies reported were resource intensive and required multiple office visits.⁶

STUDY SUMMARY
Brief interventions effective
This study was a meta-analysis of randomized controlled trials in which “minimal interventions” were compared with usual care for their effectiveness in reducing or eliminating benzodiazepine use in primary care patients. A minimal intervention was defined as a letter, self-help information, or short consultation with a primary care provider. In each case, the message to the patient included (a) an expression of concern about the patient’s long-term use of the medication, (b) information about the potential adverse effects of the medication, and (c) advice on how to gradually reduce or stop use.

Three studies met the inclusion criteria for randomization, blinding, and analysis by intention-to-treat.^7-9 All three (n = 615) had a six-month follow-up period, a higher proportion of women (> 60%), and participants with a mean age > 60. Each study compared a letter with usual care; two of the three had a third arm that included both a letter and a short consultation.

Pooled results showed twice the reduction in benzodiazepine use in the intervention groups compared with the control groups (risk ratio [RR] = 2.04). The RR for cessation of benzodiazepine use was 2.3. The number needed to treat for a reduction or cessation of use was 12. The studies reported benzodiazepine reduction rates of 20% to 35% in the intervention groups versus 6% to 15% in the usual care groups.^7-9 There appeared to be no additional benefit to adding the brief consultation compared with the letter alone.

WHAT’S NEW
Easy-to-implement strategy
While many methods to reduce benzodiazepine use have been studied, most involved levels of skill and resources that are not feasible for widespread use. This study found that a letter, stating the risks of continued use of the medication and providing a weaning schedule and tips for handling withdrawal, can be effective in reducing chronic use in a small but significant part of the population.

CAVEATS
Withdrawal effects unaddressed
The study did not adequately address the adverse effects of withdrawal from benzodiazepines, with one of the studies reporting significantly worse qualitative (but not quantitative) withdrawal symptoms at six months.7 This is of particular concern, as withdrawal symptoms are associated with the potential for relapse and concomitant abuse of other drugs and alcohol. We recommend that primary care clinicians screen for substance abuse prior to the intervention and arrange for adequate follow-up.

All three studies in the meta-analysis lasted six months; no longer-term outcomes were reported. In addition, the study did not yield enough information to identify patients who would be most likely to respond to this brief intervention.

CHALLENGES TO

IMPLEMENTATION
Which patients to target?
Identifying patients who are appropriate candidates for this brief intervention and providing adequate monitoring for adverse effects of withdrawal are the main challenges of this practice changer. Nonetheless, chronic benzodiazepine use is of considerable concern, and we believe that this is a useful, and manageable, intervention.

REFERENCES
1. Mugunthan K, McGuire T, Glasziou P. Minimal interventions to decrease long-term use of benzodiazepines in primary care: a systematic review and meta-analysis. Br J Gen Pract. 2011;61:e573-e578.
2. Tyrer PJ. Benzodiazepines on trial. Br Med J. 1984;288:1101-1102.
3. CDC. Deaths: Leading causes for 2008. June 6, 2012. www.cdc.gov/nchs/data/nvsr/nvsr60/nvsr60_06.pdf. Accessed October 10, 2012.
4. National Institute on Drug Abuse. Topics in brief: Prescription drug abuse. www.drugabuse.gov/publications/topics-in-brief/prescription-drug-abuse. Accessed October 11, 2012.
5. Morin CM, Bastien C, Guay B, et al. Randomized clinical trial of supervised tapering and cognitive behavior therapy to facilitate benzodiazepine discontinuation in older adults with chronic insomnia. Am J Psychiatry. 2004;161:332-342.
6. Oude Voshaar RC, Couvee JE, van Balkorn AJ, et al. Strategies for discontinuing long-term benzodiazepine use: a meta-analysis. Br J Psychiatr. 2006;189: 213-220.
7. Bashir K, King M, Ashworth M. Controlled evaluation of brief intervention by general practitioners to reduce chronic use of benzodiazepines. Br J Gen Pract. 1994;44:408-412.
8. Cormack MA, Sweeney KG, Hughes-Jones H, et al. Evaluation of an easy, cost-effective strategy to cut benzodiazepine use in general practice. Br J Gen Pract. 1994;44:5-8
9. Heather NA, Bowie A, Ashton H, et al. Randomized controlled trial of two brief interventions against long-term benzodiazepine use: outcome of intervention. Addict Res Theory. 2004;12:141-145.

ACKNOWLEDGEMENT
The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.
Copyright © 2012. The Family Physicians Inquiries Network. All rights reserved.
Reprinted with permission from the Family Physicians Inquiries Network and The Journal of Family Practice. 2012;61(11):671-672.

Practice Changer
Express your concern about long-term use of benzodiazepines in a letter—a simple intervention that patients often respond to.¹

Strength of recommendation
A: Based on a well-done meta-analysis with few clinical trials.

Illustrative Case
A 65-year-old patient has been taking lorazepam for insomnia for more than a year. You are concerned about her ongoing use and want to wean her from the medication. What strategies can you use to decrease, or eliminate, her use of the drug?

Benzodiazepines are commonly used medications, with an estimated 12-month prevalence of use of 8.6% in the United States.² While short-term use of these antianxiety medications can be effective, long-term use (> 3 months) is associated with significant risk.

Abuse linked to chronic use
Prescription drug abuse has recently become the nation’s leading cause of accidental death, overtaking motor vehicle accidents.³ And tranquilizers, including benzodiazepines, are the second most abused prescription medication, after pain relievers.⁴ In addition to dependence and withdrawal, chronic use of benzodiazepines is associated with daytime somnolence, blunted reflexes, memory loss, cognitive impairment, and an increased risk for falls and fractures—particularly in older patients.⁵
Reducing long-term use of benzodiazepines in a primary care setting is important but challenging. Until recently, most of the successful strategies reported were resource intensive and required multiple office visits.⁶

STUDY SUMMARY
Brief interventions effective
This study was a meta-analysis of randomized controlled trials in which “minimal interventions” were compared with usual care for their effectiveness in reducing or eliminating benzodiazepine use in primary care patients. A minimal intervention was defined as a letter, self-help information, or short consultation with a primary care provider. In each case, the message to the patient included (a) an expression of concern about the patient’s long-term use of the medication, (b) information about the potential adverse effects of the medication, and (c) advice on how to gradually reduce or stop use.

Three studies met the inclusion criteria for randomization, blinding, and analysis by intention-to-treat.^7-9 All three (n = 615) had a six-month follow-up period, a higher proportion of women (> 60%), and participants with a mean age > 60. Each study compared a letter with usual care; two of the three had a third arm that included both a letter and a short consultation.

Pooled results showed twice the reduction in benzodiazepine use in the intervention groups compared with the control groups (risk ratio [RR] = 2.04). The RR for cessation of benzodiazepine use was 2.3. The number needed to treat for a reduction or cessation of use was 12. The studies reported benzodiazepine reduction rates of 20% to 35% in the intervention groups versus 6% to 15% in the usual care groups.^7-9 There appeared to be no additional benefit to adding the brief consultation compared with the letter alone.

WHAT’S NEW
Easy-to-implement strategy
While many methods to reduce benzodiazepine use have been studied, most involved levels of skill and resources that are not feasible for widespread use. This study found that a letter, stating the risks of continued use of the medication and providing a weaning schedule and tips for handling withdrawal, can be effective in reducing chronic use in a small but significant part of the population.

CAVEATS
Withdrawal effects unaddressed
The study did not adequately address the adverse effects of withdrawal from benzodiazepines, with one of the studies reporting significantly worse qualitative (but not quantitative) withdrawal symptoms at six months.7 This is of particular concern, as withdrawal symptoms are associated with the potential for relapse and concomitant abuse of other drugs and alcohol. We recommend that primary care clinicians screen for substance abuse prior to the intervention and arrange for adequate follow-up.

All three studies in the meta-analysis lasted six months; no longer-term outcomes were reported. In addition, the study did not yield enough information to identify patients who would be most likely to respond to this brief intervention.

CHALLENGES TO

IMPLEMENTATION
Which patients to target?
Identifying patients who are appropriate candidates for this brief intervention and providing adequate monitoring for adverse effects of withdrawal are the main challenges of this practice changer. Nonetheless, chronic benzodiazepine use is of considerable concern, and we believe that this is a useful, and manageable, intervention.

REFERENCES
1. Mugunthan K, McGuire T, Glasziou P. Minimal interventions to decrease long-term use of benzodiazepines in primary care: a systematic review and meta-analysis. Br J Gen Pract. 2011;61:e573-e578.
2. Tyrer PJ. Benzodiazepines on trial. Br Med J. 1984;288:1101-1102.
3. CDC. Deaths: Leading causes for 2008. June 6, 2012. www.cdc.gov/nchs/data/nvsr/nvsr60/nvsr60_06.pdf. Accessed October 10, 2012.
4. National Institute on Drug Abuse. Topics in brief: Prescription drug abuse. www.drugabuse.gov/publications/topics-in-brief/prescription-drug-abuse. Accessed October 11, 2012.
5. Morin CM, Bastien C, Guay B, et al. Randomized clinical trial of supervised tapering and cognitive behavior therapy to facilitate benzodiazepine discontinuation in older adults with chronic insomnia. Am J Psychiatry. 2004;161:332-342.
6. Oude Voshaar RC, Couvee JE, van Balkorn AJ, et al. Strategies for discontinuing long-term benzodiazepine use: a meta-analysis. Br J Psychiatr. 2006;189: 213-220.
7. Bashir K, King M, Ashworth M. Controlled evaluation of brief intervention by general practitioners to reduce chronic use of benzodiazepines. Br J Gen Pract. 1994;44:408-412.
8. Cormack MA, Sweeney KG, Hughes-Jones H, et al. Evaluation of an easy, cost-effective strategy to cut benzodiazepine use in general practice. Br J Gen Pract. 1994;44:5-8
9. Heather NA, Bowie A, Ashton H, et al. Randomized controlled trial of two brief interventions against long-term benzodiazepine use: outcome of intervention. Addict Res Theory. 2004;12:141-145.

ACKNOWLEDGEMENT
The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.
Copyright © 2012. The Family Physicians Inquiries Network. All rights reserved.
Reprinted with permission from the Family Physicians Inquiries Network and The Journal of Family Practice. 2012;61(11):671-672.

ILLUSTRATIVE CASE

A healthy 27-year-old woman schedules a visit to discuss birth control options. Should you screen her for IPV and if so, what instrument should you use?

Each year in the United States, an estimated 5.3 million women ages 18 and older are affected by IPV, resulting in nearly 2 million injuries and more than $4 billion in direct medical and mental health costs.² In addition to the immediate effects, which include death as well as injuries from physical and sexual assault,² IPV has long-term consequences, such as chronic physical and mental illness and substance abuse.³

Too little evidence of benefit?
In 2011, the Institute of Medicine (IOM) recommended for the first time that all women of childbearing age be screened for IPV-and identified IPV screening as one of a number of preventive services that are important to women’s health.⁴ The IOM’s recommendation is in line with positions held by the American Medical Association’s National Advisory Council on Violence and Abuse⁵ and the American College of Obstetrics and Gynecology.⁶ These recommendations differ from that of the US Preventive Services Task Force (USPSTF), which determined in 2004 that there was insufficient evidence for or against screening women for IPV.⁷ In issuing its “I” rating, the USPSTF cited a lack of studies evaluating the accuracy of screening tools for identifying IPV and a lack of evidence as to whether interventions lead to a reduction in harm.

The 2012 systemic review detailed below was undertaken on behalf of the USPSTF to assess the latest evidence and update its recommendation. The USPSTF and Agency for Healthcare Research and Quality (AHRQ) determined the focus and scope of the review.

STUDY SUMMARY: USPSTF issues a B recommendation for IPV screening

Thirty-four studies of women who sought care in either primary care settings or emergency departments (EDs) but had no complaints related to IPV were included in the review, which addressed 4 key questions.

Question 1: Does screening women for current, past, or increased risk of IPV reduce exposure to IPV, morbidity, or mortality?

No, according to one large RCT whose validity was compromised by high dropout rates. The researchers reviewed a multicenter RCT with 6743 participants ages 18 to 64 years to answer that question. (The study was deemed to be of fair quality because of the high percentage of dropouts from both the screened and unscreened groups.)

The women, recruited from primary care, acute care, and obstetrics and gynecology clinics in Canada, were randomly assigned to either screening with the Woman Abuse Screening Tool (WAST)—an 8-question, self-administered and validated tool—or no screening. Primary outcomes were exposure to abuse and quality of life in the 18 months after screening; secondary outcomes included both mental and physical ailments.

Those in the intervention group underwent screening before seeing their clinicians, who received the positive results before the patient encounter but were not told how, or whether, to respond. Women in both the screened and unscreened groups had access to IPV resources, including psychologists, social workers, crisis hotlines, sexual assault crisis centers, counseling services, and women’s shelters, as well as physician visits. In addition, all participants completed a validated Composite Abuse Scale, a broader (30-question) self-administered measure of IPV, at the end of the visit. Those with positive scores were followed for 18 months.

At follow-up, women in both the screened and unscreened groups had accessed additional health care services. Both groups also had reduced IPV, posttraumatic stress disorder, depression, and alcohol problems, and improved quality of life and mental health. There was no statistical difference in outcomes between the groups.

Question 2: How effective are the screening techniques?

The efficacy of at least 5 tools has been demonstrated. Fifteen diagnostic accuracy studies, using cross-sectional and prospective data, evaluated a total of 13 screening instruments.

Five of the 13 screening tools—the face-to-face Hurt, Insult, Threaten, and Scream (HITS) tool, the self-administered Ongoing Violence Assessment Tool (OVAT), the face-to-face Slapped, Threatened and Throw (STaT) instrument, the self-administered Humiliation, Afraid, Rape, Kick (HARK) tool, and the WAST—were at least 80% sensitive and 50% specific in identifying IPV in asymptomatic women.

Question 3: How well do the interventions reduce exposure to IPV, morbidity, or mortality in women with positive screens?

Interventions improve outcomes, according to several studies. One good-quality RCT comparing prenatal behavioral counseling by psychologists or social workers with usual care found that the intervention led to decreased IPV up to 10 weeks’ postpartum and improved birth outcomes. These included a reduction in preterm births, increased mean gestational age, and decreased rates of very low birth weight, although the difference for very low birth weight was not statistically significant.

One fair-quality trial comparing home visitation by paraprofessionals with usual care for postpartum women led to lower rates of IPV for those in the home visitation group 3 years after the intervention.

Another study compared a counseling intervention with usual care for women who had reported recent IPV. The intervention led to a decrease in pregnancy coercion—being physically or verbally threatened with pregnancy or prevented from using contraception—and an increase in the likelihood of ending an unsafe relationship.

Two trials evaluating counseling vs wallet-sized referral cards and nurse management vs usual care during pregnancy showed improved outcomes in both the intervention and control groups, with no statistically significant difference between them.

Question 4: What are the adverse effects of screening for IPV and interventions to reduce harm?

There are few—if any—adverse effects, according to 3 RCTs and several descriptive studies. The RCTs found no adverse effects of screening or IPV interventions. Descriptive studies showed low levels of harm among a wide range of study populations and a variety of methods. However, some women experienced loss of privacy, emotional distress, and concerns about further abuse.

WHAT’S NEW: B recommendation is finalized

Given the relative safety of screening, the potential benefits of interventions for women who have positive screens, and the availability of accurate screening instruments, the USPSTF disseminated a draft recommendation that health care providers screen all women between 14 and 46 years old for IPV.At presstime in late January, the recommendation was finalized.⁸

CAVEATS: Universal screening questions remain

While the findings from this systematic review led the USPSTF to upgrade its recommendation for IPV screening from an I (insufficient evidence) to a B (moderate to substantial benefit of screening), additional high-quality studies are needed to definitively reveal the benefit of screening.

The validity of the large multicenter RCT that found no benefit from IPV screening was compromised by high dropout rates and, potentially, by the fact that women in the control group had access to materials that increased IPV awareness. Overall, the trials included in this systematic review ranged from fair to good quality and had relatively high and differential rates of loss to follow-up, enrollment of dissimilar groups, and concern for the Hawthorne effect (in which participants change their behavior simply as a result of being involved in a study).

What’s more, some trials used narrowly defined populations, which could limit applicability. And, while some earlier studies had found higher rates of IPV disclosure using self-administered instruments compared with face-to-face questioning, more research is needed to identify the optimal screening method.⁹

CHALLENGES TO IMPLEMENTATION: The right screen—and reliable follow-up

Five of the screening instruments used in studies included in this systematic review accurately identified women with past or present IPV. Three of these are suitable for use in primary care:

• HARK, a self-administered screen available at www.ncbi.nlm.nih.gov/pmc/articles/PMC2034562/table/T1
• HITS, a face-to-face screen
• WAST, a self-administered screen (more information about these screens is available at http://www.cdc.gov/ncipc/pub-res/images/ipvandsvscreening.pdf).

After deciding which instrument to use, family physicians still must determine how to incorporate screening into a busy practice.

Finally, physicians should not screen for IPV until reliable procedures and resources for follow-up of patients who screen positive have been identified. Resources are readily available through local and national hotline numbers. The number of the National Domestic Violence Hotline is 800-799-SAFE.

Acknowledgement

The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center For Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center For Research Resources or the National Institutes of Health.

ILLUSTRATIVE CASE

A healthy 27-year-old woman schedules a visit to discuss birth control options. Should you screen her for IPV and if so, what instrument should you use?

Each year in the United States, an estimated 5.3 million women ages 18 and older are affected by IPV, resulting in nearly 2 million injuries and more than $4 billion in direct medical and mental health costs.² In addition to the immediate effects, which include death as well as injuries from physical and sexual assault,² IPV has long-term consequences, such as chronic physical and mental illness and substance abuse.³

Too little evidence of benefit?
In 2011, the Institute of Medicine (IOM) recommended for the first time that all women of childbearing age be screened for IPV-and identified IPV screening as one of a number of preventive services that are important to women’s health.⁴ The IOM’s recommendation is in line with positions held by the American Medical Association’s National Advisory Council on Violence and Abuse⁵ and the American College of Obstetrics and Gynecology.⁶ These recommendations differ from that of the US Preventive Services Task Force (USPSTF), which determined in 2004 that there was insufficient evidence for or against screening women for IPV.⁷ In issuing its “I” rating, the USPSTF cited a lack of studies evaluating the accuracy of screening tools for identifying IPV and a lack of evidence as to whether interventions lead to a reduction in harm.

The 2012 systemic review detailed below was undertaken on behalf of the USPSTF to assess the latest evidence and update its recommendation. The USPSTF and Agency for Healthcare Research and Quality (AHRQ) determined the focus and scope of the review.

STUDY SUMMARY: USPSTF issues a B recommendation for IPV screening

Thirty-four studies of women who sought care in either primary care settings or emergency departments (EDs) but had no complaints related to IPV were included in the review, which addressed 4 key questions.

Question 1: Does screening women for current, past, or increased risk of IPV reduce exposure to IPV, morbidity, or mortality?

No, according to one large RCT whose validity was compromised by high dropout rates. The researchers reviewed a multicenter RCT with 6743 participants ages 18 to 64 years to answer that question. (The study was deemed to be of fair quality because of the high percentage of dropouts from both the screened and unscreened groups.)

The women, recruited from primary care, acute care, and obstetrics and gynecology clinics in Canada, were randomly assigned to either screening with the Woman Abuse Screening Tool (WAST)—an 8-question, self-administered and validated tool—or no screening. Primary outcomes were exposure to abuse and quality of life in the 18 months after screening; secondary outcomes included both mental and physical ailments.

Those in the intervention group underwent screening before seeing their clinicians, who received the positive results before the patient encounter but were not told how, or whether, to respond. Women in both the screened and unscreened groups had access to IPV resources, including psychologists, social workers, crisis hotlines, sexual assault crisis centers, counseling services, and women’s shelters, as well as physician visits. In addition, all participants completed a validated Composite Abuse Scale, a broader (30-question) self-administered measure of IPV, at the end of the visit. Those with positive scores were followed for 18 months.

At follow-up, women in both the screened and unscreened groups had accessed additional health care services. Both groups also had reduced IPV, posttraumatic stress disorder, depression, and alcohol problems, and improved quality of life and mental health. There was no statistical difference in outcomes between the groups.

Question 2: How effective are the screening techniques?

The efficacy of at least 5 tools has been demonstrated. Fifteen diagnostic accuracy studies, using cross-sectional and prospective data, evaluated a total of 13 screening instruments.

Five of the 13 screening tools—the face-to-face Hurt, Insult, Threaten, and Scream (HITS) tool, the self-administered Ongoing Violence Assessment Tool (OVAT), the face-to-face Slapped, Threatened and Throw (STaT) instrument, the self-administered Humiliation, Afraid, Rape, Kick (HARK) tool, and the WAST—were at least 80% sensitive and 50% specific in identifying IPV in asymptomatic women.

Question 3: How well do the interventions reduce exposure to IPV, morbidity, or mortality in women with positive screens?

Interventions improve outcomes, according to several studies. One good-quality RCT comparing prenatal behavioral counseling by psychologists or social workers with usual care found that the intervention led to decreased IPV up to 10 weeks’ postpartum and improved birth outcomes. These included a reduction in preterm births, increased mean gestational age, and decreased rates of very low birth weight, although the difference for very low birth weight was not statistically significant.

One fair-quality trial comparing home visitation by paraprofessionals with usual care for postpartum women led to lower rates of IPV for those in the home visitation group 3 years after the intervention.

Another study compared a counseling intervention with usual care for women who had reported recent IPV. The intervention led to a decrease in pregnancy coercion—being physically or verbally threatened with pregnancy or prevented from using contraception—and an increase in the likelihood of ending an unsafe relationship.

Two trials evaluating counseling vs wallet-sized referral cards and nurse management vs usual care during pregnancy showed improved outcomes in both the intervention and control groups, with no statistically significant difference between them.

Question 4: What are the adverse effects of screening for IPV and interventions to reduce harm?

There are few—if any—adverse effects, according to 3 RCTs and several descriptive studies. The RCTs found no adverse effects of screening or IPV interventions. Descriptive studies showed low levels of harm among a wide range of study populations and a variety of methods. However, some women experienced loss of privacy, emotional distress, and concerns about further abuse.

WHAT’S NEW: B recommendation is finalized

Given the relative safety of screening, the potential benefits of interventions for women who have positive screens, and the availability of accurate screening instruments, the USPSTF disseminated a draft recommendation that health care providers screen all women between 14 and 46 years old for IPV.At presstime in late January, the recommendation was finalized.⁸

CAVEATS: Universal screening questions remain

While the findings from this systematic review led the USPSTF to upgrade its recommendation for IPV screening from an I (insufficient evidence) to a B (moderate to substantial benefit of screening), additional high-quality studies are needed to definitively reveal the benefit of screening.

The validity of the large multicenter RCT that found no benefit from IPV screening was compromised by high dropout rates and, potentially, by the fact that women in the control group had access to materials that increased IPV awareness. Overall, the trials included in this systematic review ranged from fair to good quality and had relatively high and differential rates of loss to follow-up, enrollment of dissimilar groups, and concern for the Hawthorne effect (in which participants change their behavior simply as a result of being involved in a study).

What’s more, some trials used narrowly defined populations, which could limit applicability. And, while some earlier studies had found higher rates of IPV disclosure using self-administered instruments compared with face-to-face questioning, more research is needed to identify the optimal screening method.⁹

CHALLENGES TO IMPLEMENTATION: The right screen—and reliable follow-up

Five of the screening instruments used in studies included in this systematic review accurately identified women with past or present IPV. Three of these are suitable for use in primary care:

• HARK, a self-administered screen available at www.ncbi.nlm.nih.gov/pmc/articles/PMC2034562/table/T1
• HITS, a face-to-face screen
• WAST, a self-administered screen (more information about these screens is available at http://www.cdc.gov/ncipc/pub-res/images/ipvandsvscreening.pdf).

After deciding which instrument to use, family physicians still must determine how to incorporate screening into a busy practice.

Finally, physicians should not screen for IPV until reliable procedures and resources for follow-up of patients who screen positive have been identified. Resources are readily available through local and national hotline numbers. The number of the National Domestic Violence Hotline is 800-799-SAFE.

Acknowledgement

The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center For Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center For Research Resources or the National Institutes of Health.

ILLUSTRATIVE CASE

A healthy 27-year-old woman schedules a visit to discuss birth control options. Should you screen her for IPV and if so, what instrument should you use?

Each year in the United States, an estimated 5.3 million women ages 18 and older are affected by IPV, resulting in nearly 2 million injuries and more than $4 billion in direct medical and mental health costs.² In addition to the immediate effects, which include death as well as injuries from physical and sexual assault,² IPV has long-term consequences, such as chronic physical and mental illness and substance abuse.³

Too little evidence of benefit?
In 2011, the Institute of Medicine (IOM) recommended for the first time that all women of childbearing age be screened for IPV-and identified IPV screening as one of a number of preventive services that are important to women’s health.⁴ The IOM’s recommendation is in line with positions held by the American Medical Association’s National Advisory Council on Violence and Abuse⁵ and the American College of Obstetrics and Gynecology.⁶ These recommendations differ from that of the US Preventive Services Task Force (USPSTF), which determined in 2004 that there was insufficient evidence for or against screening women for IPV.⁷ In issuing its “I” rating, the USPSTF cited a lack of studies evaluating the accuracy of screening tools for identifying IPV and a lack of evidence as to whether interventions lead to a reduction in harm.

The 2012 systemic review detailed below was undertaken on behalf of the USPSTF to assess the latest evidence and update its recommendation. The USPSTF and Agency for Healthcare Research and Quality (AHRQ) determined the focus and scope of the review.

STUDY SUMMARY: USPSTF issues a B recommendation for IPV screening

Thirty-four studies of women who sought care in either primary care settings or emergency departments (EDs) but had no complaints related to IPV were included in the review, which addressed 4 key questions.

Question 1: Does screening women for current, past, or increased risk of IPV reduce exposure to IPV, morbidity, or mortality?

No, according to one large RCT whose validity was compromised by high dropout rates. The researchers reviewed a multicenter RCT with 6743 participants ages 18 to 64 years to answer that question. (The study was deemed to be of fair quality because of the high percentage of dropouts from both the screened and unscreened groups.)

The women, recruited from primary care, acute care, and obstetrics and gynecology clinics in Canada, were randomly assigned to either screening with the Woman Abuse Screening Tool (WAST)—an 8-question, self-administered and validated tool—or no screening. Primary outcomes were exposure to abuse and quality of life in the 18 months after screening; secondary outcomes included both mental and physical ailments.

Those in the intervention group underwent screening before seeing their clinicians, who received the positive results before the patient encounter but were not told how, or whether, to respond. Women in both the screened and unscreened groups had access to IPV resources, including psychologists, social workers, crisis hotlines, sexual assault crisis centers, counseling services, and women’s shelters, as well as physician visits. In addition, all participants completed a validated Composite Abuse Scale, a broader (30-question) self-administered measure of IPV, at the end of the visit. Those with positive scores were followed for 18 months.

At follow-up, women in both the screened and unscreened groups had accessed additional health care services. Both groups also had reduced IPV, posttraumatic stress disorder, depression, and alcohol problems, and improved quality of life and mental health. There was no statistical difference in outcomes between the groups.

Question 2: How effective are the screening techniques?

The efficacy of at least 5 tools has been demonstrated. Fifteen diagnostic accuracy studies, using cross-sectional and prospective data, evaluated a total of 13 screening instruments.

Five of the 13 screening tools—the face-to-face Hurt, Insult, Threaten, and Scream (HITS) tool, the self-administered Ongoing Violence Assessment Tool (OVAT), the face-to-face Slapped, Threatened and Throw (STaT) instrument, the self-administered Humiliation, Afraid, Rape, Kick (HARK) tool, and the WAST—were at least 80% sensitive and 50% specific in identifying IPV in asymptomatic women.

Question 3: How well do the interventions reduce exposure to IPV, morbidity, or mortality in women with positive screens?

Interventions improve outcomes, according to several studies. One good-quality RCT comparing prenatal behavioral counseling by psychologists or social workers with usual care found that the intervention led to decreased IPV up to 10 weeks’ postpartum and improved birth outcomes. These included a reduction in preterm births, increased mean gestational age, and decreased rates of very low birth weight, although the difference for very low birth weight was not statistically significant.

One fair-quality trial comparing home visitation by paraprofessionals with usual care for postpartum women led to lower rates of IPV for those in the home visitation group 3 years after the intervention.

Another study compared a counseling intervention with usual care for women who had reported recent IPV. The intervention led to a decrease in pregnancy coercion—being physically or verbally threatened with pregnancy or prevented from using contraception—and an increase in the likelihood of ending an unsafe relationship.

Two trials evaluating counseling vs wallet-sized referral cards and nurse management vs usual care during pregnancy showed improved outcomes in both the intervention and control groups, with no statistically significant difference between them.

Question 4: What are the adverse effects of screening for IPV and interventions to reduce harm?

There are few—if any—adverse effects, according to 3 RCTs and several descriptive studies. The RCTs found no adverse effects of screening or IPV interventions. Descriptive studies showed low levels of harm among a wide range of study populations and a variety of methods. However, some women experienced loss of privacy, emotional distress, and concerns about further abuse.

WHAT’S NEW: B recommendation is finalized

Given the relative safety of screening, the potential benefits of interventions for women who have positive screens, and the availability of accurate screening instruments, the USPSTF disseminated a draft recommendation that health care providers screen all women between 14 and 46 years old for IPV.At presstime in late January, the recommendation was finalized.⁸

CAVEATS: Universal screening questions remain

While the findings from this systematic review led the USPSTF to upgrade its recommendation for IPV screening from an I (insufficient evidence) to a B (moderate to substantial benefit of screening), additional high-quality studies are needed to definitively reveal the benefit of screening.

The validity of the large multicenter RCT that found no benefit from IPV screening was compromised by high dropout rates and, potentially, by the fact that women in the control group had access to materials that increased IPV awareness. Overall, the trials included in this systematic review ranged from fair to good quality and had relatively high and differential rates of loss to follow-up, enrollment of dissimilar groups, and concern for the Hawthorne effect (in which participants change their behavior simply as a result of being involved in a study).

What’s more, some trials used narrowly defined populations, which could limit applicability. And, while some earlier studies had found higher rates of IPV disclosure using self-administered instruments compared with face-to-face questioning, more research is needed to identify the optimal screening method.⁹

CHALLENGES TO IMPLEMENTATION: The right screen—and reliable follow-up

Five of the screening instruments used in studies included in this systematic review accurately identified women with past or present IPV. Three of these are suitable for use in primary care:

• HARK, a self-administered screen available at www.ncbi.nlm.nih.gov/pmc/articles/PMC2034562/table/T1
• HITS, a face-to-face screen
• WAST, a self-administered screen (more information about these screens is available at http://www.cdc.gov/ncipc/pub-res/images/ipvandsvscreening.pdf).

After deciding which instrument to use, family physicians still must determine how to incorporate screening into a busy practice.

Finally, physicians should not screen for IPV until reliable procedures and resources for follow-up of patients who screen positive have been identified. Resources are readily available through local and national hotline numbers. The number of the National Domestic Violence Hotline is 800-799-SAFE.

Acknowledgement

The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center For Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center For Research Resources or the National Institutes of Health.

Practice Changer
Reconsider the intervals at which you recommend rescreening for osteoporosis; for postmenopausal women with a baseline of normal bone mineral density (BMD) or mild osteopenia, a 15-year interval is probably sufficient.¹

Strength of recommendation
B: Based on a single cohort study.

Illustrative Case
A 67-year-old woman whose recent dual-energy x-ray absorptiometry (DEXA) scan showed mild osteopenia asks when she should have her next bone scan. What should you tell her?

One in five people who sustain a hip fracture die within a year,² and as many as 36% die prematurely.³ Osteoporosis is the primary predictor of fracture risk and, in older white women in particular, low BMD increases the likelihood of fracture by 70% to 80%.⁴

Optimal Screening Frequency Not Known
The US Preventive Services Task Force (USPSTF) guideline for osteoporosis screening concludes that there is a lack of evidence about optimal rescreening intervals and states that intervals > 2 years may be necessary to better predict fracture risk.⁵ In addition, the USPSTF cites a prospective study showing that repeat measurement of BMD after eight years added little predictive value, compared with baseline DEXA scan results.⁶

The prospective cohort study detailed below was undertaken to help guide decisions about how frequently to screen.

Study Summary
Longer intervals are reasonable for those at low risk
Gourlay et al followed 4,957 women ages 67 or older with normal BMD or osteopenia and no history of hip or clinical vertebral fracture or osteoporosis treatment. The primary outcome was the estimated time it would take for 10% of the women to develop osteoporosis. The time until 2% of the women developed such a fracture was the secondary outcome.

Participants had baseline DEXA scans, which were repeated at years 2, 6, 8, 10, and 16. The researchers followed the women until they were diagnosed with osteoporosis, started on medication for osteoporosis, or developed a hip or clinical vertebral fracture.

After adjusting for multiple covariates (age, body mass index, smoking status, use of glucocorticoids, fracture after age 50, estrogen use, and rheumatoid arthritis), the intervals between baseline testing and the development of osteoporosis were:

• 16.8 years for women with normal BMD

• 17.3 years for women with mild osteopenia

• 4.7 years for women with moderate osteopenia

• 1.1 year for women with advanced osteopenia.

Intervals until 2% of the cohort developed fractures were similar.

Overall, a sensible approach was used to estimate reasonable intervals between DEXA screenings: 15 years for women with normal/mild osteopenia (T-score, > –1.50), five years for those with moderate osteopenia (–1.50 to –1.99), and one year for those with advanced osteopenia (–2.00 to –2.49).

What’s New
Many DEXA scans can be eliminated
Rescreening all postmenopausal women every two years is unlikely to reduce osteoporotic fractures. This cohort study provides evidence that rescreening can often be delayed for many years, depending on the patient’s baseline risk. Changing practice based on these findings can reduce resource utilization without adversely affecting women’s health.

Caveats
Questions about applicability may remain
This analysis was limited to women ≥ 67, so different results might be obtained from analyses that included younger postmenopausal women. In addition, 99% of the participants were white. Because the prevalence of osteoporosis of the hip among white women is equal to or slightly higher than it is among nonwhite women, it is likely that the suggested intervals are reasonable estimates for women of all races.

In women older than 80, the interval between baseline testing and the development of osteoporosis was shorter than that of their younger counterparts. Thus, it might be reasonable to reduce rescreening intervals by one-third for women in their 80s.

Challenges to Implementation
Education needed for patients and clinicians
This study is the best so far to address the frequency of rescreening. In order to implement it, patients as well as clinicians will need to be educated. Effective long-term (> 10 y) reminder systems would improve implementation.

The recommendations of professional associations may also be a factor. The National Osteoporosis Foundation recommends assessing BMD every two years, but notes that more frequent testing may sometimes be warranted.⁷ The American College of Preventive Medicine recommends that screening for osteoporosis not occur more often than every two years.⁸             

REFERENCES
1. Gourlay ML, Fine JP, Preisser JS, et al. Bone-density testing interval and transition to osteoporosis in older women. N Engl J Med. 2012;366:225-233.

2. Leibson CL, Tosteson AN, Gabriel SE, et al. Mortality, disability, and nursing home use for persons with and without hip fracture. J Am Geriatr Soc. 2002;50:1644-1650.

3. Abrahamsen B, van Staa T, Ariely R, et al. Excess mortality following hip fracture: a systematic epidemiological review. Osteoporosis Int. 2009;20: 1633-1650.

4. Smith J, Shoukri K. Diagnosis of osteoporosis. Clin Cornerstone. 2000;2:22-33.

5. US Preventive Services Task Force. Screening for osteoporosis: US Preventive Services Task Force recommendation statement. www.uspreventiveservicestaskforce.org/uspstf10/osteoporosis/osteors.htm. Accessed June 15, 2012.

6. Hillier TA, Stone KL, Bauer DC, et al. Evaluating the value of repeat bone mineral density measurement and prediction of fractures in older women. Arch Intern Med. 2007;167:155-160.

7. National Osteoporosis Foundation. Clinician’s guide to prevention and treatment of osteoporosis. 2010. www.nof.org/sites/default/files/pdfs/NOF_ClinicianGuide2009_v7.pdf. Accessed June 30, 2012.

8. Lim LS, Hoeksema LJ, Sherin K; ACPM Prevention Practice Committee. Screening for osteoporosis in the adult US population: ACPM position statement on preventive practice. Am J Prev Med. 2009;36:366-375.

Acknowledgement
The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

Copyright © 2012 The Family Physicians Inquiries Network. All rights reserved. Reprinted with permission from the Family Physicians Inquiries Network and The Journal of Family Practice. 2012;61(9):555-556.

Practice Changer
Reconsider the intervals at which you recommend rescreening for osteoporosis; for postmenopausal women with a baseline of normal bone mineral density (BMD) or mild osteopenia, a 15-year interval is probably sufficient.¹

Strength of recommendation
B: Based on a single cohort study.

Illustrative Case
A 67-year-old woman whose recent dual-energy x-ray absorptiometry (DEXA) scan showed mild osteopenia asks when she should have her next bone scan. What should you tell her?

One in five people who sustain a hip fracture die within a year,² and as many as 36% die prematurely.³ Osteoporosis is the primary predictor of fracture risk and, in older white women in particular, low BMD increases the likelihood of fracture by 70% to 80%.⁴

Optimal Screening Frequency Not Known
The US Preventive Services Task Force (USPSTF) guideline for osteoporosis screening concludes that there is a lack of evidence about optimal rescreening intervals and states that intervals > 2 years may be necessary to better predict fracture risk.⁵ In addition, the USPSTF cites a prospective study showing that repeat measurement of BMD after eight years added little predictive value, compared with baseline DEXA scan results.⁶

The prospective cohort study detailed below was undertaken to help guide decisions about how frequently to screen.

Study Summary
Longer intervals are reasonable for those at low risk
Gourlay et al followed 4,957 women ages 67 or older with normal BMD or osteopenia and no history of hip or clinical vertebral fracture or osteoporosis treatment. The primary outcome was the estimated time it would take for 10% of the women to develop osteoporosis. The time until 2% of the women developed such a fracture was the secondary outcome.

Participants had baseline DEXA scans, which were repeated at years 2, 6, 8, 10, and 16. The researchers followed the women until they were diagnosed with osteoporosis, started on medication for osteoporosis, or developed a hip or clinical vertebral fracture.

After adjusting for multiple covariates (age, body mass index, smoking status, use of glucocorticoids, fracture after age 50, estrogen use, and rheumatoid arthritis), the intervals between baseline testing and the development of osteoporosis were:

• 16.8 years for women with normal BMD

• 17.3 years for women with mild osteopenia

• 4.7 years for women with moderate osteopenia

• 1.1 year for women with advanced osteopenia.

Intervals until 2% of the cohort developed fractures were similar.

Overall, a sensible approach was used to estimate reasonable intervals between DEXA screenings: 15 years for women with normal/mild osteopenia (T-score, > –1.50), five years for those with moderate osteopenia (–1.50 to –1.99), and one year for those with advanced osteopenia (–2.00 to –2.49).

What’s New
Many DEXA scans can be eliminated
Rescreening all postmenopausal women every two years is unlikely to reduce osteoporotic fractures. This cohort study provides evidence that rescreening can often be delayed for many years, depending on the patient’s baseline risk. Changing practice based on these findings can reduce resource utilization without adversely affecting women’s health.

Caveats
Questions about applicability may remain
This analysis was limited to women ≥ 67, so different results might be obtained from analyses that included younger postmenopausal women. In addition, 99% of the participants were white. Because the prevalence of osteoporosis of the hip among white women is equal to or slightly higher than it is among nonwhite women, it is likely that the suggested intervals are reasonable estimates for women of all races.

In women older than 80, the interval between baseline testing and the development of osteoporosis was shorter than that of their younger counterparts. Thus, it might be reasonable to reduce rescreening intervals by one-third for women in their 80s.

Challenges to Implementation
Education needed for patients and clinicians
This study is the best so far to address the frequency of rescreening. In order to implement it, patients as well as clinicians will need to be educated. Effective long-term (> 10 y) reminder systems would improve implementation.

The recommendations of professional associations may also be a factor. The National Osteoporosis Foundation recommends assessing BMD every two years, but notes that more frequent testing may sometimes be warranted.⁷ The American College of Preventive Medicine recommends that screening for osteoporosis not occur more often than every two years.⁸             

REFERENCES
1. Gourlay ML, Fine JP, Preisser JS, et al. Bone-density testing interval and transition to osteoporosis in older women. N Engl J Med. 2012;366:225-233.

2. Leibson CL, Tosteson AN, Gabriel SE, et al. Mortality, disability, and nursing home use for persons with and without hip fracture. J Am Geriatr Soc. 2002;50:1644-1650.

3. Abrahamsen B, van Staa T, Ariely R, et al. Excess mortality following hip fracture: a systematic epidemiological review. Osteoporosis Int. 2009;20: 1633-1650.

4. Smith J, Shoukri K. Diagnosis of osteoporosis. Clin Cornerstone. 2000;2:22-33.

5. US Preventive Services Task Force. Screening for osteoporosis: US Preventive Services Task Force recommendation statement. www.uspreventiveservicestaskforce.org/uspstf10/osteoporosis/osteors.htm. Accessed June 15, 2012.

6. Hillier TA, Stone KL, Bauer DC, et al. Evaluating the value of repeat bone mineral density measurement and prediction of fractures in older women. Arch Intern Med. 2007;167:155-160.

7. National Osteoporosis Foundation. Clinician’s guide to prevention and treatment of osteoporosis. 2010. www.nof.org/sites/default/files/pdfs/NOF_ClinicianGuide2009_v7.pdf. Accessed June 30, 2012.

8. Lim LS, Hoeksema LJ, Sherin K; ACPM Prevention Practice Committee. Screening for osteoporosis in the adult US population: ACPM position statement on preventive practice. Am J Prev Med. 2009;36:366-375.

Acknowledgement
The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

Copyright © 2012 The Family Physicians Inquiries Network. All rights reserved. Reprinted with permission from the Family Physicians Inquiries Network and The Journal of Family Practice. 2012;61(9):555-556.

Practice Changer
Reconsider the intervals at which you recommend rescreening for osteoporosis; for postmenopausal women with a baseline of normal bone mineral density (BMD) or mild osteopenia, a 15-year interval is probably sufficient.¹

Strength of recommendation
B: Based on a single cohort study.

Illustrative Case
A 67-year-old woman whose recent dual-energy x-ray absorptiometry (DEXA) scan showed mild osteopenia asks when she should have her next bone scan. What should you tell her?

One in five people who sustain a hip fracture die within a year,² and as many as 36% die prematurely.³ Osteoporosis is the primary predictor of fracture risk and, in older white women in particular, low BMD increases the likelihood of fracture by 70% to 80%.⁴

Optimal Screening Frequency Not Known
The US Preventive Services Task Force (USPSTF) guideline for osteoporosis screening concludes that there is a lack of evidence about optimal rescreening intervals and states that intervals > 2 years may be necessary to better predict fracture risk.⁵ In addition, the USPSTF cites a prospective study showing that repeat measurement of BMD after eight years added little predictive value, compared with baseline DEXA scan results.⁶

The prospective cohort study detailed below was undertaken to help guide decisions about how frequently to screen.

Study Summary
Longer intervals are reasonable for those at low risk
Gourlay et al followed 4,957 women ages 67 or older with normal BMD or osteopenia and no history of hip or clinical vertebral fracture or osteoporosis treatment. The primary outcome was the estimated time it would take for 10% of the women to develop osteoporosis. The time until 2% of the women developed such a fracture was the secondary outcome.

Participants had baseline DEXA scans, which were repeated at years 2, 6, 8, 10, and 16. The researchers followed the women until they were diagnosed with osteoporosis, started on medication for osteoporosis, or developed a hip or clinical vertebral fracture.

After adjusting for multiple covariates (age, body mass index, smoking status, use of glucocorticoids, fracture after age 50, estrogen use, and rheumatoid arthritis), the intervals between baseline testing and the development of osteoporosis were:

• 16.8 years for women with normal BMD

• 17.3 years for women with mild osteopenia

• 4.7 years for women with moderate osteopenia

• 1.1 year for women with advanced osteopenia.

Intervals until 2% of the cohort developed fractures were similar.

Overall, a sensible approach was used to estimate reasonable intervals between DEXA screenings: 15 years for women with normal/mild osteopenia (T-score, > –1.50), five years for those with moderate osteopenia (–1.50 to –1.99), and one year for those with advanced osteopenia (–2.00 to –2.49).

What’s New
Many DEXA scans can be eliminated
Rescreening all postmenopausal women every two years is unlikely to reduce osteoporotic fractures. This cohort study provides evidence that rescreening can often be delayed for many years, depending on the patient’s baseline risk. Changing practice based on these findings can reduce resource utilization without adversely affecting women’s health.

Caveats
Questions about applicability may remain
This analysis was limited to women ≥ 67, so different results might be obtained from analyses that included younger postmenopausal women. In addition, 99% of the participants were white. Because the prevalence of osteoporosis of the hip among white women is equal to or slightly higher than it is among nonwhite women, it is likely that the suggested intervals are reasonable estimates for women of all races.

In women older than 80, the interval between baseline testing and the development of osteoporosis was shorter than that of their younger counterparts. Thus, it might be reasonable to reduce rescreening intervals by one-third for women in their 80s.

Challenges to Implementation
Education needed for patients and clinicians
This study is the best so far to address the frequency of rescreening. In order to implement it, patients as well as clinicians will need to be educated. Effective long-term (> 10 y) reminder systems would improve implementation.

The recommendations of professional associations may also be a factor. The National Osteoporosis Foundation recommends assessing BMD every two years, but notes that more frequent testing may sometimes be warranted.⁷ The American College of Preventive Medicine recommends that screening for osteoporosis not occur more often than every two years.⁸             

REFERENCES
1. Gourlay ML, Fine JP, Preisser JS, et al. Bone-density testing interval and transition to osteoporosis in older women. N Engl J Med. 2012;366:225-233.

2. Leibson CL, Tosteson AN, Gabriel SE, et al. Mortality, disability, and nursing home use for persons with and without hip fracture. J Am Geriatr Soc. 2002;50:1644-1650.

3. Abrahamsen B, van Staa T, Ariely R, et al. Excess mortality following hip fracture: a systematic epidemiological review. Osteoporosis Int. 2009;20: 1633-1650.

4. Smith J, Shoukri K. Diagnosis of osteoporosis. Clin Cornerstone. 2000;2:22-33.

5. US Preventive Services Task Force. Screening for osteoporosis: US Preventive Services Task Force recommendation statement. www.uspreventiveservicestaskforce.org/uspstf10/osteoporosis/osteors.htm. Accessed June 15, 2012.

6. Hillier TA, Stone KL, Bauer DC, et al. Evaluating the value of repeat bone mineral density measurement and prediction of fractures in older women. Arch Intern Med. 2007;167:155-160.

7. National Osteoporosis Foundation. Clinician’s guide to prevention and treatment of osteoporosis. 2010. www.nof.org/sites/default/files/pdfs/NOF_ClinicianGuide2009_v7.pdf. Accessed June 30, 2012.

8. Lim LS, Hoeksema LJ, Sherin K; ACPM Prevention Practice Committee. Screening for osteoporosis in the adult US population: ACPM position statement on preventive practice. Am J Prev Med. 2009;36:366-375.

Acknowledgement
The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

Copyright © 2012 The Family Physicians Inquiries Network. All rights reserved. Reprinted with permission from the Family Physicians Inquiries Network and The Journal of Family Practice. 2012;61(9):555-556.

Practice Changer
Recommend antispasmodics or antidepressants for patients with irritable bowel syndrome (IBS) and explain that, while fiber may have other benefits, it is unlikely to relieve IBS symptoms.¹

Strength of recommendation
A: Based on a meta-analysis.

Illustrative Case
A 25-year-old woman has intermittent bouts of abdominal pain, constipation, gas, and bloating. You believe she can benefit from treatment for IBS. What should you recommend?

IBS is the most common functional disorder of the gastrointestinal (GI) tract, affecting approximately 15% of the US population² and accounting for annual health care costs of roughly $30 billion.³ The primary symptoms are bloating, gas, and abdominal pain that often improves immediately after a bowel movement. Patients may have intermittent diarrhea and constipation, as well.

IBS may be related to “brain-gut dysfunction”
The etiology of IBS is unclear, but many agree that a combination of abnormal GI motility, visceral hypersensitivity, and “brain-gut dysfunction”—the inability of the brain to send signals that turn down pain produced in the GI tract—are contributing factors. Although IBS is not life threatening, it has a significant personal, social, and psychological impact. Despite its high prevalence and impact, only a limited number of large studies have assessed the effectiveness of various treatments.

Study Summary
Antispasmodics, antidepressants offer relief—fiber does not
This Cochrane review included 56 randomized controlled trials (RCTs) comparing the efficacy of bulking agents (fiber supplements), antispasmodics, or antidepressants with placebo for the treatment of IBS. Twelve RCTs (n = 621) focused on bulking agents, 29 (n = 2,333) on antispasmodics, and 15 (n = 922) on antidepressants. Inclusion criteria included age > 12 years and an IBS diagnosis. The outcomes analyzed were improvement in abdominal pain, global health assessments, and IBS symptom scores. Adverse effects were not evaluated.

• Bulking agents. In studies ranging from four to 16 weeks, bulking agents were found to have no significant effect on abdominal pain (4 studies; standardized mean difference [SMD], 0.03) or global functioning (11 studies; risk ratio [RR], 1.11). Nor was there an improvement in IBS symptom score (3 studies; SMD, 0.00).

• Antispasmodics. Assessed in RCTs ranging from one week to six months, antispasmodics significantly improved abdominal pain (RR, 1.3; number needed to treat [NNT], 7); global functioning (RR, 1.5; NNT, 5), and IBS symptom score (RR, 1.9; NNT, 3). Ten different antispasmodic agents were studied; in subgroup analyses, five of them—cimetropium/dicyclomine, peppermint oil, pinaverium, and trimebutine—were found to have statistically significant benefits.

• Antidepressants. In studies of both tricyclics and SSRIs, antidepressants were found to have a significant effect on improving abdominal pain (RR, 1.5; NNT, 5), global functioning (RR, 1.6; NNT, 4), and IBS symptom score (RR, 2.0; NNT, 4). Subgroup analyses found statistically significant benefits in global functioning for SSRIs, and in abdominal pain and symptom scores for tricyclics.

What’s New
More evidence against fiber
This review confirms earlier findings—that both antispasmodics and antidepressants are effective treatments for IBS, but bulking agents are not. This is an important finding because dietary fiber adjustment is still among the first recommendations made by leading organizations.^4,5

Caveats
Limitations of included studies
Adverse effects of antispasmodics and antidepressants, which may limit compliance and treatment efficacy, were not addressed. The total number of participants in trials of bulking agents was much smaller than that of the other treatments, so it is possible that clinically meaningful improvements were missed. In addition, the duration of interventions was highly variable, ranging from one to four months for bulking agents and antidepressants and from one week to six months for antispasmodics.

It is also important to note that eight of the 12 studies of bulking agents were conducted in GI clinics. Given the possibility that patients referred to GI clinics have already tried and failed to respond to fiber (and thus, that those who do respond to fiber are not given referrals), it may be reasonable for clinicians to recommend a trial of bulking agents for patients with IBS and to monitor them for symptom improvement.

Challenges to Implementation
Patients may favor fiber
Patients with IBS may be reluctant to take antidepressants or antispasmodics, due to concern about adverse effects or because of a preference for what they see as a more “natural” remedy. It may be helpful to explain that while fiber may have some health benefits, such as lowering cholesterol,⁶ antispasmodics and antidepressants have been found to improve IBS symptoms but thus far, fiber has not.

REFERENCES
1. Ruepert L, Quartero AO, deWit NJ, et al. Bulking agents, antispasmodics and antidepressants for the treatment of irritable bowel syndrome. Cochrane Database Syst Rev. 2011;(8):CD003460.

2. Saito YA, Schoenfeld P, Locke GR 3rd. The epidemiology of irritable bowel syndrome in North America: a systematic review. Am J Gastroenterol. 2002;97:1910-1915.

3. Hulisz D. The burden of illness of irritable bowel syndrome: current challenges and hope for the future. J Manag Care Pharm. 2004;10:299-309.

4. American Gastroenterological Association. IBS: A patient’s guide to living with irritable bowel syndrome. www.gastro.org/patient-center/digestive-conditions/irritable-bowel-syndrome. Accessed March 21, 2012.

5. World Gastroenterology Organisation. WGO practice guideline—irritable bowel syndrome: a global perspective (2009). www.worldgastroenterology.org/irritable-bowel-syndrome.html. Accessed March 16, 2012.

6. Gunness P, Gidley MJ. Mechanisms underlying the cholesterol-lowering properties of soluble dietary fibre polysaccharides. Food Funct. 2010; 1:149-155.

Acknowledgement
The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

Copyright © 2012. The Family Physicians Inquiries Network. All rights reserved. Reprinted with permission from the Family Physicians Inquiries Network and The Journal of Family Practice. 2012;61(4):213-214.

Practice Changer
Recommend antispasmodics or antidepressants for patients with irritable bowel syndrome (IBS) and explain that, while fiber may have other benefits, it is unlikely to relieve IBS symptoms.¹

Strength of recommendation
A: Based on a meta-analysis.

Illustrative Case
A 25-year-old woman has intermittent bouts of abdominal pain, constipation, gas, and bloating. You believe she can benefit from treatment for IBS. What should you recommend?

IBS is the most common functional disorder of the gastrointestinal (GI) tract, affecting approximately 15% of the US population² and accounting for annual health care costs of roughly $30 billion.³ The primary symptoms are bloating, gas, and abdominal pain that often improves immediately after a bowel movement. Patients may have intermittent diarrhea and constipation, as well.

IBS may be related to “brain-gut dysfunction”
The etiology of IBS is unclear, but many agree that a combination of abnormal GI motility, visceral hypersensitivity, and “brain-gut dysfunction”—the inability of the brain to send signals that turn down pain produced in the GI tract—are contributing factors. Although IBS is not life threatening, it has a significant personal, social, and psychological impact. Despite its high prevalence and impact, only a limited number of large studies have assessed the effectiveness of various treatments.

Study Summary
Antispasmodics, antidepressants offer relief—fiber does not
This Cochrane review included 56 randomized controlled trials (RCTs) comparing the efficacy of bulking agents (fiber supplements), antispasmodics, or antidepressants with placebo for the treatment of IBS. Twelve RCTs (n = 621) focused on bulking agents, 29 (n = 2,333) on antispasmodics, and 15 (n = 922) on antidepressants. Inclusion criteria included age > 12 years and an IBS diagnosis. The outcomes analyzed were improvement in abdominal pain, global health assessments, and IBS symptom scores. Adverse effects were not evaluated.

• Bulking agents. In studies ranging from four to 16 weeks, bulking agents were found to have no significant effect on abdominal pain (4 studies; standardized mean difference [SMD], 0.03) or global functioning (11 studies; risk ratio [RR], 1.11). Nor was there an improvement in IBS symptom score (3 studies; SMD, 0.00).

• Antispasmodics. Assessed in RCTs ranging from one week to six months, antispasmodics significantly improved abdominal pain (RR, 1.3; number needed to treat [NNT], 7); global functioning (RR, 1.5; NNT, 5), and IBS symptom score (RR, 1.9; NNT, 3). Ten different antispasmodic agents were studied; in subgroup analyses, five of them—cimetropium/dicyclomine, peppermint oil, pinaverium, and trimebutine—were found to have statistically significant benefits.

• Antidepressants. In studies of both tricyclics and SSRIs, antidepressants were found to have a significant effect on improving abdominal pain (RR, 1.5; NNT, 5), global functioning (RR, 1.6; NNT, 4), and IBS symptom score (RR, 2.0; NNT, 4). Subgroup analyses found statistically significant benefits in global functioning for SSRIs, and in abdominal pain and symptom scores for tricyclics.

What’s New
More evidence against fiber
This review confirms earlier findings—that both antispasmodics and antidepressants are effective treatments for IBS, but bulking agents are not. This is an important finding because dietary fiber adjustment is still among the first recommendations made by leading organizations.^4,5

Caveats
Limitations of included studies
Adverse effects of antispasmodics and antidepressants, which may limit compliance and treatment efficacy, were not addressed. The total number of participants in trials of bulking agents was much smaller than that of the other treatments, so it is possible that clinically meaningful improvements were missed. In addition, the duration of interventions was highly variable, ranging from one to four months for bulking agents and antidepressants and from one week to six months for antispasmodics.

It is also important to note that eight of the 12 studies of bulking agents were conducted in GI clinics. Given the possibility that patients referred to GI clinics have already tried and failed to respond to fiber (and thus, that those who do respond to fiber are not given referrals), it may be reasonable for clinicians to recommend a trial of bulking agents for patients with IBS and to monitor them for symptom improvement.

Challenges to Implementation
Patients may favor fiber
Patients with IBS may be reluctant to take antidepressants or antispasmodics, due to concern about adverse effects or because of a preference for what they see as a more “natural” remedy. It may be helpful to explain that while fiber may have some health benefits, such as lowering cholesterol,⁶ antispasmodics and antidepressants have been found to improve IBS symptoms but thus far, fiber has not.

REFERENCES
1. Ruepert L, Quartero AO, deWit NJ, et al. Bulking agents, antispasmodics and antidepressants for the treatment of irritable bowel syndrome. Cochrane Database Syst Rev. 2011;(8):CD003460.

2. Saito YA, Schoenfeld P, Locke GR 3rd. The epidemiology of irritable bowel syndrome in North America: a systematic review. Am J Gastroenterol. 2002;97:1910-1915.

3. Hulisz D. The burden of illness of irritable bowel syndrome: current challenges and hope for the future. J Manag Care Pharm. 2004;10:299-309.

4. American Gastroenterological Association. IBS: A patient’s guide to living with irritable bowel syndrome. www.gastro.org/patient-center/digestive-conditions/irritable-bowel-syndrome. Accessed March 21, 2012.

5. World Gastroenterology Organisation. WGO practice guideline—irritable bowel syndrome: a global perspective (2009). www.worldgastroenterology.org/irritable-bowel-syndrome.html. Accessed March 16, 2012.

6. Gunness P, Gidley MJ. Mechanisms underlying the cholesterol-lowering properties of soluble dietary fibre polysaccharides. Food Funct. 2010; 1:149-155.

Acknowledgement
The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

Copyright © 2012. The Family Physicians Inquiries Network. All rights reserved. Reprinted with permission from the Family Physicians Inquiries Network and The Journal of Family Practice. 2012;61(4):213-214.

Practice Changer
Recommend antispasmodics or antidepressants for patients with irritable bowel syndrome (IBS) and explain that, while fiber may have other benefits, it is unlikely to relieve IBS symptoms.¹

Strength of recommendation
A: Based on a meta-analysis.

Illustrative Case
A 25-year-old woman has intermittent bouts of abdominal pain, constipation, gas, and bloating. You believe she can benefit from treatment for IBS. What should you recommend?

IBS is the most common functional disorder of the gastrointestinal (GI) tract, affecting approximately 15% of the US population² and accounting for annual health care costs of roughly $30 billion.³ The primary symptoms are bloating, gas, and abdominal pain that often improves immediately after a bowel movement. Patients may have intermittent diarrhea and constipation, as well.

IBS may be related to “brain-gut dysfunction”
The etiology of IBS is unclear, but many agree that a combination of abnormal GI motility, visceral hypersensitivity, and “brain-gut dysfunction”—the inability of the brain to send signals that turn down pain produced in the GI tract—are contributing factors. Although IBS is not life threatening, it has a significant personal, social, and psychological impact. Despite its high prevalence and impact, only a limited number of large studies have assessed the effectiveness of various treatments.

Study Summary
Antispasmodics, antidepressants offer relief—fiber does not
This Cochrane review included 56 randomized controlled trials (RCTs) comparing the efficacy of bulking agents (fiber supplements), antispasmodics, or antidepressants with placebo for the treatment of IBS. Twelve RCTs (n = 621) focused on bulking agents, 29 (n = 2,333) on antispasmodics, and 15 (n = 922) on antidepressants. Inclusion criteria included age > 12 years and an IBS diagnosis. The outcomes analyzed were improvement in abdominal pain, global health assessments, and IBS symptom scores. Adverse effects were not evaluated.

• Bulking agents. In studies ranging from four to 16 weeks, bulking agents were found to have no significant effect on abdominal pain (4 studies; standardized mean difference [SMD], 0.03) or global functioning (11 studies; risk ratio [RR], 1.11). Nor was there an improvement in IBS symptom score (3 studies; SMD, 0.00).

• Antispasmodics. Assessed in RCTs ranging from one week to six months, antispasmodics significantly improved abdominal pain (RR, 1.3; number needed to treat [NNT], 7); global functioning (RR, 1.5; NNT, 5), and IBS symptom score (RR, 1.9; NNT, 3). Ten different antispasmodic agents were studied; in subgroup analyses, five of them—cimetropium/dicyclomine, peppermint oil, pinaverium, and trimebutine—were found to have statistically significant benefits.

• Antidepressants. In studies of both tricyclics and SSRIs, antidepressants were found to have a significant effect on improving abdominal pain (RR, 1.5; NNT, 5), global functioning (RR, 1.6; NNT, 4), and IBS symptom score (RR, 2.0; NNT, 4). Subgroup analyses found statistically significant benefits in global functioning for SSRIs, and in abdominal pain and symptom scores for tricyclics.

What’s New
More evidence against fiber
This review confirms earlier findings—that both antispasmodics and antidepressants are effective treatments for IBS, but bulking agents are not. This is an important finding because dietary fiber adjustment is still among the first recommendations made by leading organizations.^4,5

Caveats
Limitations of included studies
Adverse effects of antispasmodics and antidepressants, which may limit compliance and treatment efficacy, were not addressed. The total number of participants in trials of bulking agents was much smaller than that of the other treatments, so it is possible that clinically meaningful improvements were missed. In addition, the duration of interventions was highly variable, ranging from one to four months for bulking agents and antidepressants and from one week to six months for antispasmodics.

It is also important to note that eight of the 12 studies of bulking agents were conducted in GI clinics. Given the possibility that patients referred to GI clinics have already tried and failed to respond to fiber (and thus, that those who do respond to fiber are not given referrals), it may be reasonable for clinicians to recommend a trial of bulking agents for patients with IBS and to monitor them for symptom improvement.

Challenges to Implementation
Patients may favor fiber
Patients with IBS may be reluctant to take antidepressants or antispasmodics, due to concern about adverse effects or because of a preference for what they see as a more “natural” remedy. It may be helpful to explain that while fiber may have some health benefits, such as lowering cholesterol,⁶ antispasmodics and antidepressants have been found to improve IBS symptoms but thus far, fiber has not.

REFERENCES
1. Ruepert L, Quartero AO, deWit NJ, et al. Bulking agents, antispasmodics and antidepressants for the treatment of irritable bowel syndrome. Cochrane Database Syst Rev. 2011;(8):CD003460.

2. Saito YA, Schoenfeld P, Locke GR 3rd. The epidemiology of irritable bowel syndrome in North America: a systematic review. Am J Gastroenterol. 2002;97:1910-1915.

3. Hulisz D. The burden of illness of irritable bowel syndrome: current challenges and hope for the future. J Manag Care Pharm. 2004;10:299-309.

4. American Gastroenterological Association. IBS: A patient’s guide to living with irritable bowel syndrome. www.gastro.org/patient-center/digestive-conditions/irritable-bowel-syndrome. Accessed March 21, 2012.

5. World Gastroenterology Organisation. WGO practice guideline—irritable bowel syndrome: a global perspective (2009). www.worldgastroenterology.org/irritable-bowel-syndrome.html. Accessed March 16, 2012.

6. Gunness P, Gidley MJ. Mechanisms underlying the cholesterol-lowering properties of soluble dietary fibre polysaccharides. Food Funct. 2010; 1:149-155.

Acknowledgement
The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

Copyright © 2012. The Family Physicians Inquiries Network. All rights reserved. Reprinted with permission from the Family Physicians Inquiries Network and The Journal of Family Practice. 2012;61(4):213-214.

ILLUSTRATIVE CASE

A 43-year-old woman with a body mass index (BMI) of 38 kg/m² and a 5-year history of diabetes has a glycated hemoglobin (HbA1c) of 8.5% despite the use of oral hypoglycemic agents. Should you talk to her about gastric bypass surgery to treat her diabetes?

Diet and exercise are the first steps in addressing diabetes, but these interventions are often unsuccessful. The International Diabetes Federation (IDF) recommends consideration of bariatric surgery for patients who have a BMI >35 kg/m² and diabetes that lifestyle modification and pharmacotherapy have failed to control.²

Surgery for diabetes: Is there ample evidence?
Until recently, the IDF’s recommendation was based on observational data and a single RCT that found increased resolution of diabetes following various bariatric procedures.^3-5 In the study detailed in this PURL, Mingrone et al took another look.

STUDY SUMMARY: Surgery led to higher remission rates

This single-center, nonblinded RCT compared 2 malabsorptive procedures— Roux-en-Y gastric bypass and biliopancreatic diversion, a more complicated procedure not commonly performed—with medical therapy.¹ The primary outcome was the rate of diabetes remission at 2 years, defined as a fasting glucose level <100 mg/dL and an HbA1c <6.5%. Changes in BMI and cholesterol levels were among the secondary endpoints.

To be eligible, patients had to be between the ages of 30 and 60 years and have a BMI ≥35 kg/m², a history of type 2 diabetes ≥5 years, and an HbA1c ≥7.0%. Exclusion criteria included a history of type 1 diabetes, diabetes caused by an underlying disease or steroid treatment, previous bariatric surgery, pregnancy, diabetic complications, other severe medical conditions, and acute hospitalization. Both the gastric bypass and biliopancreatic diversion procedures were performed by independent surgical teams.

Participants (N=60) were evaluated at baseline and at 1, 3, 6, 9, 12, and 24 months after the intervention by a team that included a dietician, a nurse, and a physician. All received a diet plan with daily exercise designed by their team. Those in the medical therapy group had their medications titrated to reach a goal HbA1c <7%. Pharmacotherapy was stopped based on normalization of blood sugars or HbA1c <6.5%.

Within 15 days postsurgery, patients in both surgical arms had their diabetes medications stopped based on their blood glucose levels.

At 2 years, 75% of the patients in the gastric bypass arm and 95% of the patients in the biliopancreatic diversion arm (number needed to treat=1.3 and 1, respectively) were considered to be in diabetes remission, defined as a fasting blood sugar of <100 mg/dL and an HbA1c <6.5% after one year without pharmacotherapy. (Notably, this differs from that of the American Diabetes Association, which requires an HbA1c <6.0% for classification as complete remission.) None of the patients in the medical therapy arm was in remission at the 2-year mark.

On average, blood sugars normalized for gastric bypass patients by 10±2 months, vs 4±1 months for biliopancreatic diversion patients (P=.01). The average HbA1c at the end of 2 years was significantly different among all 3 groups (6.35%±1.42 for those undergoing gastric bypass, 4.95%±0.49 for the biliopancreatic diversion group, and 7.69%±0.57 for the medical therapy group), as was the change in HbA1c from baseline (TABLE). Changes in BMI and the number of patients who achieved normalization of total cholesterol were similar for both surgical groups. Interestingly, neither baseline BMI nor amount of weight lost or pre-enrollment duration of diabetes were predictors of diabetes remission or normalization of fasting glucose levels.

TABLE
Surgery vs medical therapy for diabetes: Gastric bypass and biliopancreatic diversion are more effective

There were no deaths associated with this study. There were 2 adverse events requiring reoperation: an incisional hernia in a patient in the biliopancreatic diversion group and an intestinal obstruction in a patient in the gastric bypass group. Six patients in the biliopancreatic diversion arm developed metabolic abnormalities, including iron deficiency anemia, hypoalbuminemia, osteopenia, and osteoporosis. In the gastric bypass arm, 2 patients developed iron deficiency anemia.

WHAT’S NEW?: Evidence of efficacy has grown

This is the first RCT to evaluate biliopancreatic diversion and only the second to evaluate gastric bypass as strategies for controlling diabetes. Similar findings were demonstrated at 12 months in an RCT of 150 obese patients with diabetes in which intensive medical therapy was compared with either gastric bypass or sleeve gastrectomy,⁶ published concurrently with the Mingrone study. Like the Mingrone study, this study found that for select patients with diabetes, surgery may lead to better outcomes than medical management alone.

CAVEATS: Long-term effect is still uncertain

The long-term efficacy of surgery as a way to manage diabetes remains uncertain. Patients in this study were followed for just 2 years and the outcomes were metabolic measures rather than morbidity and mortality. A recent prospective observational study following patients for 6 years after gastric bypass found that the rate of remission for diabetes was 75% (95% confidence interval (CI), 63%-87%) at 2 years but dropped to 62% (95% CI, 49%-75%) at 6 years⁷

A larger study (N=4047) of longer duration—the Swedish Obese Subjects (SOS) cohort study —found a considerably larger drop: The diabetes remission rate for those who had surgery went from 72% at 2 years to 36% at 10-year follow-up, but that was still higher than the 10-year remission rate (13%) for the matched controls.⁴ It is still not clear exactly how long diabetic remission lasts after bariatric surgery or what effect a 10-year respite from the disease will have on the long-term morbidity and mortality of patients with diabetes.

Surgical risks. In small studies such as the one by Mingrone et al,¹ it can be difficult to see the full extent of surgical complications. The much larger SOS study found low mortality rates (0.25%). But 13% of those who underwent bariatric surgery had postoperative complications (number needed to harm = 8), with 2.2% of patients requiring reoperation.⁴ Additionally, women who become pregnant after bariatric surgery are at increased risk for internal hernias or bowel obstruction during pregnancy.⁸

Furthermore, malabsorptive-type surgeries are known to cause nutritional deficiencies, leading to disorders including anemia and osteoporosis.⁶ Importantly, while women of childbearing-age who undergo bariatric surgery decrease their risk of developing gestational hypertension and gestational diabetes, they are more likely to have nutritional deficiencies during pregnancy and to have children with these deficiencies.⁸

CHALLENGES TO IMPLEMENTATION: The ideal candidate remains unclear

It is still not clear from this study which patients should be referred for bariatric surgery. Historically, BMI has been used as the main indication for bariatric surgery, but this and other, studies have found that remission of diabetes is independent of BMI and the amount of weight lost.⁹ A predictive 10-point Diabetes Surgery Score has recently been developed: It uses age, BMI, duration of diabetes, and C-peptide levels to predict the likelihood of diabetes remission after surgery.¹⁰ This scoring system has yet to be validated in non-Asian patients, and a threshold for recommending surgery has been not established. However, this tool indicates that younger patients with a shorter duration of diabetes (which was not a factor in the outcome of the Mingrone study) and no baseline use of insulin are most likely to benefit from surgery. Thus, these patients may be the ones we need to consider referring first.

Cost of surgery. Several studies have shown that bariatric surgery is cost-effective for the treatment of diabetes, and saves money after approximately 5 years.^11,12 However, patients with diabetes and obesity may be uninsured or underinsured, and have high out-of-pocket costs. One challenge will be to ensure that surgery is a viable option for patients with financial constraints.

Acknowledgement

The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

ILLUSTRATIVE CASE

A 43-year-old woman with a body mass index (BMI) of 38 kg/m² and a 5-year history of diabetes has a glycated hemoglobin (HbA1c) of 8.5% despite the use of oral hypoglycemic agents. Should you talk to her about gastric bypass surgery to treat her diabetes?

Diet and exercise are the first steps in addressing diabetes, but these interventions are often unsuccessful. The International Diabetes Federation (IDF) recommends consideration of bariatric surgery for patients who have a BMI >35 kg/m² and diabetes that lifestyle modification and pharmacotherapy have failed to control.²

Surgery for diabetes: Is there ample evidence?
Until recently, the IDF’s recommendation was based on observational data and a single RCT that found increased resolution of diabetes following various bariatric procedures.^3-5 In the study detailed in this PURL, Mingrone et al took another look.

STUDY SUMMARY: Surgery led to higher remission rates

This single-center, nonblinded RCT compared 2 malabsorptive procedures— Roux-en-Y gastric bypass and biliopancreatic diversion, a more complicated procedure not commonly performed—with medical therapy.¹ The primary outcome was the rate of diabetes remission at 2 years, defined as a fasting glucose level <100 mg/dL and an HbA1c <6.5%. Changes in BMI and cholesterol levels were among the secondary endpoints.

To be eligible, patients had to be between the ages of 30 and 60 years and have a BMI ≥35 kg/m², a history of type 2 diabetes ≥5 years, and an HbA1c ≥7.0%. Exclusion criteria included a history of type 1 diabetes, diabetes caused by an underlying disease or steroid treatment, previous bariatric surgery, pregnancy, diabetic complications, other severe medical conditions, and acute hospitalization. Both the gastric bypass and biliopancreatic diversion procedures were performed by independent surgical teams.

Participants (N=60) were evaluated at baseline and at 1, 3, 6, 9, 12, and 24 months after the intervention by a team that included a dietician, a nurse, and a physician. All received a diet plan with daily exercise designed by their team. Those in the medical therapy group had their medications titrated to reach a goal HbA1c <7%. Pharmacotherapy was stopped based on normalization of blood sugars or HbA1c <6.5%.

Within 15 days postsurgery, patients in both surgical arms had their diabetes medications stopped based on their blood glucose levels.

At 2 years, 75% of the patients in the gastric bypass arm and 95% of the patients in the biliopancreatic diversion arm (number needed to treat=1.3 and 1, respectively) were considered to be in diabetes remission, defined as a fasting blood sugar of <100 mg/dL and an HbA1c <6.5% after one year without pharmacotherapy. (Notably, this differs from that of the American Diabetes Association, which requires an HbA1c <6.0% for classification as complete remission.) None of the patients in the medical therapy arm was in remission at the 2-year mark.

On average, blood sugars normalized for gastric bypass patients by 10±2 months, vs 4±1 months for biliopancreatic diversion patients (P=.01). The average HbA1c at the end of 2 years was significantly different among all 3 groups (6.35%±1.42 for those undergoing gastric bypass, 4.95%±0.49 for the biliopancreatic diversion group, and 7.69%±0.57 for the medical therapy group), as was the change in HbA1c from baseline (TABLE). Changes in BMI and the number of patients who achieved normalization of total cholesterol were similar for both surgical groups. Interestingly, neither baseline BMI nor amount of weight lost or pre-enrollment duration of diabetes were predictors of diabetes remission or normalization of fasting glucose levels.

TABLE
Surgery vs medical therapy for diabetes: Gastric bypass and biliopancreatic diversion are more effective

There were no deaths associated with this study. There were 2 adverse events requiring reoperation: an incisional hernia in a patient in the biliopancreatic diversion group and an intestinal obstruction in a patient in the gastric bypass group. Six patients in the biliopancreatic diversion arm developed metabolic abnormalities, including iron deficiency anemia, hypoalbuminemia, osteopenia, and osteoporosis. In the gastric bypass arm, 2 patients developed iron deficiency anemia.

WHAT’S NEW?: Evidence of efficacy has grown

This is the first RCT to evaluate biliopancreatic diversion and only the second to evaluate gastric bypass as strategies for controlling diabetes. Similar findings were demonstrated at 12 months in an RCT of 150 obese patients with diabetes in which intensive medical therapy was compared with either gastric bypass or sleeve gastrectomy,⁶ published concurrently with the Mingrone study. Like the Mingrone study, this study found that for select patients with diabetes, surgery may lead to better outcomes than medical management alone.

CAVEATS: Long-term effect is still uncertain

The long-term efficacy of surgery as a way to manage diabetes remains uncertain. Patients in this study were followed for just 2 years and the outcomes were metabolic measures rather than morbidity and mortality. A recent prospective observational study following patients for 6 years after gastric bypass found that the rate of remission for diabetes was 75% (95% confidence interval (CI), 63%-87%) at 2 years but dropped to 62% (95% CI, 49%-75%) at 6 years⁷

A larger study (N=4047) of longer duration—the Swedish Obese Subjects (SOS) cohort study —found a considerably larger drop: The diabetes remission rate for those who had surgery went from 72% at 2 years to 36% at 10-year follow-up, but that was still higher than the 10-year remission rate (13%) for the matched controls.⁴ It is still not clear exactly how long diabetic remission lasts after bariatric surgery or what effect a 10-year respite from the disease will have on the long-term morbidity and mortality of patients with diabetes.

Surgical risks. In small studies such as the one by Mingrone et al,¹ it can be difficult to see the full extent of surgical complications. The much larger SOS study found low mortality rates (0.25%). But 13% of those who underwent bariatric surgery had postoperative complications (number needed to harm = 8), with 2.2% of patients requiring reoperation.⁴ Additionally, women who become pregnant after bariatric surgery are at increased risk for internal hernias or bowel obstruction during pregnancy.⁸

Furthermore, malabsorptive-type surgeries are known to cause nutritional deficiencies, leading to disorders including anemia and osteoporosis.⁶ Importantly, while women of childbearing-age who undergo bariatric surgery decrease their risk of developing gestational hypertension and gestational diabetes, they are more likely to have nutritional deficiencies during pregnancy and to have children with these deficiencies.⁸

CHALLENGES TO IMPLEMENTATION: The ideal candidate remains unclear

It is still not clear from this study which patients should be referred for bariatric surgery. Historically, BMI has been used as the main indication for bariatric surgery, but this and other, studies have found that remission of diabetes is independent of BMI and the amount of weight lost.⁹ A predictive 10-point Diabetes Surgery Score has recently been developed: It uses age, BMI, duration of diabetes, and C-peptide levels to predict the likelihood of diabetes remission after surgery.¹⁰ This scoring system has yet to be validated in non-Asian patients, and a threshold for recommending surgery has been not established. However, this tool indicates that younger patients with a shorter duration of diabetes (which was not a factor in the outcome of the Mingrone study) and no baseline use of insulin are most likely to benefit from surgery. Thus, these patients may be the ones we need to consider referring first.

Cost of surgery. Several studies have shown that bariatric surgery is cost-effective for the treatment of diabetes, and saves money after approximately 5 years.^11,12 However, patients with diabetes and obesity may be uninsured or underinsured, and have high out-of-pocket costs. One challenge will be to ensure that surgery is a viable option for patients with financial constraints.

Acknowledgement

The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

ILLUSTRATIVE CASE

A 43-year-old woman with a body mass index (BMI) of 38 kg/m² and a 5-year history of diabetes has a glycated hemoglobin (HbA1c) of 8.5% despite the use of oral hypoglycemic agents. Should you talk to her about gastric bypass surgery to treat her diabetes?

Diet and exercise are the first steps in addressing diabetes, but these interventions are often unsuccessful. The International Diabetes Federation (IDF) recommends consideration of bariatric surgery for patients who have a BMI >35 kg/m² and diabetes that lifestyle modification and pharmacotherapy have failed to control.²

Surgery for diabetes: Is there ample evidence?
Until recently, the IDF’s recommendation was based on observational data and a single RCT that found increased resolution of diabetes following various bariatric procedures.^3-5 In the study detailed in this PURL, Mingrone et al took another look.

STUDY SUMMARY: Surgery led to higher remission rates

This single-center, nonblinded RCT compared 2 malabsorptive procedures— Roux-en-Y gastric bypass and biliopancreatic diversion, a more complicated procedure not commonly performed—with medical therapy.¹ The primary outcome was the rate of diabetes remission at 2 years, defined as a fasting glucose level <100 mg/dL and an HbA1c <6.5%. Changes in BMI and cholesterol levels were among the secondary endpoints.

To be eligible, patients had to be between the ages of 30 and 60 years and have a BMI ≥35 kg/m², a history of type 2 diabetes ≥5 years, and an HbA1c ≥7.0%. Exclusion criteria included a history of type 1 diabetes, diabetes caused by an underlying disease or steroid treatment, previous bariatric surgery, pregnancy, diabetic complications, other severe medical conditions, and acute hospitalization. Both the gastric bypass and biliopancreatic diversion procedures were performed by independent surgical teams.

Participants (N=60) were evaluated at baseline and at 1, 3, 6, 9, 12, and 24 months after the intervention by a team that included a dietician, a nurse, and a physician. All received a diet plan with daily exercise designed by their team. Those in the medical therapy group had their medications titrated to reach a goal HbA1c <7%. Pharmacotherapy was stopped based on normalization of blood sugars or HbA1c <6.5%.

Within 15 days postsurgery, patients in both surgical arms had their diabetes medications stopped based on their blood glucose levels.

At 2 years, 75% of the patients in the gastric bypass arm and 95% of the patients in the biliopancreatic diversion arm (number needed to treat=1.3 and 1, respectively) were considered to be in diabetes remission, defined as a fasting blood sugar of <100 mg/dL and an HbA1c <6.5% after one year without pharmacotherapy. (Notably, this differs from that of the American Diabetes Association, which requires an HbA1c <6.0% for classification as complete remission.) None of the patients in the medical therapy arm was in remission at the 2-year mark.

On average, blood sugars normalized for gastric bypass patients by 10±2 months, vs 4±1 months for biliopancreatic diversion patients (P=.01). The average HbA1c at the end of 2 years was significantly different among all 3 groups (6.35%±1.42 for those undergoing gastric bypass, 4.95%±0.49 for the biliopancreatic diversion group, and 7.69%±0.57 for the medical therapy group), as was the change in HbA1c from baseline (TABLE). Changes in BMI and the number of patients who achieved normalization of total cholesterol were similar for both surgical groups. Interestingly, neither baseline BMI nor amount of weight lost or pre-enrollment duration of diabetes were predictors of diabetes remission or normalization of fasting glucose levels.

TABLE
Surgery vs medical therapy for diabetes: Gastric bypass and biliopancreatic diversion are more effective

There were no deaths associated with this study. There were 2 adverse events requiring reoperation: an incisional hernia in a patient in the biliopancreatic diversion group and an intestinal obstruction in a patient in the gastric bypass group. Six patients in the biliopancreatic diversion arm developed metabolic abnormalities, including iron deficiency anemia, hypoalbuminemia, osteopenia, and osteoporosis. In the gastric bypass arm, 2 patients developed iron deficiency anemia.

WHAT’S NEW?: Evidence of efficacy has grown

This is the first RCT to evaluate biliopancreatic diversion and only the second to evaluate gastric bypass as strategies for controlling diabetes. Similar findings were demonstrated at 12 months in an RCT of 150 obese patients with diabetes in which intensive medical therapy was compared with either gastric bypass or sleeve gastrectomy,⁶ published concurrently with the Mingrone study. Like the Mingrone study, this study found that for select patients with diabetes, surgery may lead to better outcomes than medical management alone.

CAVEATS: Long-term effect is still uncertain

The long-term efficacy of surgery as a way to manage diabetes remains uncertain. Patients in this study were followed for just 2 years and the outcomes were metabolic measures rather than morbidity and mortality. A recent prospective observational study following patients for 6 years after gastric bypass found that the rate of remission for diabetes was 75% (95% confidence interval (CI), 63%-87%) at 2 years but dropped to 62% (95% CI, 49%-75%) at 6 years⁷

A larger study (N=4047) of longer duration—the Swedish Obese Subjects (SOS) cohort study —found a considerably larger drop: The diabetes remission rate for those who had surgery went from 72% at 2 years to 36% at 10-year follow-up, but that was still higher than the 10-year remission rate (13%) for the matched controls.⁴ It is still not clear exactly how long diabetic remission lasts after bariatric surgery or what effect a 10-year respite from the disease will have on the long-term morbidity and mortality of patients with diabetes.

Surgical risks. In small studies such as the one by Mingrone et al,¹ it can be difficult to see the full extent of surgical complications. The much larger SOS study found low mortality rates (0.25%). But 13% of those who underwent bariatric surgery had postoperative complications (number needed to harm = 8), with 2.2% of patients requiring reoperation.⁴ Additionally, women who become pregnant after bariatric surgery are at increased risk for internal hernias or bowel obstruction during pregnancy.⁸

Furthermore, malabsorptive-type surgeries are known to cause nutritional deficiencies, leading to disorders including anemia and osteoporosis.⁶ Importantly, while women of childbearing-age who undergo bariatric surgery decrease their risk of developing gestational hypertension and gestational diabetes, they are more likely to have nutritional deficiencies during pregnancy and to have children with these deficiencies.⁸

CHALLENGES TO IMPLEMENTATION: The ideal candidate remains unclear

It is still not clear from this study which patients should be referred for bariatric surgery. Historically, BMI has been used as the main indication for bariatric surgery, but this and other, studies have found that remission of diabetes is independent of BMI and the amount of weight lost.⁹ A predictive 10-point Diabetes Surgery Score has recently been developed: It uses age, BMI, duration of diabetes, and C-peptide levels to predict the likelihood of diabetes remission after surgery.¹⁰ This scoring system has yet to be validated in non-Asian patients, and a threshold for recommending surgery has been not established. However, this tool indicates that younger patients with a shorter duration of diabetes (which was not a factor in the outcome of the Mingrone study) and no baseline use of insulin are most likely to benefit from surgery. Thus, these patients may be the ones we need to consider referring first.

Cost of surgery. Several studies have shown that bariatric surgery is cost-effective for the treatment of diabetes, and saves money after approximately 5 years.^11,12 However, patients with diabetes and obesity may be uninsured or underinsured, and have high out-of-pocket costs. One challenge will be to ensure that surgery is a viable option for patients with financial constraints.

Acknowledgement

The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

Practice Changer
Treat low-risk patients with pulmonary embolism (PE) with low-molecular-weight heparin (LMWH) in an outpatient setting.¹

Strength of recommendation
B: Based on one good quality randomized controlled trial (RCT).

ILLUSTRATIVE CASE
Three months after undergoing surgical repair of an ankle fracture, a 50-year-old woman presents with acute-onset dyspnea at rest and pleuritic chest pain. Her left calf is tender and swollen. The patient has a history of hypertension and smokes about 10 cigarettes per day. Her temperature is 37ºC; ventricular rate, 98 beats/min; blood pressure, 135/85 mm Hg; respiratory rate, 25 breaths/min; and pulse oximetry, 92%. Spiral CT reveals a contrast filling defect indicative of a PE. Her score on the Pulmonary Embolism Severity Index (PESI) is 50, an indication of low risk. She wants to know if she can be treated at home. What should you tell her?

In the past, intravenous unfractionated heparin, administered in an inpatient setting, was the recommended initial anticoagulation therapy for patients with venous thromboembolism (VTE). LMWH, which can be administered subcutaneously and does not require laboratory monitoring, has made it possible to treat VTE without

hospitalization.

Outpatient PE care hindered

by lack of evidence
Guidelines from the American College of Physicians, the American Academy of Family Physicians, and the British Thoracic Society recommend outpatient treatment of deep vein thrombosis with LMWH, which they find to be safe and cost-effective for select patients.^2,3 Until recently, the safety and efficacy of outpatient management of PE has been less clear.

The lack of an accurate prediction tool to identify patients who could be treated safely outside the hospital was one barrier to the development of evidence-based recommendations for outpatient PE treatment. In 2005, the PESI,⁴ a validated tool that identifies patients with low risk for death from PE, was developed. Until recently, the absence of an RCT comparing inpatient and outpatient treatment for acute PE was another barrier.

STUDY SUMMARY
Outpatient treatment measures up
The Outpatient Treatment of Pulmonary Embolism (OTPE) study compared outpatient vs inpatient treatment of low-risk patients with acute PE. Participants had to be 18 or older, have acute symptomatic and objectively verified PE, and be at low risk for death based on the PESI score.4 In addition to excluding patients at moderate or high risk, the researchers identified 14 other exclusion criteria, including hypoxia, chest pain requiring opiates, and high risk for bleeding.

Patients were randomly assigned to the outpatient (n = 171) or inpatient (n = 168) group. Both groups received subcutaneous LMWH (enoxaparin, 1 mg/kg bid) for ≥ 5 days, followed by oral anticoagulation with a vitamin K antagonist for ≥ 90 days. Patients in the outpatient group were discharged from the emergency department (ED) within

24 hours of randomization, after being trained by a nurse to self-inject. Therapy after discharge was managed either by the patient's primary care physician or the hospital's anticoagulation staff.

The LMWH was discontinued in patients with an INR ≥ 2.0 for two consecutive days. All patients were followed for 90 days and contacted by the study team daily for the first week and then at 14, 30, 60, and 90 days. On each occasion, participants were asked about symptoms of recurrent VTE, bleeding, and the use of health care resources.

The primary outcome was the recurrence of symptomatic, objectively confirmed VTE within the study period. Secondary outcomes were major bleeding and all-cause mortality. Outcomes were confirmed by clinicians who were unaware of treatment

assignments.

Patients were also asked to rate their overall satisfaction with their care and their treatment preference 14 days after randomization, using a 5-point Likert questionnaire. Prior to the trial, the investigators decided that outpatient treatment would be considered noninferior to inpatient care if the difference between rates of recurrent VTE did not exceed 4%, a measure used in previous studies comparing treatment regimens for VTE and outpatient versus inpatient treatment of DVT.^5,6

Little difference in readmission rates, ED or office visits
One in 171 outpatients (0.6%) and none of the inpatients had recurrent VTE. Two outpatients (1.2%)—and no inpatients—developed major bleeding within 14 days, the result of intramuscular hematomas that occurred on days 3 and 13. There was one additional bleeding event (menometrorrhagia) in the outpatient group on day 50, but it was believed to be unrelated to the PE treatment. Per-protocol analysis, a more conservative measure used in noninferiority studies, found a difference in major bleeding rates of 3.8%. One person in each group died of non-VTE and nontreatment-related causes.

Almost all participants (99%) completed the satisfaction survey, which indicated that 92% of outpatients and 95% of inpatients were satisfied or very satisfied with their care. Hospital readmission rates, ED visits, and visits to primary care physicians were similar, with no significant differences between the

two groups. The mean time spent in the hospital was 0.5 days for outpatients and 3.9 days for inpatients. Fourteen percent of outpatients and 6% of inpatients received home nursing visits for enoxaparin injection. The total number of home visits was higher among outpatients (348 vs 105). Because both groups had extreme outliers, however, this difference was not statistically significant.

WHAT'S NEW
It's safe to keep

low-risk patients at home
This is the first RCT comparing the safety and effectiveness of outpatient and inpatient treatment of acute, symptomatic PE. Results were statistically comparable, and patients were satisfied being treated at home. Outpatient treatment was less expensive because of the shorter length of stay (0.5 vs 3.9 days) and was associated with the same rates of hospital readmission, ED visits, and visits to primary care physicians. There were more home nursing visits in the outpatient treatment group. But even if you assume a cost of $200 per home visit, the additional cost would be about $282 per individual in the outpatient group—significantly less than the cost of the additional 3.4 days in the hospital for each individual in the inpatient group.

The study also confirmed that the PESI accurately identifies low-risk patients with PE who can be treated in an outpatient setting. Thirty percent of patients who were screened for the OTPE trial met the low-risk eligibility requirement.

CAVEATS
Use of risk assessment tool is essential
The average age of patients in this study was 47 in the outpatient group and 49 in the inpatient group. In addition, only 1% to 3% of the patients were diagnosed with cancer. Older patients who have both cancer and PE would be unlikely to qualify for outpatient care.

Clinicians applying this practice changer should use the PESI to ensure that outpatient treatment for PE is used only for individuals at low risk.

CHALLENGES TO IMPLEMENTATION
ED coordination, training, and home care won't be easy
This practice changer may be difficult for primary care providers, who might not be included in emergency physicians' decisions regarding the appropriate treatment for acute PE. In this study, primary care physicians were notified of the randomized treatment plan for their patients, and 17 potential participants were excluded from the trial because of their doctors' opposition.

Outpatient management should be considered only if arrangements for adequate home nursing care can be made, if needed—and only for patients who are able to follow instructions and self-inject LMWH. Newer anticoagulation medications that are either injected once a day or taken orally might decrease the need for home nursing visits.  

REFERENCES
1. Aujesky D, Roy PM, Verschuren F, et al. Outpatient versus inpatient treatment for patients with acute pulmonary embolism: an international, open-label, randomised, non-inferiority trial. Lancet. 2011; 378:41-48.

2. Snow V, Qaseem A, Barry P, et al. Management of venous thromboembolism: a clinical practice guideline from the American College of Physicians and the American Academy of Family Physicians. Ann Intern Med. 2007;146:204-210.

3. British Thoracic Society Standards of Care Committee Pulmonary Embolism Guideline Development Group. British Thoracic Society guidelines for the management of suspected acute pulmonary embolism. Thorax. 2003;58: 470-483.

4.  Aujesky D, Obrosky DS, Stone RA, et al. Derivation and validation of a prognostic model for pulmonary embolism. Am J Respir Crit Care Med. 2005;172:1041-1046.

5. Koopman MM, Prandoni P, Piovella F, et al. Treatment of venous thrombosis with intravenous unfractionated heparin administered in the hospital as compared with subcutaneous low-molecular-weight heparin administered at home. The Tasman Study Group. N Engl J Med. 1996;334: 682-687.

6. Schulman S, Kearon C, Kakkar AK, et al. Dabigatran versus warfarin in the treatment of acute venous thromboembolism. N Engl J Med. 2009; 361:2342-2352.

ACKNOWLEDGEMENT
The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

Copyright © 2012 The Family Physicians Inquiries Network. All rights reserved.

Reprinted with permission from the Family Physicians Inquiries Network and The Journal of Family Practice. 2012;61(6):349-352.

Practice Changer
Treat low-risk patients with pulmonary embolism (PE) with low-molecular-weight heparin (LMWH) in an outpatient setting.¹

Strength of recommendation
B: Based on one good quality randomized controlled trial (RCT).

ILLUSTRATIVE CASE
Three months after undergoing surgical repair of an ankle fracture, a 50-year-old woman presents with acute-onset dyspnea at rest and pleuritic chest pain. Her left calf is tender and swollen. The patient has a history of hypertension and smokes about 10 cigarettes per day. Her temperature is 37ºC; ventricular rate, 98 beats/min; blood pressure, 135/85 mm Hg; respiratory rate, 25 breaths/min; and pulse oximetry, 92%. Spiral CT reveals a contrast filling defect indicative of a PE. Her score on the Pulmonary Embolism Severity Index (PESI) is 50, an indication of low risk. She wants to know if she can be treated at home. What should you tell her?

In the past, intravenous unfractionated heparin, administered in an inpatient setting, was the recommended initial anticoagulation therapy for patients with venous thromboembolism (VTE). LMWH, which can be administered subcutaneously and does not require laboratory monitoring, has made it possible to treat VTE without

hospitalization.

Outpatient PE care hindered

by lack of evidence
Guidelines from the American College of Physicians, the American Academy of Family Physicians, and the British Thoracic Society recommend outpatient treatment of deep vein thrombosis with LMWH, which they find to be safe and cost-effective for select patients.^2,3 Until recently, the safety and efficacy of outpatient management of PE has been less clear.

The lack of an accurate prediction tool to identify patients who could be treated safely outside the hospital was one barrier to the development of evidence-based recommendations for outpatient PE treatment. In 2005, the PESI,⁴ a validated tool that identifies patients with low risk for death from PE, was developed. Until recently, the absence of an RCT comparing inpatient and outpatient treatment for acute PE was another barrier.

STUDY SUMMARY
Outpatient treatment measures up
The Outpatient Treatment of Pulmonary Embolism (OTPE) study compared outpatient vs inpatient treatment of low-risk patients with acute PE. Participants had to be 18 or older, have acute symptomatic and objectively verified PE, and be at low risk for death based on the PESI score.4 In addition to excluding patients at moderate or high risk, the researchers identified 14 other exclusion criteria, including hypoxia, chest pain requiring opiates, and high risk for bleeding.

Patients were randomly assigned to the outpatient (n = 171) or inpatient (n = 168) group. Both groups received subcutaneous LMWH (enoxaparin, 1 mg/kg bid) for ≥ 5 days, followed by oral anticoagulation with a vitamin K antagonist for ≥ 90 days. Patients in the outpatient group were discharged from the emergency department (ED) within

24 hours of randomization, after being trained by a nurse to self-inject. Therapy after discharge was managed either by the patient's primary care physician or the hospital's anticoagulation staff.

The LMWH was discontinued in patients with an INR ≥ 2.0 for two consecutive days. All patients were followed for 90 days and contacted by the study team daily for the first week and then at 14, 30, 60, and 90 days. On each occasion, participants were asked about symptoms of recurrent VTE, bleeding, and the use of health care resources.

The primary outcome was the recurrence of symptomatic, objectively confirmed VTE within the study period. Secondary outcomes were major bleeding and all-cause mortality. Outcomes were confirmed by clinicians who were unaware of treatment

assignments.

Patients were also asked to rate their overall satisfaction with their care and their treatment preference 14 days after randomization, using a 5-point Likert questionnaire. Prior to the trial, the investigators decided that outpatient treatment would be considered noninferior to inpatient care if the difference between rates of recurrent VTE did not exceed 4%, a measure used in previous studies comparing treatment regimens for VTE and outpatient versus inpatient treatment of DVT.^5,6

Little difference in readmission rates, ED or office visits
One in 171 outpatients (0.6%) and none of the inpatients had recurrent VTE. Two outpatients (1.2%)—and no inpatients—developed major bleeding within 14 days, the result of intramuscular hematomas that occurred on days 3 and 13. There was one additional bleeding event (menometrorrhagia) in the outpatient group on day 50, but it was believed to be unrelated to the PE treatment. Per-protocol analysis, a more conservative measure used in noninferiority studies, found a difference in major bleeding rates of 3.8%. One person in each group died of non-VTE and nontreatment-related causes.

Almost all participants (99%) completed the satisfaction survey, which indicated that 92% of outpatients and 95% of inpatients were satisfied or very satisfied with their care. Hospital readmission rates, ED visits, and visits to primary care physicians were similar, with no significant differences between the

two groups. The mean time spent in the hospital was 0.5 days for outpatients and 3.9 days for inpatients. Fourteen percent of outpatients and 6% of inpatients received home nursing visits for enoxaparin injection. The total number of home visits was higher among outpatients (348 vs 105). Because both groups had extreme outliers, however, this difference was not statistically significant.

WHAT'S NEW
It's safe to keep

low-risk patients at home
This is the first RCT comparing the safety and effectiveness of outpatient and inpatient treatment of acute, symptomatic PE. Results were statistically comparable, and patients were satisfied being treated at home. Outpatient treatment was less expensive because of the shorter length of stay (0.5 vs 3.9 days) and was associated with the same rates of hospital readmission, ED visits, and visits to primary care physicians. There were more home nursing visits in the outpatient treatment group. But even if you assume a cost of $200 per home visit, the additional cost would be about $282 per individual in the outpatient group—significantly less than the cost of the additional 3.4 days in the hospital for each individual in the inpatient group.

The study also confirmed that the PESI accurately identifies low-risk patients with PE who can be treated in an outpatient setting. Thirty percent of patients who were screened for the OTPE trial met the low-risk eligibility requirement.

CAVEATS
Use of risk assessment tool is essential
The average age of patients in this study was 47 in the outpatient group and 49 in the inpatient group. In addition, only 1% to 3% of the patients were diagnosed with cancer. Older patients who have both cancer and PE would be unlikely to qualify for outpatient care.

Clinicians applying this practice changer should use the PESI to ensure that outpatient treatment for PE is used only for individuals at low risk.

CHALLENGES TO IMPLEMENTATION
ED coordination, training, and home care won't be easy
This practice changer may be difficult for primary care providers, who might not be included in emergency physicians' decisions regarding the appropriate treatment for acute PE. In this study, primary care physicians were notified of the randomized treatment plan for their patients, and 17 potential participants were excluded from the trial because of their doctors' opposition.

Outpatient management should be considered only if arrangements for adequate home nursing care can be made, if needed—and only for patients who are able to follow instructions and self-inject LMWH. Newer anticoagulation medications that are either injected once a day or taken orally might decrease the need for home nursing visits.  

REFERENCES
1. Aujesky D, Roy PM, Verschuren F, et al. Outpatient versus inpatient treatment for patients with acute pulmonary embolism: an international, open-label, randomised, non-inferiority trial. Lancet. 2011; 378:41-48.

2. Snow V, Qaseem A, Barry P, et al. Management of venous thromboembolism: a clinical practice guideline from the American College of Physicians and the American Academy of Family Physicians. Ann Intern Med. 2007;146:204-210.

3. British Thoracic Society Standards of Care Committee Pulmonary Embolism Guideline Development Group. British Thoracic Society guidelines for the management of suspected acute pulmonary embolism. Thorax. 2003;58: 470-483.

4.  Aujesky D, Obrosky DS, Stone RA, et al. Derivation and validation of a prognostic model for pulmonary embolism. Am J Respir Crit Care Med. 2005;172:1041-1046.

5. Koopman MM, Prandoni P, Piovella F, et al. Treatment of venous thrombosis with intravenous unfractionated heparin administered in the hospital as compared with subcutaneous low-molecular-weight heparin administered at home. The Tasman Study Group. N Engl J Med. 1996;334: 682-687.

6. Schulman S, Kearon C, Kakkar AK, et al. Dabigatran versus warfarin in the treatment of acute venous thromboembolism. N Engl J Med. 2009; 361:2342-2352.

ACKNOWLEDGEMENT
The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

Copyright © 2012 The Family Physicians Inquiries Network. All rights reserved.

Reprinted with permission from the Family Physicians Inquiries Network and The Journal of Family Practice. 2012;61(6):349-352.

Practice Changer
Treat low-risk patients with pulmonary embolism (PE) with low-molecular-weight heparin (LMWH) in an outpatient setting.¹

Strength of recommendation
B: Based on one good quality randomized controlled trial (RCT).

ILLUSTRATIVE CASE
Three months after undergoing surgical repair of an ankle fracture, a 50-year-old woman presents with acute-onset dyspnea at rest and pleuritic chest pain. Her left calf is tender and swollen. The patient has a history of hypertension and smokes about 10 cigarettes per day. Her temperature is 37ºC; ventricular rate, 98 beats/min; blood pressure, 135/85 mm Hg; respiratory rate, 25 breaths/min; and pulse oximetry, 92%. Spiral CT reveals a contrast filling defect indicative of a PE. Her score on the Pulmonary Embolism Severity Index (PESI) is 50, an indication of low risk. She wants to know if she can be treated at home. What should you tell her?

In the past, intravenous unfractionated heparin, administered in an inpatient setting, was the recommended initial anticoagulation therapy for patients with venous thromboembolism (VTE). LMWH, which can be administered subcutaneously and does not require laboratory monitoring, has made it possible to treat VTE without

hospitalization.

Outpatient PE care hindered

by lack of evidence
Guidelines from the American College of Physicians, the American Academy of Family Physicians, and the British Thoracic Society recommend outpatient treatment of deep vein thrombosis with LMWH, which they find to be safe and cost-effective for select patients.^2,3 Until recently, the safety and efficacy of outpatient management of PE has been less clear.

The lack of an accurate prediction tool to identify patients who could be treated safely outside the hospital was one barrier to the development of evidence-based recommendations for outpatient PE treatment. In 2005, the PESI,⁴ a validated tool that identifies patients with low risk for death from PE, was developed. Until recently, the absence of an RCT comparing inpatient and outpatient treatment for acute PE was another barrier.

STUDY SUMMARY
Outpatient treatment measures up
The Outpatient Treatment of Pulmonary Embolism (OTPE) study compared outpatient vs inpatient treatment of low-risk patients with acute PE. Participants had to be 18 or older, have acute symptomatic and objectively verified PE, and be at low risk for death based on the PESI score.4 In addition to excluding patients at moderate or high risk, the researchers identified 14 other exclusion criteria, including hypoxia, chest pain requiring opiates, and high risk for bleeding.

Patients were randomly assigned to the outpatient (n = 171) or inpatient (n = 168) group. Both groups received subcutaneous LMWH (enoxaparin, 1 mg/kg bid) for ≥ 5 days, followed by oral anticoagulation with a vitamin K antagonist for ≥ 90 days. Patients in the outpatient group were discharged from the emergency department (ED) within

24 hours of randomization, after being trained by a nurse to self-inject. Therapy after discharge was managed either by the patient's primary care physician or the hospital's anticoagulation staff.

The LMWH was discontinued in patients with an INR ≥ 2.0 for two consecutive days. All patients were followed for 90 days and contacted by the study team daily for the first week and then at 14, 30, 60, and 90 days. On each occasion, participants were asked about symptoms of recurrent VTE, bleeding, and the use of health care resources.

The primary outcome was the recurrence of symptomatic, objectively confirmed VTE within the study period. Secondary outcomes were major bleeding and all-cause mortality. Outcomes were confirmed by clinicians who were unaware of treatment

assignments.

Patients were also asked to rate their overall satisfaction with their care and their treatment preference 14 days after randomization, using a 5-point Likert questionnaire. Prior to the trial, the investigators decided that outpatient treatment would be considered noninferior to inpatient care if the difference between rates of recurrent VTE did not exceed 4%, a measure used in previous studies comparing treatment regimens for VTE and outpatient versus inpatient treatment of DVT.^5,6

Little difference in readmission rates, ED or office visits
One in 171 outpatients (0.6%) and none of the inpatients had recurrent VTE. Two outpatients (1.2%)—and no inpatients—developed major bleeding within 14 days, the result of intramuscular hematomas that occurred on days 3 and 13. There was one additional bleeding event (menometrorrhagia) in the outpatient group on day 50, but it was believed to be unrelated to the PE treatment. Per-protocol analysis, a more conservative measure used in noninferiority studies, found a difference in major bleeding rates of 3.8%. One person in each group died of non-VTE and nontreatment-related causes.

Almost all participants (99%) completed the satisfaction survey, which indicated that 92% of outpatients and 95% of inpatients were satisfied or very satisfied with their care. Hospital readmission rates, ED visits, and visits to primary care physicians were similar, with no significant differences between the

two groups. The mean time spent in the hospital was 0.5 days for outpatients and 3.9 days for inpatients. Fourteen percent of outpatients and 6% of inpatients received home nursing visits for enoxaparin injection. The total number of home visits was higher among outpatients (348 vs 105). Because both groups had extreme outliers, however, this difference was not statistically significant.

WHAT'S NEW
It's safe to keep

low-risk patients at home
This is the first RCT comparing the safety and effectiveness of outpatient and inpatient treatment of acute, symptomatic PE. Results were statistically comparable, and patients were satisfied being treated at home. Outpatient treatment was less expensive because of the shorter length of stay (0.5 vs 3.9 days) and was associated with the same rates of hospital readmission, ED visits, and visits to primary care physicians. There were more home nursing visits in the outpatient treatment group. But even if you assume a cost of $200 per home visit, the additional cost would be about $282 per individual in the outpatient group—significantly less than the cost of the additional 3.4 days in the hospital for each individual in the inpatient group.

The study also confirmed that the PESI accurately identifies low-risk patients with PE who can be treated in an outpatient setting. Thirty percent of patients who were screened for the OTPE trial met the low-risk eligibility requirement.

CAVEATS
Use of risk assessment tool is essential
The average age of patients in this study was 47 in the outpatient group and 49 in the inpatient group. In addition, only 1% to 3% of the patients were diagnosed with cancer. Older patients who have both cancer and PE would be unlikely to qualify for outpatient care.

Clinicians applying this practice changer should use the PESI to ensure that outpatient treatment for PE is used only for individuals at low risk.

CHALLENGES TO IMPLEMENTATION
ED coordination, training, and home care won't be easy
This practice changer may be difficult for primary care providers, who might not be included in emergency physicians' decisions regarding the appropriate treatment for acute PE. In this study, primary care physicians were notified of the randomized treatment plan for their patients, and 17 potential participants were excluded from the trial because of their doctors' opposition.

Outpatient management should be considered only if arrangements for adequate home nursing care can be made, if needed—and only for patients who are able to follow instructions and self-inject LMWH. Newer anticoagulation medications that are either injected once a day or taken orally might decrease the need for home nursing visits.  

REFERENCES
1. Aujesky D, Roy PM, Verschuren F, et al. Outpatient versus inpatient treatment for patients with acute pulmonary embolism: an international, open-label, randomised, non-inferiority trial. Lancet. 2011; 378:41-48.

2. Snow V, Qaseem A, Barry P, et al. Management of venous thromboembolism: a clinical practice guideline from the American College of Physicians and the American Academy of Family Physicians. Ann Intern Med. 2007;146:204-210.

3. British Thoracic Society Standards of Care Committee Pulmonary Embolism Guideline Development Group. British Thoracic Society guidelines for the management of suspected acute pulmonary embolism. Thorax. 2003;58: 470-483.

4.  Aujesky D, Obrosky DS, Stone RA, et al. Derivation and validation of a prognostic model for pulmonary embolism. Am J Respir Crit Care Med. 2005;172:1041-1046.

5. Koopman MM, Prandoni P, Piovella F, et al. Treatment of venous thrombosis with intravenous unfractionated heparin administered in the hospital as compared with subcutaneous low-molecular-weight heparin administered at home. The Tasman Study Group. N Engl J Med. 1996;334: 682-687.

6. Schulman S, Kearon C, Kakkar AK, et al. Dabigatran versus warfarin in the treatment of acute venous thromboembolism. N Engl J Med. 2009; 361:2342-2352.

ACKNOWLEDGEMENT
The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

Copyright © 2012 The Family Physicians Inquiries Network. All rights reserved.

Reprinted with permission from the Family Physicians Inquiries Network and The Journal of Family Practice. 2012;61(6):349-352.

Practice Changer
Stop prescribing amoxicillin for acute rhinosinusitis. It’s unlikely to provide a speedier recovery than OTC remedies alone.¹

Strength of recommendation
B: Based on a single high-quality randomized controlled trial.

ILLUSTRATIVE CASE
A 28-year-old man comes to your clinic after experiencing fatigue, purulent nasal discharge, and unilateral facial pain for nearly

10 days. Overall, he appears healthy, and you diagnose acute rhinosinusitis. You suggest OTC remedies for supportive care and wonder if a course of amoxicillin would speed his

recovery.

Each year, more than 30 million Americans—about one in seven adults—are diagnosed with sinusitis.² No more than 2% of these cases are thought to be bacterial.³

CDC guidelines for the diagnosis of acute bacterial rhinosinusitis include symptoms that last seven or more days, with maxillary pain or tenderness in the face or teeth and purulent nasal secretions.⁴ Patients with symptoms lasting less than seven days are unlikely to have a bacterial infection. But the nonspecific signs and symptoms included in the CDC guidelines limit their usefulness in determining whether the cause of the sinusitis is bacterial or viral on clinical grounds alone.

Most cases of sinusitis spontaneously resolve
In patients with acute bacterial sinusitis, the American Academy of Otolaryngology–Head and Neck Surgery (AAO–HNS) guidelines advocate watchful waiting and symptom relief with nasal oxymetazoline, pseudoephedrine, and saline nasal irrigation.³ The rate of spontaneous resolution is high: 80% of patients with clinically diagnosed sinusitis improve without treatment within two weeks.^1,5

Traditional decongestants and mucolytics have not demonstrated efficacy in resolving sinusitis, although rigorous evaluation is lacking. Other treatments, such as saline irrigation and intranasal corticosteroids, are of unclear benefit and need further study.^6-8

Lack of evidence has done little to curtail antibiotic use
A previous PURL that was based on a meta-analysis of antibiotic treatment trials for sinusitis recommended that we stop prescribing antibiotics for adults with acute sinusitis unless their symptoms are severe.^9,10 Yet antibiotics remain the mainstay of treatment.

Despite the AAO–HNS guidelines, evidence of spontaneous resolution, and accumulating data on the lack of efficacy of antimicrobials for sinusitis, 81% of patients diagnosed with acute sinusitis were given prescriptions for antibiotics, a study of primary care practices showed.¹¹ Frequent use of antibiotics contributes to high rates of drug resistance, and adverse events related to antibiotic use account for an estimated 142,500 emergency department visits annually.¹²

STUDY SUMMARY
Little benefit from amoxicillin, even for severe cases
Garbutt and colleagues revisited the issue, randomizing 166 patients from 10 primary care practices to receive amoxicillin plus symptomatic treatment or placebo plus symptomatic treatment for acute rhinosinusitis.¹

To be eligible for the study, patients had to be between the ages of 18 and 70, meet CDC diagnostic criteria for acute rhinosinusitis, and have moderate to very severe symptoms that were of seven- to-28-day duration and worsening or not improving, or of < 7-day duration but had worsened after an initial improvement. Exclusion criteria included complications from sinusitis, a history of allergy to penicillin or amoxicillin, antibiotic use in the past four weeks, comorbidities that impair immune function, cystic fibrosis, pregnancy, and mild symptoms.

Both groups had similar baseline characteristics, with participants who were predominantly white (79%) and female (64%). All the participants received a supply of symptomatic treatments: acetaminophen, guaifenesin, dextromethorphan, and sustained-release pseudoephedrine. The treatment group also received amoxicillin 1,500 mg/d, divided into three doses; the placebo group received identical-looking placebo pills.

Patients were assessed with the Sino-nasal Outcome Test-16 (SNOT-16), a validated measure that asks patients to assess both the severity and frequency of 16 sinus symptoms. SNOT-16 uses a 0-to-3 rating scale (0 = no problem; 3 = severe problem), with a clinically important difference of ≥ 0.5 on the mean score. The test was administered at enrollment and at days 3, 7, and 10. The disease-specific quality of life at day 3 was the primary outcome.

There was no statistically significantly difference in SNOT-16 scores between the amoxicillin and placebo groups on days 3 and 10. On day 7, there was a small statistically significant improvement in the amoxicillin group, but it did not reach the level of clinical importance (≥ 0.5) based on SNOT-16’s mean score.

The authors also asked participants to retrospectively assess symptom change since enrollment on a six-point scale. Those who reported that their symptoms were “a lot better” or “absent” were characterized as significantly improved. The results correlated with the data from the SNOT-16, showing no difference between the amoxicillin and control groups at days 3 and 10. On day 7, 74% of patients treated with amoxicillin self-

reported significant improvement in symptoms since the start of the study, versus 56% in the control group. The number needed to treat was 6 (95% confidence interval, 3 to 34; P = .02) for a reduction in symptoms at day 7.

Patients in both groups had similar rates of absenteeism, inability to perform usual activities, relapse and recurrence, and use of additional health care. Satisfaction with treatment was similar, as well.

No serious adverse effects occurred. Both groups reported similar frequencies (< 10%) of nausea, diarrhea, abdominal pain, or vaginitis.

WHAT’S NEW
Even severe sinusitis resolves without antibiotics
Previous studies recommended foregoing antibiotics for acute sinusitis, except when symptoms are severe. This study—in which more than half (52%) of patients in each group had symptoms rated severe or very severe—found no benefit to adding amoxicillin to supportive treatments.¹ Antibiotics did not shorten the duration of illness, prevent relapse and recurrence, or improve satisfaction with treatment. The researchers found a statistically significant difference between groups on day 7 of 0.19 points, but no clinically meaningful difference (≥ 0.5) based on the SNOT-16 mean score.

CAVEATS
Guidelines, risk of complications may give reason to pause
The 2012 Infectious Diseases Society of America guidelines recommend amoxicillin with clavulanic acid as empiric therapy for acute bacterial rhinosinusitis.⁷ The findings of the study by Garbutt et al—conducted at a time when the incidence of beta-lactamase-producing organisms was low and amoxicillin was the treatment of choice—suggest otherwise.

Serious complications of sinusitis, such as brain abscess, periorbital cellulitis, and meningitis, can occur, however. Patients who deteriorate clinically or develop high fever or severe headache require close follow-up, which may include further diagnostic evaluation or consultation with an otolaryngologist. Evidence is lacking as to whether antibiotics prevent such complications.⁵

CHALLENGES TO IMPLEMENTATION
Managing patient expectations
Many patients with symptoms of acute rhinosinusitis think they need an antibiotic. Managing their expectations and providing instructions about supportive treatments are time consuming and may be

difficult.

Nonetheless, we’re optimistic: We think that most patients today are aware of the problems associated with antibiotic resistance and wary of “superbugs,” and will therefore be receptive to this practice change. Clinicians can help by reminding patients of the adverse effects of antibiotics and the natural course of rhinosinusitis, as well as by offering symptomatic treatments.

REFERENCES
1. Garbutt J, Banister C, Spitznagel E, et al. Amoxicillin for acute rhinosinusitis: a randomized controlled trial. JAMA. 2012;307:685-692.

2. Centers for Disease Control and Prevention. Summary health statistics for US adults: National Health Interview Survey 2010. January 2012. www.cdc.gov/nchs/data/series/sr_10/sr10_252.pdf. Accessed July 9, 2012.

3. Rosenfeld RM, Andes D, Bhattacharyya N, et al. Clinical practice guideline: adult sinusitis. Otolaryngol Head Neck Surg. 2007:137(3 suppl):S1-S31.

4. Hickner JM, Bartlett JG, Besser RE, et al; American Academy of Family Physians; American College of Physicians; American Society of Internal Medicine; Centers for Disease Control; Infectious Disease Society of America. Principles of appropriate antibiotic use for acute rhinosinusitis in adults; background. Ann Intern Med. 2001;134:498-505.

5. Ahovuo-Saloranta A, Borisenk OV, Kovanen N, et al. Antibiotics for acute maxillary sinusitis. Cochrane Database Sys Rev. 2008(2):CD000243.

6. Allen G, Kelsberg G, Jankowski TA. Do nasal decongestants relieve symptoms? J Fam Pract. 2003;52:714-724.

7. Chow AW, Benninger MS, Brook I, et al. IDSA clinical practice guideline for acute bacterial rhinosinusitis in children and adults. Clin Infect Dis. 2012;54:e72-e112.

8. Zalmanovici A, Yaphe J. Intranasal steroids for acute sinusitis. Cochrane Database Syst Rev. 2009(4):CD005149.

9. Schumann A, Hickner, J. Patients insist on antibiotics for sinusitis? Here is a good reason to say “no.” J Fam Pract. 2008;57:464-468.

10. Young J, De Sutter A, Merenstein D, et al. Antibiotics for adults with clinically diagnosed acute rhinosinusitis: a meta-analysis of individual patient data. Lancet. 2008;371:908-914.

11. Gill JM, Fleischut P, Haas S. Use of antibiotics for adult upper respiratory infections in outpatient settings: a national ambulatory network study. Fam Med. 2006;38:349-354.

12. Shehab N, Patel PR, Srinivasan A, et al. Emergency department visits for antibiotic-associated adverse events. Clin Infect Dis. 2008; 47:735-743.

ACKNOWLEDGEMENT
The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

Copyright © 2012 The Family Physicians Inquiries Network. All rights reserved.

Reprinted with permission from the Family Physicians Inquiries Network and The Journal of Family Practice. 2012;61(10):610-612.

Practice Changer
Stop prescribing amoxicillin for acute rhinosinusitis. It’s unlikely to provide a speedier recovery than OTC remedies alone.¹

Strength of recommendation
B: Based on a single high-quality randomized controlled trial.

ILLUSTRATIVE CASE
A 28-year-old man comes to your clinic after experiencing fatigue, purulent nasal discharge, and unilateral facial pain for nearly

10 days. Overall, he appears healthy, and you diagnose acute rhinosinusitis. You suggest OTC remedies for supportive care and wonder if a course of amoxicillin would speed his

recovery.

Each year, more than 30 million Americans—about one in seven adults—are diagnosed with sinusitis.² No more than 2% of these cases are thought to be bacterial.³

CDC guidelines for the diagnosis of acute bacterial rhinosinusitis include symptoms that last seven or more days, with maxillary pain or tenderness in the face or teeth and purulent nasal secretions.⁴ Patients with symptoms lasting less than seven days are unlikely to have a bacterial infection. But the nonspecific signs and symptoms included in the CDC guidelines limit their usefulness in determining whether the cause of the sinusitis is bacterial or viral on clinical grounds alone.

Most cases of sinusitis spontaneously resolve
In patients with acute bacterial sinusitis, the American Academy of Otolaryngology–Head and Neck Surgery (AAO–HNS) guidelines advocate watchful waiting and symptom relief with nasal oxymetazoline, pseudoephedrine, and saline nasal irrigation.³ The rate of spontaneous resolution is high: 80% of patients with clinically diagnosed sinusitis improve without treatment within two weeks.^1,5

Traditional decongestants and mucolytics have not demonstrated efficacy in resolving sinusitis, although rigorous evaluation is lacking. Other treatments, such as saline irrigation and intranasal corticosteroids, are of unclear benefit and need further study.^6-8

Lack of evidence has done little to curtail antibiotic use
A previous PURL that was based on a meta-analysis of antibiotic treatment trials for sinusitis recommended that we stop prescribing antibiotics for adults with acute sinusitis unless their symptoms are severe.^9,10 Yet antibiotics remain the mainstay of treatment.

Despite the AAO–HNS guidelines, evidence of spontaneous resolution, and accumulating data on the lack of efficacy of antimicrobials for sinusitis, 81% of patients diagnosed with acute sinusitis were given prescriptions for antibiotics, a study of primary care practices showed.¹¹ Frequent use of antibiotics contributes to high rates of drug resistance, and adverse events related to antibiotic use account for an estimated 142,500 emergency department visits annually.¹²

STUDY SUMMARY
Little benefit from amoxicillin, even for severe cases
Garbutt and colleagues revisited the issue, randomizing 166 patients from 10 primary care practices to receive amoxicillin plus symptomatic treatment or placebo plus symptomatic treatment for acute rhinosinusitis.¹

To be eligible for the study, patients had to be between the ages of 18 and 70, meet CDC diagnostic criteria for acute rhinosinusitis, and have moderate to very severe symptoms that were of seven- to-28-day duration and worsening or not improving, or of < 7-day duration but had worsened after an initial improvement. Exclusion criteria included complications from sinusitis, a history of allergy to penicillin or amoxicillin, antibiotic use in the past four weeks, comorbidities that impair immune function, cystic fibrosis, pregnancy, and mild symptoms.

Both groups had similar baseline characteristics, with participants who were predominantly white (79%) and female (64%). All the participants received a supply of symptomatic treatments: acetaminophen, guaifenesin, dextromethorphan, and sustained-release pseudoephedrine. The treatment group also received amoxicillin 1,500 mg/d, divided into three doses; the placebo group received identical-looking placebo pills.

Patients were assessed with the Sino-nasal Outcome Test-16 (SNOT-16), a validated measure that asks patients to assess both the severity and frequency of 16 sinus symptoms. SNOT-16 uses a 0-to-3 rating scale (0 = no problem; 3 = severe problem), with a clinically important difference of ≥ 0.5 on the mean score. The test was administered at enrollment and at days 3, 7, and 10. The disease-specific quality of life at day 3 was the primary outcome.

There was no statistically significantly difference in SNOT-16 scores between the amoxicillin and placebo groups on days 3 and 10. On day 7, there was a small statistically significant improvement in the amoxicillin group, but it did not reach the level of clinical importance (≥ 0.5) based on SNOT-16’s mean score.

The authors also asked participants to retrospectively assess symptom change since enrollment on a six-point scale. Those who reported that their symptoms were “a lot better” or “absent” were characterized as significantly improved. The results correlated with the data from the SNOT-16, showing no difference between the amoxicillin and control groups at days 3 and 10. On day 7, 74% of patients treated with amoxicillin self-

reported significant improvement in symptoms since the start of the study, versus 56% in the control group. The number needed to treat was 6 (95% confidence interval, 3 to 34; P = .02) for a reduction in symptoms at day 7.

Patients in both groups had similar rates of absenteeism, inability to perform usual activities, relapse and recurrence, and use of additional health care. Satisfaction with treatment was similar, as well.

No serious adverse effects occurred. Both groups reported similar frequencies (< 10%) of nausea, diarrhea, abdominal pain, or vaginitis.

WHAT’S NEW
Even severe sinusitis resolves without antibiotics
Previous studies recommended foregoing antibiotics for acute sinusitis, except when symptoms are severe. This study—in which more than half (52%) of patients in each group had symptoms rated severe or very severe—found no benefit to adding amoxicillin to supportive treatments.¹ Antibiotics did not shorten the duration of illness, prevent relapse and recurrence, or improve satisfaction with treatment. The researchers found a statistically significant difference between groups on day 7 of 0.19 points, but no clinically meaningful difference (≥ 0.5) based on the SNOT-16 mean score.

CAVEATS
Guidelines, risk of complications may give reason to pause
The 2012 Infectious Diseases Society of America guidelines recommend amoxicillin with clavulanic acid as empiric therapy for acute bacterial rhinosinusitis.⁷ The findings of the study by Garbutt et al—conducted at a time when the incidence of beta-lactamase-producing organisms was low and amoxicillin was the treatment of choice—suggest otherwise.

Serious complications of sinusitis, such as brain abscess, periorbital cellulitis, and meningitis, can occur, however. Patients who deteriorate clinically or develop high fever or severe headache require close follow-up, which may include further diagnostic evaluation or consultation with an otolaryngologist. Evidence is lacking as to whether antibiotics prevent such complications.⁵

CHALLENGES TO IMPLEMENTATION
Managing patient expectations
Many patients with symptoms of acute rhinosinusitis think they need an antibiotic. Managing their expectations and providing instructions about supportive treatments are time consuming and may be

difficult.

Nonetheless, we’re optimistic: We think that most patients today are aware of the problems associated with antibiotic resistance and wary of “superbugs,” and will therefore be receptive to this practice change. Clinicians can help by reminding patients of the adverse effects of antibiotics and the natural course of rhinosinusitis, as well as by offering symptomatic treatments.

REFERENCES
1. Garbutt J, Banister C, Spitznagel E, et al. Amoxicillin for acute rhinosinusitis: a randomized controlled trial. JAMA. 2012;307:685-692.

2. Centers for Disease Control and Prevention. Summary health statistics for US adults: National Health Interview Survey 2010. January 2012. www.cdc.gov/nchs/data/series/sr_10/sr10_252.pdf. Accessed July 9, 2012.

3. Rosenfeld RM, Andes D, Bhattacharyya N, et al. Clinical practice guideline: adult sinusitis. Otolaryngol Head Neck Surg. 2007:137(3 suppl):S1-S31.

4. Hickner JM, Bartlett JG, Besser RE, et al; American Academy of Family Physians; American College of Physicians; American Society of Internal Medicine; Centers for Disease Control; Infectious Disease Society of America. Principles of appropriate antibiotic use for acute rhinosinusitis in adults; background. Ann Intern Med. 2001;134:498-505.

5. Ahovuo-Saloranta A, Borisenk OV, Kovanen N, et al. Antibiotics for acute maxillary sinusitis. Cochrane Database Sys Rev. 2008(2):CD000243.

6. Allen G, Kelsberg G, Jankowski TA. Do nasal decongestants relieve symptoms? J Fam Pract. 2003;52:714-724.

7. Chow AW, Benninger MS, Brook I, et al. IDSA clinical practice guideline for acute bacterial rhinosinusitis in children and adults. Clin Infect Dis. 2012;54:e72-e112.

8. Zalmanovici A, Yaphe J. Intranasal steroids for acute sinusitis. Cochrane Database Syst Rev. 2009(4):CD005149.

9. Schumann A, Hickner, J. Patients insist on antibiotics for sinusitis? Here is a good reason to say “no.” J Fam Pract. 2008;57:464-468.

10. Young J, De Sutter A, Merenstein D, et al. Antibiotics for adults with clinically diagnosed acute rhinosinusitis: a meta-analysis of individual patient data. Lancet. 2008;371:908-914.

11. Gill JM, Fleischut P, Haas S. Use of antibiotics for adult upper respiratory infections in outpatient settings: a national ambulatory network study. Fam Med. 2006;38:349-354.

12. Shehab N, Patel PR, Srinivasan A, et al. Emergency department visits for antibiotic-associated adverse events. Clin Infect Dis. 2008; 47:735-743.

ACKNOWLEDGEMENT
The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

Copyright © 2012 The Family Physicians Inquiries Network. All rights reserved.

Reprinted with permission from the Family Physicians Inquiries Network and The Journal of Family Practice. 2012;61(10):610-612.

Practice Changer
Stop prescribing amoxicillin for acute rhinosinusitis. It’s unlikely to provide a speedier recovery than OTC remedies alone.¹

Strength of recommendation
B: Based on a single high-quality randomized controlled trial.

ILLUSTRATIVE CASE
A 28-year-old man comes to your clinic after experiencing fatigue, purulent nasal discharge, and unilateral facial pain for nearly

10 days. Overall, he appears healthy, and you diagnose acute rhinosinusitis. You suggest OTC remedies for supportive care and wonder if a course of amoxicillin would speed his

recovery.

Each year, more than 30 million Americans—about one in seven adults—are diagnosed with sinusitis.² No more than 2% of these cases are thought to be bacterial.³

CDC guidelines for the diagnosis of acute bacterial rhinosinusitis include symptoms that last seven or more days, with maxillary pain or tenderness in the face or teeth and purulent nasal secretions.⁴ Patients with symptoms lasting less than seven days are unlikely to have a bacterial infection. But the nonspecific signs and symptoms included in the CDC guidelines limit their usefulness in determining whether the cause of the sinusitis is bacterial or viral on clinical grounds alone.

Most cases of sinusitis spontaneously resolve
In patients with acute bacterial sinusitis, the American Academy of Otolaryngology–Head and Neck Surgery (AAO–HNS) guidelines advocate watchful waiting and symptom relief with nasal oxymetazoline, pseudoephedrine, and saline nasal irrigation.³ The rate of spontaneous resolution is high: 80% of patients with clinically diagnosed sinusitis improve without treatment within two weeks.^1,5

Traditional decongestants and mucolytics have not demonstrated efficacy in resolving sinusitis, although rigorous evaluation is lacking. Other treatments, such as saline irrigation and intranasal corticosteroids, are of unclear benefit and need further study.^6-8

Lack of evidence has done little to curtail antibiotic use
A previous PURL that was based on a meta-analysis of antibiotic treatment trials for sinusitis recommended that we stop prescribing antibiotics for adults with acute sinusitis unless their symptoms are severe.^9,10 Yet antibiotics remain the mainstay of treatment.

Despite the AAO–HNS guidelines, evidence of spontaneous resolution, and accumulating data on the lack of efficacy of antimicrobials for sinusitis, 81% of patients diagnosed with acute sinusitis were given prescriptions for antibiotics, a study of primary care practices showed.¹¹ Frequent use of antibiotics contributes to high rates of drug resistance, and adverse events related to antibiotic use account for an estimated 142,500 emergency department visits annually.¹²

STUDY SUMMARY
Little benefit from amoxicillin, even for severe cases
Garbutt and colleagues revisited the issue, randomizing 166 patients from 10 primary care practices to receive amoxicillin plus symptomatic treatment or placebo plus symptomatic treatment for acute rhinosinusitis.¹

To be eligible for the study, patients had to be between the ages of 18 and 70, meet CDC diagnostic criteria for acute rhinosinusitis, and have moderate to very severe symptoms that were of seven- to-28-day duration and worsening or not improving, or of < 7-day duration but had worsened after an initial improvement. Exclusion criteria included complications from sinusitis, a history of allergy to penicillin or amoxicillin, antibiotic use in the past four weeks, comorbidities that impair immune function, cystic fibrosis, pregnancy, and mild symptoms.

Both groups had similar baseline characteristics, with participants who were predominantly white (79%) and female (64%). All the participants received a supply of symptomatic treatments: acetaminophen, guaifenesin, dextromethorphan, and sustained-release pseudoephedrine. The treatment group also received amoxicillin 1,500 mg/d, divided into three doses; the placebo group received identical-looking placebo pills.

Patients were assessed with the Sino-nasal Outcome Test-16 (SNOT-16), a validated measure that asks patients to assess both the severity and frequency of 16 sinus symptoms. SNOT-16 uses a 0-to-3 rating scale (0 = no problem; 3 = severe problem), with a clinically important difference of ≥ 0.5 on the mean score. The test was administered at enrollment and at days 3, 7, and 10. The disease-specific quality of life at day 3 was the primary outcome.

There was no statistically significantly difference in SNOT-16 scores between the amoxicillin and placebo groups on days 3 and 10. On day 7, there was a small statistically significant improvement in the amoxicillin group, but it did not reach the level of clinical importance (≥ 0.5) based on SNOT-16’s mean score.

The authors also asked participants to retrospectively assess symptom change since enrollment on a six-point scale. Those who reported that their symptoms were “a lot better” or “absent” were characterized as significantly improved. The results correlated with the data from the SNOT-16, showing no difference between the amoxicillin and control groups at days 3 and 10. On day 7, 74% of patients treated with amoxicillin self-

reported significant improvement in symptoms since the start of the study, versus 56% in the control group. The number needed to treat was 6 (95% confidence interval, 3 to 34; P = .02) for a reduction in symptoms at day 7.

Patients in both groups had similar rates of absenteeism, inability to perform usual activities, relapse and recurrence, and use of additional health care. Satisfaction with treatment was similar, as well.

No serious adverse effects occurred. Both groups reported similar frequencies (< 10%) of nausea, diarrhea, abdominal pain, or vaginitis.

WHAT’S NEW
Even severe sinusitis resolves without antibiotics
Previous studies recommended foregoing antibiotics for acute sinusitis, except when symptoms are severe. This study—in which more than half (52%) of patients in each group had symptoms rated severe or very severe—found no benefit to adding amoxicillin to supportive treatments.¹ Antibiotics did not shorten the duration of illness, prevent relapse and recurrence, or improve satisfaction with treatment. The researchers found a statistically significant difference between groups on day 7 of 0.19 points, but no clinically meaningful difference (≥ 0.5) based on the SNOT-16 mean score.

CAVEATS
Guidelines, risk of complications may give reason to pause
The 2012 Infectious Diseases Society of America guidelines recommend amoxicillin with clavulanic acid as empiric therapy for acute bacterial rhinosinusitis.⁷ The findings of the study by Garbutt et al—conducted at a time when the incidence of beta-lactamase-producing organisms was low and amoxicillin was the treatment of choice—suggest otherwise.

Serious complications of sinusitis, such as brain abscess, periorbital cellulitis, and meningitis, can occur, however. Patients who deteriorate clinically or develop high fever or severe headache require close follow-up, which may include further diagnostic evaluation or consultation with an otolaryngologist. Evidence is lacking as to whether antibiotics prevent such complications.⁵

CHALLENGES TO IMPLEMENTATION
Managing patient expectations
Many patients with symptoms of acute rhinosinusitis think they need an antibiotic. Managing their expectations and providing instructions about supportive treatments are time consuming and may be

difficult.

Nonetheless, we’re optimistic: We think that most patients today are aware of the problems associated with antibiotic resistance and wary of “superbugs,” and will therefore be receptive to this practice change. Clinicians can help by reminding patients of the adverse effects of antibiotics and the natural course of rhinosinusitis, as well as by offering symptomatic treatments.

REFERENCES
1. Garbutt J, Banister C, Spitznagel E, et al. Amoxicillin for acute rhinosinusitis: a randomized controlled trial. JAMA. 2012;307:685-692.

2. Centers for Disease Control and Prevention. Summary health statistics for US adults: National Health Interview Survey 2010. January 2012. www.cdc.gov/nchs/data/series/sr_10/sr10_252.pdf. Accessed July 9, 2012.

3. Rosenfeld RM, Andes D, Bhattacharyya N, et al. Clinical practice guideline: adult sinusitis. Otolaryngol Head Neck Surg. 2007:137(3 suppl):S1-S31.

4. Hickner JM, Bartlett JG, Besser RE, et al; American Academy of Family Physians; American College of Physicians; American Society of Internal Medicine; Centers for Disease Control; Infectious Disease Society of America. Principles of appropriate antibiotic use for acute rhinosinusitis in adults; background. Ann Intern Med. 2001;134:498-505.

5. Ahovuo-Saloranta A, Borisenk OV, Kovanen N, et al. Antibiotics for acute maxillary sinusitis. Cochrane Database Sys Rev. 2008(2):CD000243.

6. Allen G, Kelsberg G, Jankowski TA. Do nasal decongestants relieve symptoms? J Fam Pract. 2003;52:714-724.

7. Chow AW, Benninger MS, Brook I, et al. IDSA clinical practice guideline for acute bacterial rhinosinusitis in children and adults. Clin Infect Dis. 2012;54:e72-e112.

8. Zalmanovici A, Yaphe J. Intranasal steroids for acute sinusitis. Cochrane Database Syst Rev. 2009(4):CD005149.

9. Schumann A, Hickner, J. Patients insist on antibiotics for sinusitis? Here is a good reason to say “no.” J Fam Pract. 2008;57:464-468.

10. Young J, De Sutter A, Merenstein D, et al. Antibiotics for adults with clinically diagnosed acute rhinosinusitis: a meta-analysis of individual patient data. Lancet. 2008;371:908-914.

11. Gill JM, Fleischut P, Haas S. Use of antibiotics for adult upper respiratory infections in outpatient settings: a national ambulatory network study. Fam Med. 2006;38:349-354.

12. Shehab N, Patel PR, Srinivasan A, et al. Emergency department visits for antibiotic-associated adverse events. Clin Infect Dis. 2008; 47:735-743.

ACKNOWLEDGEMENT
The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

Copyright © 2012 The Family Physicians Inquiries Network. All rights reserved.

Reprinted with permission from the Family Physicians Inquiries Network and The Journal of Family Practice. 2012;61(10):610-612.