Alexander R. Carbo, MD

The “Things We Do for No Reason” (TWDFNR) series reviews practices that have become common parts of hospital care but may provide little value to our patients. Practices reviewed in the TWDFNR series do not represent “black and white” conclusions or clinical practice standards but are meant as a starting place for research and active discussions among hospitalists and patients. We invite you to be part of that discussion. https://www.choosingwisely.org/

A nticoagulation is the cornerstone of acute venous thromboembolism (VTE) management. Nonetheless, the use of inferior vena cava (IVC) filters in addition to anticoagulation is increasing, with wide variation in practice patterns and a growing recognition of filter-related complications. Rigorous randomized controlled data demonstrating that IVC filters, particularly the increasingly commonly placed retrievable filters, provide a mortality benefit are sparse. Given our review of IVC filter use and the lack of evidence demonstrating that IVC filters provide a mortality benefit, we recommend using anticoagulation alone for stable medical service patients admitted with acute VTE. In nuanced cases, hospitalists should engage in multidisciplinary care to develop individualized treatment options.

A 65-year-old woman with a history of diabetes mellitus, metastatic breast cancer, and peptic ulcer disease presents to the Emergency Department for the evaluation of right thigh swelling, chest pain, and dyspnea after a transcontinental flight. Physical examination is notable for a pulse of 114 beats per minute, blood pressure of 136/93 mm Hg, respiratory rate of 14 breaths per minute, oxygen saturation of 95% on room air, and swelling of the right thigh. Computerized tomography imaging demonstrates multiple bilateral pulmonary emboli. Emergency department physicians begin anticoagulation and inform you that they have ordered the placement of a retrievable inferior vena cava (IVC) filter.

Acute venous thromboembolism (VTE) accounts for more than 500,000 hospitalizations in the United States each year.¹ Although the management of VTE centers around anticoagulation, the concurrent use of IVC filters has increased over the past several decades.² Several observational studies have attempted to quantify IVC filter usage and have shown that overall filter placement has increased at an impressive rate. Within two decades, the number of patients undergoing IVC filter placement has increased nearly 25 times from 2,000 in 1979 to 49,000 in 1999.² Recent Medicare data show that claims for IVC filter placement procedures have increased from 30,756 in 1999 to 65,041 in 2008.³ IVC filter placement rates are higher in the US than in other developed countries; one review projected that in 2012, the IVC filter placement rate in a given population in the US is 25 times higher than that in a similar population in Europe.⁴

The guidelines for IVC filter usage are largely based on expert opinion, and solid data regarding this intervention are lacking. This combination is problematic, especially because the practice is becoming commonplace, and filter-related complications are increasingly recognized. Additionally, the appropriateness of filter use varies among providers, as evidenced by a retrospective study in which three VTE experts reviewed medical records to determine the appropriateness of filter placement. They unanimously agreed that filter use was appropriate in 51% of the cases, unanimously agreed that filter use was inappropriate in 26% of the cases, and lacked consensus on the appropriateness of filter use in 23% of the cases.⁵ The striking lack of consensus among experts underscores the wide range of opinion regarding the appropriateness of IVC filter placement on a case-by-case basis. Moreover, evidence suggests that physician adherence to guidelines for appropriate IVC filter use is suboptimal. One single-center study showed that only 43.5% of filters placed by interventional radiology practitioners met the guidelines established by the American College of Chest Physicians (ACCP), with a slightly increased percentage of filter placement meeting guidelines if the requesting provider is an IM-trained physician.⁶

In theory, the concept of IVC filters makes intuitive sense—filters block the ascent of any thrombus from the lower extremities to prevent the feared complication of a pulmonary embolism (PE). Unfortunately, rigorous data are limited, and consensus guidelines vary between different specialty organizations, further obfuscating the role of IVC filter placement in the management of VTE. For example, the ACCP recommends against the use of IVC filters in most patients with VTE receiving anticoagulation and does not list any prophylactic indications.^7,8 Meanwhile, the Society of Interventional Radiology lists prophylactic indications for IVC filter placement in certain patient populations, such patients with a risk of VTE and a high risk of bleeding, and notes numerous relative indications for IVC filter placement.⁸ Notably, these differences in expert opinion likely influence practice patterns, as evidenced by the increase in IVC filter placement for relative indications.^9,10

The Prevention du Risque d’Embolie Pulmonaire par Interruption Cave (PRECIP) trial is the most robust study supporting the 2016 ACCP recommendation against IVC filter use in patients that can receive anticoagulation.^7,11 This study randomized 400 patients with deep vein thrombosis (DVT) at high risk for PE to anticoagulation with or without permanent filter placement to address VTE and mortality rates associated with IVC filter placement. The trial showed that the VTE burden shifts in the presence of IVC filters. At 2-year follow-up, the group with IVC filters had nonsignificantly fewer PEs than the control group and an increased incidence of DVT. Mortality rates did not differ between groups.¹¹ At eight-year follow-up this shift in VTE burden is again seen given that the number of PEs in patients who received IVC filters decreased and the incidence of DVTs increased. Again, mortality did not differ between groups.¹² A subsequent study randomized 399 patients with DVT and acute symptomatic PE with at least one additional marker of severity to anticoagulation with or without retrievable IVC filter placement and showed no difference in recurrent PE or mortality at 3 or 6 months.¹³ These results argue against placing retrievable filters in patients receiving anticoagulation.

The identification of associated adverse events further favor the judicious use of IVC filters. A retrospective review of the long-term complications of IVC filters based on imaging data showed a 14% fracture rate, 13% IVC thrombosis rate, and a 48% perforation rate.¹⁴ Multiple studies have shown that the associated complication rates of retrievable filters are higher than those of permanent filters; such an association is concerning given that retrievable filter usage exceeds permanent filter usage.^14,15 The increase in retrievable filter usage is likely attributable to their attractive risk-benefit calculation. In theory, retrievable IVC filters should be perfect for patients who have conditions that increase VTE risk but create temporary contraindications, such as trauma or major surgery, to anticoagulation. However, anticoagulation is preferred over IVC filters in the long term because the complication rates of IVC filters increase with dwell time.¹⁶ Given the reports of adverse events and concern that IVC filters are not appropriately removed, the Food and Drug Administration recommends removing retrievable IVC filters once the risk of filters outweighs the benefits, which appears to be 29-54 days after implantation.¹⁷ However, successful retrieval rates are low, both because of the low rates of removal attempts and because of the interference of complications, such as embedded or thrombosed filters, with removal.^10,18 As an example, in a retrospective review of all patients who received an IVC filter at an academic medical center over the period of 2003-2011, nearly 25% of patients were discharged on anticoagulation after IVC filter placement.¹⁰ This suggests that their contraindication to anticoagulation and need for IVC placement have passed by the time of discharge. Nevertheless, clinicians attempted filter retrieval in only 9.6% of these patients, representing a significant missed opportunity of treatment with anticoagulation rather than IVC filters.¹⁰

Factors such as filter plan documentation, hematology involvement, patient age ≤70 years, and establishment of dedicated IVC filter clinics are correlated with improved rates of filter removal; these correlations emphasize the importance of a clear follow-up plan in the timely removal of these devices.^18,19

IVC filter placement is inappropriate in the vast majority of patients with VTE who can be anticoagulated. However the ACCP does acknowledge that a small subset of patients – specifically, those with severe or massive PE – may fall outside this guideline.⁷ Clinicians fear that these patients have low cardiopulmonary reserve and may experience hemodynamic collapse and death with another “hit” from a recurrent PE. This recommendation is consistent with the evidence that in unstable patients with PE, IVC filter placement is associated with decreased in-hospital mortality.²⁰ Data remain limited for this situation, and the decision to place an IVC filter in anticoagulated but unstable patients is an individualized one.

In stable patients admitted to the medical service with VTE and who can be anticoagulated, there is little evidence that placement of an IVC filter will improve short- or long-term mortality. Hospitalists should anticoagulate these patients with a vitamin-K antagonist, heparin product, or novel oral anticoagulants.

• Anticoagulate hemodynamically stable patients who are admitted to the medical service with VTE and who do not have a contraindication to anticoagulation. Do not place a permanent or retrievable IVC filter.
• IVC filter placement may benefit unstable patients who may experience hemodynamic collapse with an increased PE burden. IVC filter placement should be discussed with a multidisciplinary team.
• When discharging a patient with an IVC filter, hospitalists should improve retrieval rates by scheduling subsequent removal. The discharge summary should contain information about the IVC filter, as well as clear instructions regarding the plan for removal. The instructions should include radiology follow-up information and the designation of responsible physicians in case of questions.

Although IVC filter use is increasing, the evidence does not support their use in hemodynamically stable patients who can be anticoagulated. The patient described in the initial case has no contraindication to systemic anticoagulation. Therefore, she should be started on anticoagulation, and an IVC filter should not be placed.

Do you think this is a low-value practice? Is this truly a “Thing We Do for No Reason?” Share what you do in your practice and join in the conversation online by retweeting it on Twitter (#TWDFNR) and liking it on Facebook. We invite you to propose ideas for other “Things We Do for No Reason” topics by emailingTWDFNR@hospitalmedicine.org.

Disclosures

The authors do not have any conflicts of interest to disclose

The “Things We Do for No Reason” (TWDFNR) series reviews practices that have become common parts of hospital care but may provide little value to our patients. Practices reviewed in the TWDFNR series do not represent “black and white” conclusions or clinical practice standards but are meant as a starting place for research and active discussions among hospitalists and patients. We invite you to be part of that discussion. https://www.choosingwisely.org/

A nticoagulation is the cornerstone of acute venous thromboembolism (VTE) management. Nonetheless, the use of inferior vena cava (IVC) filters in addition to anticoagulation is increasing, with wide variation in practice patterns and a growing recognition of filter-related complications. Rigorous randomized controlled data demonstrating that IVC filters, particularly the increasingly commonly placed retrievable filters, provide a mortality benefit are sparse. Given our review of IVC filter use and the lack of evidence demonstrating that IVC filters provide a mortality benefit, we recommend using anticoagulation alone for stable medical service patients admitted with acute VTE. In nuanced cases, hospitalists should engage in multidisciplinary care to develop individualized treatment options.

A 65-year-old woman with a history of diabetes mellitus, metastatic breast cancer, and peptic ulcer disease presents to the Emergency Department for the evaluation of right thigh swelling, chest pain, and dyspnea after a transcontinental flight. Physical examination is notable for a pulse of 114 beats per minute, blood pressure of 136/93 mm Hg, respiratory rate of 14 breaths per minute, oxygen saturation of 95% on room air, and swelling of the right thigh. Computerized tomography imaging demonstrates multiple bilateral pulmonary emboli. Emergency department physicians begin anticoagulation and inform you that they have ordered the placement of a retrievable inferior vena cava (IVC) filter.

Acute venous thromboembolism (VTE) accounts for more than 500,000 hospitalizations in the United States each year.¹ Although the management of VTE centers around anticoagulation, the concurrent use of IVC filters has increased over the past several decades.² Several observational studies have attempted to quantify IVC filter usage and have shown that overall filter placement has increased at an impressive rate. Within two decades, the number of patients undergoing IVC filter placement has increased nearly 25 times from 2,000 in 1979 to 49,000 in 1999.² Recent Medicare data show that claims for IVC filter placement procedures have increased from 30,756 in 1999 to 65,041 in 2008.³ IVC filter placement rates are higher in the US than in other developed countries; one review projected that in 2012, the IVC filter placement rate in a given population in the US is 25 times higher than that in a similar population in Europe.⁴

The guidelines for IVC filter usage are largely based on expert opinion, and solid data regarding this intervention are lacking. This combination is problematic, especially because the practice is becoming commonplace, and filter-related complications are increasingly recognized. Additionally, the appropriateness of filter use varies among providers, as evidenced by a retrospective study in which three VTE experts reviewed medical records to determine the appropriateness of filter placement. They unanimously agreed that filter use was appropriate in 51% of the cases, unanimously agreed that filter use was inappropriate in 26% of the cases, and lacked consensus on the appropriateness of filter use in 23% of the cases.⁵ The striking lack of consensus among experts underscores the wide range of opinion regarding the appropriateness of IVC filter placement on a case-by-case basis. Moreover, evidence suggests that physician adherence to guidelines for appropriate IVC filter use is suboptimal. One single-center study showed that only 43.5% of filters placed by interventional radiology practitioners met the guidelines established by the American College of Chest Physicians (ACCP), with a slightly increased percentage of filter placement meeting guidelines if the requesting provider is an IM-trained physician.⁶

In theory, the concept of IVC filters makes intuitive sense—filters block the ascent of any thrombus from the lower extremities to prevent the feared complication of a pulmonary embolism (PE). Unfortunately, rigorous data are limited, and consensus guidelines vary between different specialty organizations, further obfuscating the role of IVC filter placement in the management of VTE. For example, the ACCP recommends against the use of IVC filters in most patients with VTE receiving anticoagulation and does not list any prophylactic indications.^7,8 Meanwhile, the Society of Interventional Radiology lists prophylactic indications for IVC filter placement in certain patient populations, such patients with a risk of VTE and a high risk of bleeding, and notes numerous relative indications for IVC filter placement.⁸ Notably, these differences in expert opinion likely influence practice patterns, as evidenced by the increase in IVC filter placement for relative indications.^9,10

The Prevention du Risque d’Embolie Pulmonaire par Interruption Cave (PRECIP) trial is the most robust study supporting the 2016 ACCP recommendation against IVC filter use in patients that can receive anticoagulation.^7,11 This study randomized 400 patients with deep vein thrombosis (DVT) at high risk for PE to anticoagulation with or without permanent filter placement to address VTE and mortality rates associated with IVC filter placement. The trial showed that the VTE burden shifts in the presence of IVC filters. At 2-year follow-up, the group with IVC filters had nonsignificantly fewer PEs than the control group and an increased incidence of DVT. Mortality rates did not differ between groups.¹¹ At eight-year follow-up this shift in VTE burden is again seen given that the number of PEs in patients who received IVC filters decreased and the incidence of DVTs increased. Again, mortality did not differ between groups.¹² A subsequent study randomized 399 patients with DVT and acute symptomatic PE with at least one additional marker of severity to anticoagulation with or without retrievable IVC filter placement and showed no difference in recurrent PE or mortality at 3 or 6 months.¹³ These results argue against placing retrievable filters in patients receiving anticoagulation.

The identification of associated adverse events further favor the judicious use of IVC filters. A retrospective review of the long-term complications of IVC filters based on imaging data showed a 14% fracture rate, 13% IVC thrombosis rate, and a 48% perforation rate.¹⁴ Multiple studies have shown that the associated complication rates of retrievable filters are higher than those of permanent filters; such an association is concerning given that retrievable filter usage exceeds permanent filter usage.^14,15 The increase in retrievable filter usage is likely attributable to their attractive risk-benefit calculation. In theory, retrievable IVC filters should be perfect for patients who have conditions that increase VTE risk but create temporary contraindications, such as trauma or major surgery, to anticoagulation. However, anticoagulation is preferred over IVC filters in the long term because the complication rates of IVC filters increase with dwell time.¹⁶ Given the reports of adverse events and concern that IVC filters are not appropriately removed, the Food and Drug Administration recommends removing retrievable IVC filters once the risk of filters outweighs the benefits, which appears to be 29-54 days after implantation.¹⁷ However, successful retrieval rates are low, both because of the low rates of removal attempts and because of the interference of complications, such as embedded or thrombosed filters, with removal.^10,18 As an example, in a retrospective review of all patients who received an IVC filter at an academic medical center over the period of 2003-2011, nearly 25% of patients were discharged on anticoagulation after IVC filter placement.¹⁰ This suggests that their contraindication to anticoagulation and need for IVC placement have passed by the time of discharge. Nevertheless, clinicians attempted filter retrieval in only 9.6% of these patients, representing a significant missed opportunity of treatment with anticoagulation rather than IVC filters.¹⁰

Factors such as filter plan documentation, hematology involvement, patient age ≤70 years, and establishment of dedicated IVC filter clinics are correlated with improved rates of filter removal; these correlations emphasize the importance of a clear follow-up plan in the timely removal of these devices.^18,19

IVC filter placement is inappropriate in the vast majority of patients with VTE who can be anticoagulated. However the ACCP does acknowledge that a small subset of patients – specifically, those with severe or massive PE – may fall outside this guideline.⁷ Clinicians fear that these patients have low cardiopulmonary reserve and may experience hemodynamic collapse and death with another “hit” from a recurrent PE. This recommendation is consistent with the evidence that in unstable patients with PE, IVC filter placement is associated with decreased in-hospital mortality.²⁰ Data remain limited for this situation, and the decision to place an IVC filter in anticoagulated but unstable patients is an individualized one.

In stable patients admitted to the medical service with VTE and who can be anticoagulated, there is little evidence that placement of an IVC filter will improve short- or long-term mortality. Hospitalists should anticoagulate these patients with a vitamin-K antagonist, heparin product, or novel oral anticoagulants.

• Anticoagulate hemodynamically stable patients who are admitted to the medical service with VTE and who do not have a contraindication to anticoagulation. Do not place a permanent or retrievable IVC filter.
• IVC filter placement may benefit unstable patients who may experience hemodynamic collapse with an increased PE burden. IVC filter placement should be discussed with a multidisciplinary team.
• When discharging a patient with an IVC filter, hospitalists should improve retrieval rates by scheduling subsequent removal. The discharge summary should contain information about the IVC filter, as well as clear instructions regarding the plan for removal. The instructions should include radiology follow-up information and the designation of responsible physicians in case of questions.

Although IVC filter use is increasing, the evidence does not support their use in hemodynamically stable patients who can be anticoagulated. The patient described in the initial case has no contraindication to systemic anticoagulation. Therefore, she should be started on anticoagulation, and an IVC filter should not be placed.

Do you think this is a low-value practice? Is this truly a “Thing We Do for No Reason?” Share what you do in your practice and join in the conversation online by retweeting it on Twitter (#TWDFNR) and liking it on Facebook. We invite you to propose ideas for other “Things We Do for No Reason” topics by emailingTWDFNR@hospitalmedicine.org.

Disclosures

The authors do not have any conflicts of interest to disclose

The “Things We Do for No Reason” (TWDFNR) series reviews practices that have become common parts of hospital care but may provide little value to our patients. Practices reviewed in the TWDFNR series do not represent “black and white” conclusions or clinical practice standards but are meant as a starting place for research and active discussions among hospitalists and patients. We invite you to be part of that discussion. https://www.choosingwisely.org/

A nticoagulation is the cornerstone of acute venous thromboembolism (VTE) management. Nonetheless, the use of inferior vena cava (IVC) filters in addition to anticoagulation is increasing, with wide variation in practice patterns and a growing recognition of filter-related complications. Rigorous randomized controlled data demonstrating that IVC filters, particularly the increasingly commonly placed retrievable filters, provide a mortality benefit are sparse. Given our review of IVC filter use and the lack of evidence demonstrating that IVC filters provide a mortality benefit, we recommend using anticoagulation alone for stable medical service patients admitted with acute VTE. In nuanced cases, hospitalists should engage in multidisciplinary care to develop individualized treatment options.

A 65-year-old woman with a history of diabetes mellitus, metastatic breast cancer, and peptic ulcer disease presents to the Emergency Department for the evaluation of right thigh swelling, chest pain, and dyspnea after a transcontinental flight. Physical examination is notable for a pulse of 114 beats per minute, blood pressure of 136/93 mm Hg, respiratory rate of 14 breaths per minute, oxygen saturation of 95% on room air, and swelling of the right thigh. Computerized tomography imaging demonstrates multiple bilateral pulmonary emboli. Emergency department physicians begin anticoagulation and inform you that they have ordered the placement of a retrievable inferior vena cava (IVC) filter.

Acute venous thromboembolism (VTE) accounts for more than 500,000 hospitalizations in the United States each year.¹ Although the management of VTE centers around anticoagulation, the concurrent use of IVC filters has increased over the past several decades.² Several observational studies have attempted to quantify IVC filter usage and have shown that overall filter placement has increased at an impressive rate. Within two decades, the number of patients undergoing IVC filter placement has increased nearly 25 times from 2,000 in 1979 to 49,000 in 1999.² Recent Medicare data show that claims for IVC filter placement procedures have increased from 30,756 in 1999 to 65,041 in 2008.³ IVC filter placement rates are higher in the US than in other developed countries; one review projected that in 2012, the IVC filter placement rate in a given population in the US is 25 times higher than that in a similar population in Europe.⁴

The guidelines for IVC filter usage are largely based on expert opinion, and solid data regarding this intervention are lacking. This combination is problematic, especially because the practice is becoming commonplace, and filter-related complications are increasingly recognized. Additionally, the appropriateness of filter use varies among providers, as evidenced by a retrospective study in which three VTE experts reviewed medical records to determine the appropriateness of filter placement. They unanimously agreed that filter use was appropriate in 51% of the cases, unanimously agreed that filter use was inappropriate in 26% of the cases, and lacked consensus on the appropriateness of filter use in 23% of the cases.⁵ The striking lack of consensus among experts underscores the wide range of opinion regarding the appropriateness of IVC filter placement on a case-by-case basis. Moreover, evidence suggests that physician adherence to guidelines for appropriate IVC filter use is suboptimal. One single-center study showed that only 43.5% of filters placed by interventional radiology practitioners met the guidelines established by the American College of Chest Physicians (ACCP), with a slightly increased percentage of filter placement meeting guidelines if the requesting provider is an IM-trained physician.⁶

In theory, the concept of IVC filters makes intuitive sense—filters block the ascent of any thrombus from the lower extremities to prevent the feared complication of a pulmonary embolism (PE). Unfortunately, rigorous data are limited, and consensus guidelines vary between different specialty organizations, further obfuscating the role of IVC filter placement in the management of VTE. For example, the ACCP recommends against the use of IVC filters in most patients with VTE receiving anticoagulation and does not list any prophylactic indications.^7,8 Meanwhile, the Society of Interventional Radiology lists prophylactic indications for IVC filter placement in certain patient populations, such patients with a risk of VTE and a high risk of bleeding, and notes numerous relative indications for IVC filter placement.⁸ Notably, these differences in expert opinion likely influence practice patterns, as evidenced by the increase in IVC filter placement for relative indications.^9,10

The Prevention du Risque d’Embolie Pulmonaire par Interruption Cave (PRECIP) trial is the most robust study supporting the 2016 ACCP recommendation against IVC filter use in patients that can receive anticoagulation.^7,11 This study randomized 400 patients with deep vein thrombosis (DVT) at high risk for PE to anticoagulation with or without permanent filter placement to address VTE and mortality rates associated with IVC filter placement. The trial showed that the VTE burden shifts in the presence of IVC filters. At 2-year follow-up, the group with IVC filters had nonsignificantly fewer PEs than the control group and an increased incidence of DVT. Mortality rates did not differ between groups.¹¹ At eight-year follow-up this shift in VTE burden is again seen given that the number of PEs in patients who received IVC filters decreased and the incidence of DVTs increased. Again, mortality did not differ between groups.¹² A subsequent study randomized 399 patients with DVT and acute symptomatic PE with at least one additional marker of severity to anticoagulation with or without retrievable IVC filter placement and showed no difference in recurrent PE or mortality at 3 or 6 months.¹³ These results argue against placing retrievable filters in patients receiving anticoagulation.

The identification of associated adverse events further favor the judicious use of IVC filters. A retrospective review of the long-term complications of IVC filters based on imaging data showed a 14% fracture rate, 13% IVC thrombosis rate, and a 48% perforation rate.¹⁴ Multiple studies have shown that the associated complication rates of retrievable filters are higher than those of permanent filters; such an association is concerning given that retrievable filter usage exceeds permanent filter usage.^14,15 The increase in retrievable filter usage is likely attributable to their attractive risk-benefit calculation. In theory, retrievable IVC filters should be perfect for patients who have conditions that increase VTE risk but create temporary contraindications, such as trauma or major surgery, to anticoagulation. However, anticoagulation is preferred over IVC filters in the long term because the complication rates of IVC filters increase with dwell time.¹⁶ Given the reports of adverse events and concern that IVC filters are not appropriately removed, the Food and Drug Administration recommends removing retrievable IVC filters once the risk of filters outweighs the benefits, which appears to be 29-54 days after implantation.¹⁷ However, successful retrieval rates are low, both because of the low rates of removal attempts and because of the interference of complications, such as embedded or thrombosed filters, with removal.^10,18 As an example, in a retrospective review of all patients who received an IVC filter at an academic medical center over the period of 2003-2011, nearly 25% of patients were discharged on anticoagulation after IVC filter placement.¹⁰ This suggests that their contraindication to anticoagulation and need for IVC placement have passed by the time of discharge. Nevertheless, clinicians attempted filter retrieval in only 9.6% of these patients, representing a significant missed opportunity of treatment with anticoagulation rather than IVC filters.¹⁰

Factors such as filter plan documentation, hematology involvement, patient age ≤70 years, and establishment of dedicated IVC filter clinics are correlated with improved rates of filter removal; these correlations emphasize the importance of a clear follow-up plan in the timely removal of these devices.^18,19

IVC filter placement is inappropriate in the vast majority of patients with VTE who can be anticoagulated. However the ACCP does acknowledge that a small subset of patients – specifically, those with severe or massive PE – may fall outside this guideline.⁷ Clinicians fear that these patients have low cardiopulmonary reserve and may experience hemodynamic collapse and death with another “hit” from a recurrent PE. This recommendation is consistent with the evidence that in unstable patients with PE, IVC filter placement is associated with decreased in-hospital mortality.²⁰ Data remain limited for this situation, and the decision to place an IVC filter in anticoagulated but unstable patients is an individualized one.

In stable patients admitted to the medical service with VTE and who can be anticoagulated, there is little evidence that placement of an IVC filter will improve short- or long-term mortality. Hospitalists should anticoagulate these patients with a vitamin-K antagonist, heparin product, or novel oral anticoagulants.

• Anticoagulate hemodynamically stable patients who are admitted to the medical service with VTE and who do not have a contraindication to anticoagulation. Do not place a permanent or retrievable IVC filter.
• IVC filter placement may benefit unstable patients who may experience hemodynamic collapse with an increased PE burden. IVC filter placement should be discussed with a multidisciplinary team.
• When discharging a patient with an IVC filter, hospitalists should improve retrieval rates by scheduling subsequent removal. The discharge summary should contain information about the IVC filter, as well as clear instructions regarding the plan for removal. The instructions should include radiology follow-up information and the designation of responsible physicians in case of questions.

Although IVC filter use is increasing, the evidence does not support their use in hemodynamically stable patients who can be anticoagulated. The patient described in the initial case has no contraindication to systemic anticoagulation. Therefore, she should be started on anticoagulation, and an IVC filter should not be placed.

Do you think this is a low-value practice? Is this truly a “Thing We Do for No Reason?” Share what you do in your practice and join in the conversation online by retweeting it on Twitter (#TWDFNR) and liking it on Facebook. We invite you to propose ideas for other “Things We Do for No Reason” topics by emailingTWDFNR@hospitalmedicine.org.

Disclosures

The authors do not have any conflicts of interest to disclose

Semi-Recumbent Position to Prevent Ventilator-Associated Pneumonia: Is It Possible?

By Joseph Ming Wah Li, MD

Van Nieuwenhoven CA, Vandenbroucke-Grauls C, van Tiel FH, et al. Feasibility and effects of the semirecumbent position to prevent ventilator-associated pneumonia: a randomized study. Crit Care Med. 2006 Feb;34(2):396-402.

Ventilator-associated pneumonia (VAP) is a cause of significant morbidity and mortality among mechanically ventilated patients. Studies with radioactive-labeled enteral feeds have demonstrated an increased frequency of endotracheal aspiration of gastric contents in supine patients. The CDC guidelines for prevention of nosocomial pneumonia advise placement of mechanically ventilated patients in a semi-recumbent position as a VAP prevention measure.

Only one previous study, by Drakulovic and colleagues, has assessed this strategy to prevent VAP.1 That study demonstrated a 75% decrease in the incidence of VAP. But van Nieuwenhoven and colleagues raised two important questions about the findings from the previous study: First, the Drakulovic study placed control patients in a horizontal (zero degrees) position, which is not the standard of care in most ICUs. Most patients are placed at 10 degrees, and this position is elevated as patients are weaned. Second, the Drakulovic study measured patients only once daily but did not monitor their body positions in between the daily measurements.

Dr. van Nieuwenhoven and colleagues set out to determine whether it is feasible to keep mechanically ventilated patients in a semi-recumbent position on a continual basis and whether this measure would prevent VAP. This was a prospective multi-centered trial in which mechanically ventilated patients were randomly assigned to the semi-recumbent position with a target backrest elevation of 45 degrees or standard of care (supine position) with a backrest elevation of 10 degrees. They used a transducer with a pendulum, which was placed on the bed frame to measure the backrest elevation every 60 seconds for up to seven days. They calculated a mean degree of elevation for each patient daily. Nurses always respected the patient’s request for positioning, but a dedicated research nurse restored backrest position to the randomized position whenever possible.

Baseline characteristics for both groups were similar. For the supine (control) group, average elevations were 9.8 degrees on day one and 16.1 degrees on day seven. For the semi-recumbent group, average elevations were 28.1 degrees on day one and 22.6 degrees on day seven. There were no significant differences in numbers of patients who developed VAP in either group.

This study suggests that, despite the use of dedicated research nurses to maintain positioning, it may not be possible to keep patients’ backrests elevated to 45 degrees. Keeping patients’ backrests at an elevation of nearly 30 degrees does not appear to prevent VAP more than keeping patients’ backrests at 10 degrees, the present standard of care.

Reference

1. Drakulovic MB, Torres A, Bauer TT, et al. Supine body position as a risk factor for nosocomial pneumonia in mechanically ventilated patients: a randomised trial. Lancet. 1999;354(9193):1851-1858.

Bar Codes in Medicine: An Opportunity for Quality Improvement

By Alex Carbo, MD

Poon EG, Cina JL, Churchill W, et al. Medication dispensing errors and potential adverse drug events before and after implementing bar code technology in the pharmacy. Ann Intern Med. 2006;145:426-434.

Medication errors and adverse drug events (ADEs) have received much attention in the literature; the use of health information technology to mitigate these errors and ADEs has now been proposed in many areas of healthcare. In an effort to decrease medication-dispensing errors, the U.S. Food and Drug Administration (FDA) mandated bar code use for all medications in hospitals, beginning in April 2006. While this technology has been extensively studied in other industries, there is little data describing its effects in the healthcare system.

Poon and colleagues set out to evaluate whether implementation of bar code technology reduced dispensing errors and the ADEs that might be caused by these miscalculations. In a before-and-after evaluation, they studied more than 350,000 dispensed medication doses in an academic medical center between February 2003 and September 2004.

During the bar code conversion process, the hospital pharmacy built a dedicated repackaging center, which was responsible for affixing a bar code to every dose of medication. These medications were then dispensed in three different configurations: two configurations required staff to verify all doses at least once using bar code scanning, and the third configuration—for commonly dispensed medications that could not be accommodated in a standard carousel machine because of their size or need for refrigeration—required scanning only one dose from each batch.

The authors found a 93% to 96% relative reduction in the incidence of target dispensing errors (P<0.001) and an 86% to 97% relative reduction in the incidence of potential ADEs (P<0.001) in the two configurations that required staff to verify all doses by scanning. The greatest reductions were seen in wrong medication errors (56%), wrong strength/dose errors (71%), wrong formulation errors (90%), and expired medication errors (100%).

In the configuration that did not require scanning of every dose, however, there was a 60% relative reduction in the incidence of target dispensing errors (P<0.001), but a 2.4-fold increase in the incidence of target potential ADEs. This included new errors attributable to wrong strength and wrong medication dispensing.

In light of the FDA’s mandate regarding bar codes, it seems that every hospital has the opportunity to improve patient safety and decrease medication error rates with the use of bar code technology. This study suggests that in order to achieve this benefit these systems should be designed to ensure that every medication dose is verified by scanning during the dispensing process.

Evaluation of a Guideline to Guide Resuscitation

By Cindy Lien, MD

Morrison LJ, Visentin LM, Kiss A, et al. Validation of a rule for termination of resuscitation in out-of-hospital cardiac arrest. N Engl J Med. 2006 Aug 3;355(5):478-487.

The survival rate of patients with out-of-hospital cardiac arrest is very low. Thus, guidelines have been developed for termination of resuscitation for those patients who have had no response to advanced cardiac life support provided by emergency medical service (EMS) personnel. Similar guidelines have not yet been developed, however, for situations in which patients receive basic life support from emergency workers trained in the use of an automated external cardiac defibrillator. Patients with little potential for survival are routinely transported to emergency departments, at significant cost to the healthcare system.

Morrison and colleagues present results from the Termination of Resuscitation (TOR) study, a prospective evaluation of a clinical prediction rule for the termination of basic life support by emergency medical personnel trained in the use of automated external defibrillators. The clinical prediction rule, previously developed in a retrospective review of case records from a large urban EMS system, recommends termination of resuscitation if there is no return of spontaneous circulation, no shock administered, and no witness of the arrest by EMS personnel.

In the current study, the authors obtained follow-up data for 1,240 adult patients in Ontario, Canada, who had suffered an arrest of presumed cardiac cause and were subsequently transported to the emergency department after resuscitative efforts. Twenty-four EMS systems participated in the study. The study found that only 0.5% of the patients for whom the clinical prediction rule recommended termination survived (four out of 776 patients). Of the 1,240 total study patients, 41 (3%) survived. The clinical prediction rule recommended continuation of resuscitative efforts for 37 of these 41 patients, resulting in a specificity of 90.2%. The positive predictive value for death was calculated to be 99.5% when termination was recommended.

The TOR trial also determined whether the addition of other criteria to the original prediction rule could further refine the specificity and positive predictive value. They found that the addition to the criteria of a response time greater than eight minutes increased the positive predictive value and specificity to 99.7% and 97.6%, respectively. When the variable “not witnessed by bystander” was added to the clinical prediction rule, both the positive predictive value and specificity increased to 100%. In other words, no patients survived if they had had a completely unwitnessed arrest, no return of spontaneous circulation, and no shocks delivered.

This study identifies a subpopulation of patients with presumed cardiac arrest for whom termination of resuscitative efforts in the field appears reasonable. The authors note that a survival rate of 1% or less has been suggested in past literature as reflective of medical futility. The TOR investigators acknowledge that their study took place before the 2005 AHA Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care were released and that their study protocols were consistent with the 2000 resuscitation guidelines. In light of this information, continued validity testing of the clinical prediction rule under the 2005 AHA protocols is needed.

Nonetheless, it is quite impressive that use of their clinical prediction rule would have resulted in transportation of only 37% of patients (464 of 1,240), rather than 100% of patients, as is currently the practice. If the guidelines described in this article are to be implemented, further studies are necessary to address the training of EMS personnel, who would carry responsibility for terminating resuscitation and notifying families of patients’ deaths.

Prevention of Ventilator-Associated Pneumonia

By Diane Sliwka, MD

Koeman M, van der Ven AJ, Hak E, et al. Oral decontamination with chlorhexidine reduces the incidence of ventilator-associated pneumonia. Am J Respir Crit Care Med. 2006 Jun;173(12):1348-1355. Epub 2006 Apr 7.

Ventilator-associated pneumonia (VAP) is an important nosocomial source of morbidity and mortality. The use of prophylactic antimicrobials to decrease VAP raises concern for antimicrobial resistance. This study evaluates the topical antiseptic chlorhexidine (CHX) as an alternative prophylactic intervention for VAP. CHX has previously been shown to decrease VAP in cardiac surgical patients, but has not been studied in higher risk, long-term-ventilated patients. Because CHX works better for gram-positive organisms, the combination of colistin and CHX (COL + CHX) was also studied for improved gram-negative coverage.

This multi-center, randomized, double-blind, placebo-controlled trial enrolled 385 adult patients. Patients who were expected to be intubated for longer than 48 hours were randomized to 3 arms: CHX alone, CHX + COL, and placebo. Exclusion criteria included known preadmission immunocompromised state, pregnancy, and physical limitation to oral application. Pneumonia was defined by clinical decision-making, which was later confirmed by three blinded intensivists’ reviews of the case records and supported by daily clinical pulmonary infection scores.

The primary endpoint of VAP was diagnosed in 52 of 385 patients: 18% placebo, 13% CHX, and 10% CHX + COL. Rate of VAP in the two treatment groups was lower than placebo and reached statistical significance when compared to placebo. The daily hazard ratio for CHX versus placebo was .352 (95% CI .160, .791); for CHX + COL versus placebo, it was .454 (95% CI .224, .925), showing a 65% and 55% reduction in the rate of pneumonia development. Multivariate analysis of variables such as gender, pulmonary admission diagnosis, colonization at time of admission, and antimicrobial use on admission did not affect the data.

The secondary endpoint of endotracheal colonization was evaluated by a twice-weekly endotracheal culture. There was no statistically significant difference in colonization among the three groups in the first (days 1-4) or third (days 9-12) time frames. During the second time frame (days 5-8), there was a statistically significant decrease in colonization for the CHX + COL treatment group when compared to both placebo (16% versus 40% p<.007) and to CHX (16% versus 38%, p<.011); this decrease is thought to be due to gram-negative coverage by COL.

The secondary endpoint of oropharyngeal colonization was evaluated for 87% of all patient days. CHX and CHX + COL were similarly effective for gram-positive bacteria when compared to placebo, with 30% and 27% reduction in rates of colonization, respectively: HR 0.695 for CHX (95% CI, 0.606, 0.796; p < 0.001) and 0.732 (95% CI, 0.640, 0.838; p < 0.001) for CHX + COL. The CHX + COL combination was more effective for gram-negative bacteria: daily HR .534 (95% CI, 0.455, 0.626; p <0.001) alone with a 47% reduction in gram-negative colonization compared to CHX.

No difference was seen in ICU mortality, duration of mechanical ventilation, or duration of ICU stay. One adverse event (tongue swelling) occurred in the CHX + COL group.

Limitations of the study include the following:

• Daily assessments on all patients were not performed;
• The placebo group had more males and more infections on admission than the other two groups, raising the question of randomization error;
• Clinical versus quantitative diagnosis of pneumonia may overestimate VAP in this study;
• It is not known how many patients were not enrolled in the study due to short anticipated ventilator times, but who later had prolonged ventilations; and
• The lack of effect on ventilator time, ICU length of stay, and mortality raises the question of the significance of these findings.

Despite these limitations, the low cost of these treatments, minimal adverse events, low risk of promoting significant antimicrobial resistance, and the finding of decreased VAP and bacterial colonization risk shown in this study support the potential benefit of topical decontamination with CHX and COL in conjunction with other measures of VAP prevention. TH

Reference

1. De Riso AJ II, Ladowski JS, Dillon TA, et al. Chlorhexidine gluconate 0.12% oral rinse reduces the incidence of total nosocomial respiratory infection and nonprophylactic systemic antibiotic use in patients undergoing heart surgery. Chest. 1996;109:1556-1561.

Semi-Recumbent Position to Prevent Ventilator-Associated Pneumonia: Is It Possible?

By Joseph Ming Wah Li, MD

Van Nieuwenhoven CA, Vandenbroucke-Grauls C, van Tiel FH, et al. Feasibility and effects of the semirecumbent position to prevent ventilator-associated pneumonia: a randomized study. Crit Care Med. 2006 Feb;34(2):396-402.

Ventilator-associated pneumonia (VAP) is a cause of significant morbidity and mortality among mechanically ventilated patients. Studies with radioactive-labeled enteral feeds have demonstrated an increased frequency of endotracheal aspiration of gastric contents in supine patients. The CDC guidelines for prevention of nosocomial pneumonia advise placement of mechanically ventilated patients in a semi-recumbent position as a VAP prevention measure.

Only one previous study, by Drakulovic and colleagues, has assessed this strategy to prevent VAP.1 That study demonstrated a 75% decrease in the incidence of VAP. But van Nieuwenhoven and colleagues raised two important questions about the findings from the previous study: First, the Drakulovic study placed control patients in a horizontal (zero degrees) position, which is not the standard of care in most ICUs. Most patients are placed at 10 degrees, and this position is elevated as patients are weaned. Second, the Drakulovic study measured patients only once daily but did not monitor their body positions in between the daily measurements.

Dr. van Nieuwenhoven and colleagues set out to determine whether it is feasible to keep mechanically ventilated patients in a semi-recumbent position on a continual basis and whether this measure would prevent VAP. This was a prospective multi-centered trial in which mechanically ventilated patients were randomly assigned to the semi-recumbent position with a target backrest elevation of 45 degrees or standard of care (supine position) with a backrest elevation of 10 degrees. They used a transducer with a pendulum, which was placed on the bed frame to measure the backrest elevation every 60 seconds for up to seven days. They calculated a mean degree of elevation for each patient daily. Nurses always respected the patient’s request for positioning, but a dedicated research nurse restored backrest position to the randomized position whenever possible.

Baseline characteristics for both groups were similar. For the supine (control) group, average elevations were 9.8 degrees on day one and 16.1 degrees on day seven. For the semi-recumbent group, average elevations were 28.1 degrees on day one and 22.6 degrees on day seven. There were no significant differences in numbers of patients who developed VAP in either group.

This study suggests that, despite the use of dedicated research nurses to maintain positioning, it may not be possible to keep patients’ backrests elevated to 45 degrees. Keeping patients’ backrests at an elevation of nearly 30 degrees does not appear to prevent VAP more than keeping patients’ backrests at 10 degrees, the present standard of care.

Reference

1. Drakulovic MB, Torres A, Bauer TT, et al. Supine body position as a risk factor for nosocomial pneumonia in mechanically ventilated patients: a randomised trial. Lancet. 1999;354(9193):1851-1858.

Bar Codes in Medicine: An Opportunity for Quality Improvement

By Alex Carbo, MD

Poon EG, Cina JL, Churchill W, et al. Medication dispensing errors and potential adverse drug events before and after implementing bar code technology in the pharmacy. Ann Intern Med. 2006;145:426-434.

Medication errors and adverse drug events (ADEs) have received much attention in the literature; the use of health information technology to mitigate these errors and ADEs has now been proposed in many areas of healthcare. In an effort to decrease medication-dispensing errors, the U.S. Food and Drug Administration (FDA) mandated bar code use for all medications in hospitals, beginning in April 2006. While this technology has been extensively studied in other industries, there is little data describing its effects in the healthcare system.

Poon and colleagues set out to evaluate whether implementation of bar code technology reduced dispensing errors and the ADEs that might be caused by these miscalculations. In a before-and-after evaluation, they studied more than 350,000 dispensed medication doses in an academic medical center between February 2003 and September 2004.

During the bar code conversion process, the hospital pharmacy built a dedicated repackaging center, which was responsible for affixing a bar code to every dose of medication. These medications were then dispensed in three different configurations: two configurations required staff to verify all doses at least once using bar code scanning, and the third configuration—for commonly dispensed medications that could not be accommodated in a standard carousel machine because of their size or need for refrigeration—required scanning only one dose from each batch.

The authors found a 93% to 96% relative reduction in the incidence of target dispensing errors (P<0.001) and an 86% to 97% relative reduction in the incidence of potential ADEs (P<0.001) in the two configurations that required staff to verify all doses by scanning. The greatest reductions were seen in wrong medication errors (56%), wrong strength/dose errors (71%), wrong formulation errors (90%), and expired medication errors (100%).

In the configuration that did not require scanning of every dose, however, there was a 60% relative reduction in the incidence of target dispensing errors (P<0.001), but a 2.4-fold increase in the incidence of target potential ADEs. This included new errors attributable to wrong strength and wrong medication dispensing.

In light of the FDA’s mandate regarding bar codes, it seems that every hospital has the opportunity to improve patient safety and decrease medication error rates with the use of bar code technology. This study suggests that in order to achieve this benefit these systems should be designed to ensure that every medication dose is verified by scanning during the dispensing process.

Evaluation of a Guideline to Guide Resuscitation

By Cindy Lien, MD

Morrison LJ, Visentin LM, Kiss A, et al. Validation of a rule for termination of resuscitation in out-of-hospital cardiac arrest. N Engl J Med. 2006 Aug 3;355(5):478-487.

The survival rate of patients with out-of-hospital cardiac arrest is very low. Thus, guidelines have been developed for termination of resuscitation for those patients who have had no response to advanced cardiac life support provided by emergency medical service (EMS) personnel. Similar guidelines have not yet been developed, however, for situations in which patients receive basic life support from emergency workers trained in the use of an automated external cardiac defibrillator. Patients with little potential for survival are routinely transported to emergency departments, at significant cost to the healthcare system.

Morrison and colleagues present results from the Termination of Resuscitation (TOR) study, a prospective evaluation of a clinical prediction rule for the termination of basic life support by emergency medical personnel trained in the use of automated external defibrillators. The clinical prediction rule, previously developed in a retrospective review of case records from a large urban EMS system, recommends termination of resuscitation if there is no return of spontaneous circulation, no shock administered, and no witness of the arrest by EMS personnel.

In the current study, the authors obtained follow-up data for 1,240 adult patients in Ontario, Canada, who had suffered an arrest of presumed cardiac cause and were subsequently transported to the emergency department after resuscitative efforts. Twenty-four EMS systems participated in the study. The study found that only 0.5% of the patients for whom the clinical prediction rule recommended termination survived (four out of 776 patients). Of the 1,240 total study patients, 41 (3%) survived. The clinical prediction rule recommended continuation of resuscitative efforts for 37 of these 41 patients, resulting in a specificity of 90.2%. The positive predictive value for death was calculated to be 99.5% when termination was recommended.

The TOR trial also determined whether the addition of other criteria to the original prediction rule could further refine the specificity and positive predictive value. They found that the addition to the criteria of a response time greater than eight minutes increased the positive predictive value and specificity to 99.7% and 97.6%, respectively. When the variable “not witnessed by bystander” was added to the clinical prediction rule, both the positive predictive value and specificity increased to 100%. In other words, no patients survived if they had had a completely unwitnessed arrest, no return of spontaneous circulation, and no shocks delivered.

This study identifies a subpopulation of patients with presumed cardiac arrest for whom termination of resuscitative efforts in the field appears reasonable. The authors note that a survival rate of 1% or less has been suggested in past literature as reflective of medical futility. The TOR investigators acknowledge that their study took place before the 2005 AHA Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care were released and that their study protocols were consistent with the 2000 resuscitation guidelines. In light of this information, continued validity testing of the clinical prediction rule under the 2005 AHA protocols is needed.

Nonetheless, it is quite impressive that use of their clinical prediction rule would have resulted in transportation of only 37% of patients (464 of 1,240), rather than 100% of patients, as is currently the practice. If the guidelines described in this article are to be implemented, further studies are necessary to address the training of EMS personnel, who would carry responsibility for terminating resuscitation and notifying families of patients’ deaths.

Prevention of Ventilator-Associated Pneumonia

By Diane Sliwka, MD

Koeman M, van der Ven AJ, Hak E, et al. Oral decontamination with chlorhexidine reduces the incidence of ventilator-associated pneumonia. Am J Respir Crit Care Med. 2006 Jun;173(12):1348-1355. Epub 2006 Apr 7.

Ventilator-associated pneumonia (VAP) is an important nosocomial source of morbidity and mortality. The use of prophylactic antimicrobials to decrease VAP raises concern for antimicrobial resistance. This study evaluates the topical antiseptic chlorhexidine (CHX) as an alternative prophylactic intervention for VAP. CHX has previously been shown to decrease VAP in cardiac surgical patients, but has not been studied in higher risk, long-term-ventilated patients. Because CHX works better for gram-positive organisms, the combination of colistin and CHX (COL + CHX) was also studied for improved gram-negative coverage.

This multi-center, randomized, double-blind, placebo-controlled trial enrolled 385 adult patients. Patients who were expected to be intubated for longer than 48 hours were randomized to 3 arms: CHX alone, CHX + COL, and placebo. Exclusion criteria included known preadmission immunocompromised state, pregnancy, and physical limitation to oral application. Pneumonia was defined by clinical decision-making, which was later confirmed by three blinded intensivists’ reviews of the case records and supported by daily clinical pulmonary infection scores.

The primary endpoint of VAP was diagnosed in 52 of 385 patients: 18% placebo, 13% CHX, and 10% CHX + COL. Rate of VAP in the two treatment groups was lower than placebo and reached statistical significance when compared to placebo. The daily hazard ratio for CHX versus placebo was .352 (95% CI .160, .791); for CHX + COL versus placebo, it was .454 (95% CI .224, .925), showing a 65% and 55% reduction in the rate of pneumonia development. Multivariate analysis of variables such as gender, pulmonary admission diagnosis, colonization at time of admission, and antimicrobial use on admission did not affect the data.

The secondary endpoint of endotracheal colonization was evaluated by a twice-weekly endotracheal culture. There was no statistically significant difference in colonization among the three groups in the first (days 1-4) or third (days 9-12) time frames. During the second time frame (days 5-8), there was a statistically significant decrease in colonization for the CHX + COL treatment group when compared to both placebo (16% versus 40% p<.007) and to CHX (16% versus 38%, p<.011); this decrease is thought to be due to gram-negative coverage by COL.

The secondary endpoint of oropharyngeal colonization was evaluated for 87% of all patient days. CHX and CHX + COL were similarly effective for gram-positive bacteria when compared to placebo, with 30% and 27% reduction in rates of colonization, respectively: HR 0.695 for CHX (95% CI, 0.606, 0.796; p < 0.001) and 0.732 (95% CI, 0.640, 0.838; p < 0.001) for CHX + COL. The CHX + COL combination was more effective for gram-negative bacteria: daily HR .534 (95% CI, 0.455, 0.626; p <0.001) alone with a 47% reduction in gram-negative colonization compared to CHX.

No difference was seen in ICU mortality, duration of mechanical ventilation, or duration of ICU stay. One adverse event (tongue swelling) occurred in the CHX + COL group.

Limitations of the study include the following:

• Daily assessments on all patients were not performed;
• The placebo group had more males and more infections on admission than the other two groups, raising the question of randomization error;
• Clinical versus quantitative diagnosis of pneumonia may overestimate VAP in this study;
• It is not known how many patients were not enrolled in the study due to short anticipated ventilator times, but who later had prolonged ventilations; and
• The lack of effect on ventilator time, ICU length of stay, and mortality raises the question of the significance of these findings.

Despite these limitations, the low cost of these treatments, minimal adverse events, low risk of promoting significant antimicrobial resistance, and the finding of decreased VAP and bacterial colonization risk shown in this study support the potential benefit of topical decontamination with CHX and COL in conjunction with other measures of VAP prevention. TH

Reference

1. De Riso AJ II, Ladowski JS, Dillon TA, et al. Chlorhexidine gluconate 0.12% oral rinse reduces the incidence of total nosocomial respiratory infection and nonprophylactic systemic antibiotic use in patients undergoing heart surgery. Chest. 1996;109:1556-1561.

Semi-Recumbent Position to Prevent Ventilator-Associated Pneumonia: Is It Possible?

By Joseph Ming Wah Li, MD

Van Nieuwenhoven CA, Vandenbroucke-Grauls C, van Tiel FH, et al. Feasibility and effects of the semirecumbent position to prevent ventilator-associated pneumonia: a randomized study. Crit Care Med. 2006 Feb;34(2):396-402.

Ventilator-associated pneumonia (VAP) is a cause of significant morbidity and mortality among mechanically ventilated patients. Studies with radioactive-labeled enteral feeds have demonstrated an increased frequency of endotracheal aspiration of gastric contents in supine patients. The CDC guidelines for prevention of nosocomial pneumonia advise placement of mechanically ventilated patients in a semi-recumbent position as a VAP prevention measure.

Only one previous study, by Drakulovic and colleagues, has assessed this strategy to prevent VAP.1 That study demonstrated a 75% decrease in the incidence of VAP. But van Nieuwenhoven and colleagues raised two important questions about the findings from the previous study: First, the Drakulovic study placed control patients in a horizontal (zero degrees) position, which is not the standard of care in most ICUs. Most patients are placed at 10 degrees, and this position is elevated as patients are weaned. Second, the Drakulovic study measured patients only once daily but did not monitor their body positions in between the daily measurements.

Dr. van Nieuwenhoven and colleagues set out to determine whether it is feasible to keep mechanically ventilated patients in a semi-recumbent position on a continual basis and whether this measure would prevent VAP. This was a prospective multi-centered trial in which mechanically ventilated patients were randomly assigned to the semi-recumbent position with a target backrest elevation of 45 degrees or standard of care (supine position) with a backrest elevation of 10 degrees. They used a transducer with a pendulum, which was placed on the bed frame to measure the backrest elevation every 60 seconds for up to seven days. They calculated a mean degree of elevation for each patient daily. Nurses always respected the patient’s request for positioning, but a dedicated research nurse restored backrest position to the randomized position whenever possible.

Baseline characteristics for both groups were similar. For the supine (control) group, average elevations were 9.8 degrees on day one and 16.1 degrees on day seven. For the semi-recumbent group, average elevations were 28.1 degrees on day one and 22.6 degrees on day seven. There were no significant differences in numbers of patients who developed VAP in either group.

This study suggests that, despite the use of dedicated research nurses to maintain positioning, it may not be possible to keep patients’ backrests elevated to 45 degrees. Keeping patients’ backrests at an elevation of nearly 30 degrees does not appear to prevent VAP more than keeping patients’ backrests at 10 degrees, the present standard of care.

Reference

1. Drakulovic MB, Torres A, Bauer TT, et al. Supine body position as a risk factor for nosocomial pneumonia in mechanically ventilated patients: a randomised trial. Lancet. 1999;354(9193):1851-1858.

Bar Codes in Medicine: An Opportunity for Quality Improvement

By Alex Carbo, MD

Poon EG, Cina JL, Churchill W, et al. Medication dispensing errors and potential adverse drug events before and after implementing bar code technology in the pharmacy. Ann Intern Med. 2006;145:426-434.

Medication errors and adverse drug events (ADEs) have received much attention in the literature; the use of health information technology to mitigate these errors and ADEs has now been proposed in many areas of healthcare. In an effort to decrease medication-dispensing errors, the U.S. Food and Drug Administration (FDA) mandated bar code use for all medications in hospitals, beginning in April 2006. While this technology has been extensively studied in other industries, there is little data describing its effects in the healthcare system.

Poon and colleagues set out to evaluate whether implementation of bar code technology reduced dispensing errors and the ADEs that might be caused by these miscalculations. In a before-and-after evaluation, they studied more than 350,000 dispensed medication doses in an academic medical center between February 2003 and September 2004.

During the bar code conversion process, the hospital pharmacy built a dedicated repackaging center, which was responsible for affixing a bar code to every dose of medication. These medications were then dispensed in three different configurations: two configurations required staff to verify all doses at least once using bar code scanning, and the third configuration—for commonly dispensed medications that could not be accommodated in a standard carousel machine because of their size or need for refrigeration—required scanning only one dose from each batch.

The authors found a 93% to 96% relative reduction in the incidence of target dispensing errors (P<0.001) and an 86% to 97% relative reduction in the incidence of potential ADEs (P<0.001) in the two configurations that required staff to verify all doses by scanning. The greatest reductions were seen in wrong medication errors (56%), wrong strength/dose errors (71%), wrong formulation errors (90%), and expired medication errors (100%).

In the configuration that did not require scanning of every dose, however, there was a 60% relative reduction in the incidence of target dispensing errors (P<0.001), but a 2.4-fold increase in the incidence of target potential ADEs. This included new errors attributable to wrong strength and wrong medication dispensing.

In light of the FDA’s mandate regarding bar codes, it seems that every hospital has the opportunity to improve patient safety and decrease medication error rates with the use of bar code technology. This study suggests that in order to achieve this benefit these systems should be designed to ensure that every medication dose is verified by scanning during the dispensing process.

Evaluation of a Guideline to Guide Resuscitation

By Cindy Lien, MD

Morrison LJ, Visentin LM, Kiss A, et al. Validation of a rule for termination of resuscitation in out-of-hospital cardiac arrest. N Engl J Med. 2006 Aug 3;355(5):478-487.

The survival rate of patients with out-of-hospital cardiac arrest is very low. Thus, guidelines have been developed for termination of resuscitation for those patients who have had no response to advanced cardiac life support provided by emergency medical service (EMS) personnel. Similar guidelines have not yet been developed, however, for situations in which patients receive basic life support from emergency workers trained in the use of an automated external cardiac defibrillator. Patients with little potential for survival are routinely transported to emergency departments, at significant cost to the healthcare system.

Morrison and colleagues present results from the Termination of Resuscitation (TOR) study, a prospective evaluation of a clinical prediction rule for the termination of basic life support by emergency medical personnel trained in the use of automated external defibrillators. The clinical prediction rule, previously developed in a retrospective review of case records from a large urban EMS system, recommends termination of resuscitation if there is no return of spontaneous circulation, no shock administered, and no witness of the arrest by EMS personnel.

In the current study, the authors obtained follow-up data for 1,240 adult patients in Ontario, Canada, who had suffered an arrest of presumed cardiac cause and were subsequently transported to the emergency department after resuscitative efforts. Twenty-four EMS systems participated in the study. The study found that only 0.5% of the patients for whom the clinical prediction rule recommended termination survived (four out of 776 patients). Of the 1,240 total study patients, 41 (3%) survived. The clinical prediction rule recommended continuation of resuscitative efforts for 37 of these 41 patients, resulting in a specificity of 90.2%. The positive predictive value for death was calculated to be 99.5% when termination was recommended.

The TOR trial also determined whether the addition of other criteria to the original prediction rule could further refine the specificity and positive predictive value. They found that the addition to the criteria of a response time greater than eight minutes increased the positive predictive value and specificity to 99.7% and 97.6%, respectively. When the variable “not witnessed by bystander” was added to the clinical prediction rule, both the positive predictive value and specificity increased to 100%. In other words, no patients survived if they had had a completely unwitnessed arrest, no return of spontaneous circulation, and no shocks delivered.

This study identifies a subpopulation of patients with presumed cardiac arrest for whom termination of resuscitative efforts in the field appears reasonable. The authors note that a survival rate of 1% or less has been suggested in past literature as reflective of medical futility. The TOR investigators acknowledge that their study took place before the 2005 AHA Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care were released and that their study protocols were consistent with the 2000 resuscitation guidelines. In light of this information, continued validity testing of the clinical prediction rule under the 2005 AHA protocols is needed.

Nonetheless, it is quite impressive that use of their clinical prediction rule would have resulted in transportation of only 37% of patients (464 of 1,240), rather than 100% of patients, as is currently the practice. If the guidelines described in this article are to be implemented, further studies are necessary to address the training of EMS personnel, who would carry responsibility for terminating resuscitation and notifying families of patients’ deaths.

Prevention of Ventilator-Associated Pneumonia

By Diane Sliwka, MD

Koeman M, van der Ven AJ, Hak E, et al. Oral decontamination with chlorhexidine reduces the incidence of ventilator-associated pneumonia. Am J Respir Crit Care Med. 2006 Jun;173(12):1348-1355. Epub 2006 Apr 7.

Ventilator-associated pneumonia (VAP) is an important nosocomial source of morbidity and mortality. The use of prophylactic antimicrobials to decrease VAP raises concern for antimicrobial resistance. This study evaluates the topical antiseptic chlorhexidine (CHX) as an alternative prophylactic intervention for VAP. CHX has previously been shown to decrease VAP in cardiac surgical patients, but has not been studied in higher risk, long-term-ventilated patients. Because CHX works better for gram-positive organisms, the combination of colistin and CHX (COL + CHX) was also studied for improved gram-negative coverage.

This multi-center, randomized, double-blind, placebo-controlled trial enrolled 385 adult patients. Patients who were expected to be intubated for longer than 48 hours were randomized to 3 arms: CHX alone, CHX + COL, and placebo. Exclusion criteria included known preadmission immunocompromised state, pregnancy, and physical limitation to oral application. Pneumonia was defined by clinical decision-making, which was later confirmed by three blinded intensivists’ reviews of the case records and supported by daily clinical pulmonary infection scores.

The primary endpoint of VAP was diagnosed in 52 of 385 patients: 18% placebo, 13% CHX, and 10% CHX + COL. Rate of VAP in the two treatment groups was lower than placebo and reached statistical significance when compared to placebo. The daily hazard ratio for CHX versus placebo was .352 (95% CI .160, .791); for CHX + COL versus placebo, it was .454 (95% CI .224, .925), showing a 65% and 55% reduction in the rate of pneumonia development. Multivariate analysis of variables such as gender, pulmonary admission diagnosis, colonization at time of admission, and antimicrobial use on admission did not affect the data.

The secondary endpoint of endotracheal colonization was evaluated by a twice-weekly endotracheal culture. There was no statistically significant difference in colonization among the three groups in the first (days 1-4) or third (days 9-12) time frames. During the second time frame (days 5-8), there was a statistically significant decrease in colonization for the CHX + COL treatment group when compared to both placebo (16% versus 40% p<.007) and to CHX (16% versus 38%, p<.011); this decrease is thought to be due to gram-negative coverage by COL.

The secondary endpoint of oropharyngeal colonization was evaluated for 87% of all patient days. CHX and CHX + COL were similarly effective for gram-positive bacteria when compared to placebo, with 30% and 27% reduction in rates of colonization, respectively: HR 0.695 for CHX (95% CI, 0.606, 0.796; p < 0.001) and 0.732 (95% CI, 0.640, 0.838; p < 0.001) for CHX + COL. The CHX + COL combination was more effective for gram-negative bacteria: daily HR .534 (95% CI, 0.455, 0.626; p <0.001) alone with a 47% reduction in gram-negative colonization compared to CHX.

No difference was seen in ICU mortality, duration of mechanical ventilation, or duration of ICU stay. One adverse event (tongue swelling) occurred in the CHX + COL group.

Limitations of the study include the following:

• Daily assessments on all patients were not performed;
• The placebo group had more males and more infections on admission than the other two groups, raising the question of randomization error;
• Clinical versus quantitative diagnosis of pneumonia may overestimate VAP in this study;
• It is not known how many patients were not enrolled in the study due to short anticipated ventilator times, but who later had prolonged ventilations; and
• The lack of effect on ventilator time, ICU length of stay, and mortality raises the question of the significance of these findings.

Despite these limitations, the low cost of these treatments, minimal adverse events, low risk of promoting significant antimicrobial resistance, and the finding of decreased VAP and bacterial colonization risk shown in this study support the potential benefit of topical decontamination with CHX and COL in conjunction with other measures of VAP prevention. TH

Reference

1. De Riso AJ II, Ladowski JS, Dillon TA, et al. Chlorhexidine gluconate 0.12% oral rinse reduces the incidence of total nosocomial respiratory infection and nonprophylactic systemic antibiotic use in patients undergoing heart surgery. Chest. 1996;109:1556-1561.