David Kagen, MD

Clinical question: Do on-screen, computer-based clinical reminders improve adherence to target processes of care or clinical outcomes?

Background: Gaps between practice guidelines and routine care are caused, in part, by the inability of clinicians to access or recall information at the point of care. Although automated reminder systems offer the promise of “just in time” recommendations, studies of electronic reminders have demonstrated mixed results.

Study design: Literature review and meta-analysis.

Setting: Multiple databases and information repositories, including MEDLINE, EMBASE, and CINAHL.

Synopsis: The authors conducted a literature search to identify randomized and quasi-randomized controlled trials measuring the effect of computer-based reminders on process measures or clinical outcomes. To avoid statistical challenges inherent in unit-of-analysis errors, the authors reported median improvement in process adherence or median change in clinical endpoints.

Out of a pool of 2,036 citations, 28 studies detailing 32 comparative analyses were included. Across the 28 studies, reminders resulted in a median improvement in target process adherence of 4.2% (3.3% for prescribing behavior, 2.8% for test ordering). Eight comparisons reported dichotomous clinical endpoints and collectively showed a median absolute improvement of 2.5%.

The greatest contribution to measured treatment effects came from large academic centers with well-established electronic health records and robust informatics departments. No characteristics of the reminder system or the clinical context were associated with the magnitude of impact. A potential limitation in reporting median effects across studies is that all studies were given equal weight.

Bottom line: Electronic reminders appear to have a small, positive effect on clinician adherence to recommended processes, although it is uncertain what contextual or design features are responsible for the greatest treatment effect.

Citation: Shojania K, Jennings A, Mayhew A, Ramsay CR, Eccles MP, Grimshaw J. The effects of on-screen, point of care computer reminders on processes and outcomes of care. Cochrane Database Syst Rev. 2009(3):CD001096. TH

Clinical question: Do on-screen, computer-based clinical reminders improve adherence to target processes of care or clinical outcomes?

Background: Gaps between practice guidelines and routine care are caused, in part, by the inability of clinicians to access or recall information at the point of care. Although automated reminder systems offer the promise of “just in time” recommendations, studies of electronic reminders have demonstrated mixed results.

Study design: Literature review and meta-analysis.

Setting: Multiple databases and information repositories, including MEDLINE, EMBASE, and CINAHL.

Synopsis: The authors conducted a literature search to identify randomized and quasi-randomized controlled trials measuring the effect of computer-based reminders on process measures or clinical outcomes. To avoid statistical challenges inherent in unit-of-analysis errors, the authors reported median improvement in process adherence or median change in clinical endpoints.

Out of a pool of 2,036 citations, 28 studies detailing 32 comparative analyses were included. Across the 28 studies, reminders resulted in a median improvement in target process adherence of 4.2% (3.3% for prescribing behavior, 2.8% for test ordering). Eight comparisons reported dichotomous clinical endpoints and collectively showed a median absolute improvement of 2.5%.

The greatest contribution to measured treatment effects came from large academic centers with well-established electronic health records and robust informatics departments. No characteristics of the reminder system or the clinical context were associated with the magnitude of impact. A potential limitation in reporting median effects across studies is that all studies were given equal weight.

Bottom line: Electronic reminders appear to have a small, positive effect on clinician adherence to recommended processes, although it is uncertain what contextual or design features are responsible for the greatest treatment effect.

Citation: Shojania K, Jennings A, Mayhew A, Ramsay CR, Eccles MP, Grimshaw J. The effects of on-screen, point of care computer reminders on processes and outcomes of care. Cochrane Database Syst Rev. 2009(3):CD001096. TH

Clinical question: Do on-screen, computer-based clinical reminders improve adherence to target processes of care or clinical outcomes?

Background: Gaps between practice guidelines and routine care are caused, in part, by the inability of clinicians to access or recall information at the point of care. Although automated reminder systems offer the promise of “just in time” recommendations, studies of electronic reminders have demonstrated mixed results.

Study design: Literature review and meta-analysis.

Setting: Multiple databases and information repositories, including MEDLINE, EMBASE, and CINAHL.

Synopsis: The authors conducted a literature search to identify randomized and quasi-randomized controlled trials measuring the effect of computer-based reminders on process measures or clinical outcomes. To avoid statistical challenges inherent in unit-of-analysis errors, the authors reported median improvement in process adherence or median change in clinical endpoints.

Out of a pool of 2,036 citations, 28 studies detailing 32 comparative analyses were included. Across the 28 studies, reminders resulted in a median improvement in target process adherence of 4.2% (3.3% for prescribing behavior, 2.8% for test ordering). Eight comparisons reported dichotomous clinical endpoints and collectively showed a median absolute improvement of 2.5%.

The greatest contribution to measured treatment effects came from large academic centers with well-established electronic health records and robust informatics departments. No characteristics of the reminder system or the clinical context were associated with the magnitude of impact. A potential limitation in reporting median effects across studies is that all studies were given equal weight.

Bottom line: Electronic reminders appear to have a small, positive effect on clinician adherence to recommended processes, although it is uncertain what contextual or design features are responsible for the greatest treatment effect.

Citation: Shojania K, Jennings A, Mayhew A, Ramsay CR, Eccles MP, Grimshaw J. The effects of on-screen, point of care computer reminders on processes and outcomes of care. Cochrane Database Syst Rev. 2009(3):CD001096. TH

Clinical question: Does the inclusion of a medication adherence counseling session during a hospital discharge reconciliation process reduce discrepancies in the final medication regimen?

Background: Inadvertent medication prescribing errors are an important cause of preventable adverse drug events and commonly occur at transitions of care. Although medication reconciliation processes can identify errors, the best strategies for implementation remain unclear.

Study design: Prospective, observational cohort.

Setting: A 550-bed teaching hospital in the Netherlands.

Synopsis: Of 437 patients admitted to a pulmonary ward and screened for eligibility, 267 were included in the analysis. A pharmacy specialist reviewed all available community prescription records, inpatient documentation, and discharge medication lists in an effort to identify discrepancies. Potential errors were discussed with the prescriber. Then, the pharmacy specialist interviewed the patient and provided additional counseling. Any new discrepancies were discussed with the prescriber. All questions raised by the pharmacist were recorded, as were all subsequent prescriber interventions.

The primary outcome measure was the number of interventions made as a result of pharmacy review. A total of 940 questions were asked. At least one intervention was recorded for 87% of patients before counseling (mean 2.7 interventions/patient) and for 97% of patients after (mean 5.3 interventions/patient). Discrepancies were addressed for 63.7% of patients before counseling and 72.5% after. Pharmacotherapy was optimized for 67.2% of patients before counseling and 76.3% after.

Bottom line: Patient engagement in the medication reconciliation process incrementally improves the quality of the history and helps identify clinically meaningful discrepancies at the time of hospital discharge.

Citation: Karapinar-Carkit F, Borgsteede S, Zoer J, Smit HJ, Egberts AC, van den Bemt P. Effect of medication reconciliation with and without patient counseling on the number of pharmaceutical interventions among patients discharged from the hospital. Ann Pharmacother. 2009;43(6):1001-1010.

Clinical question: Does the inclusion of a medication adherence counseling session during a hospital discharge reconciliation process reduce discrepancies in the final medication regimen?

Background: Inadvertent medication prescribing errors are an important cause of preventable adverse drug events and commonly occur at transitions of care. Although medication reconciliation processes can identify errors, the best strategies for implementation remain unclear.

Study design: Prospective, observational cohort.

Setting: A 550-bed teaching hospital in the Netherlands.

Synopsis: Of 437 patients admitted to a pulmonary ward and screened for eligibility, 267 were included in the analysis. A pharmacy specialist reviewed all available community prescription records, inpatient documentation, and discharge medication lists in an effort to identify discrepancies. Potential errors were discussed with the prescriber. Then, the pharmacy specialist interviewed the patient and provided additional counseling. Any new discrepancies were discussed with the prescriber. All questions raised by the pharmacist were recorded, as were all subsequent prescriber interventions.

The primary outcome measure was the number of interventions made as a result of pharmacy review. A total of 940 questions were asked. At least one intervention was recorded for 87% of patients before counseling (mean 2.7 interventions/patient) and for 97% of patients after (mean 5.3 interventions/patient). Discrepancies were addressed for 63.7% of patients before counseling and 72.5% after. Pharmacotherapy was optimized for 67.2% of patients before counseling and 76.3% after.

Bottom line: Patient engagement in the medication reconciliation process incrementally improves the quality of the history and helps identify clinically meaningful discrepancies at the time of hospital discharge.

Citation: Karapinar-Carkit F, Borgsteede S, Zoer J, Smit HJ, Egberts AC, van den Bemt P. Effect of medication reconciliation with and without patient counseling on the number of pharmaceutical interventions among patients discharged from the hospital. Ann Pharmacother. 2009;43(6):1001-1010.

Clinical question: Does the inclusion of a medication adherence counseling session during a hospital discharge reconciliation process reduce discrepancies in the final medication regimen?

Background: Inadvertent medication prescribing errors are an important cause of preventable adverse drug events and commonly occur at transitions of care. Although medication reconciliation processes can identify errors, the best strategies for implementation remain unclear.

Study design: Prospective, observational cohort.

Setting: A 550-bed teaching hospital in the Netherlands.

Synopsis: Of 437 patients admitted to a pulmonary ward and screened for eligibility, 267 were included in the analysis. A pharmacy specialist reviewed all available community prescription records, inpatient documentation, and discharge medication lists in an effort to identify discrepancies. Potential errors were discussed with the prescriber. Then, the pharmacy specialist interviewed the patient and provided additional counseling. Any new discrepancies were discussed with the prescriber. All questions raised by the pharmacist were recorded, as were all subsequent prescriber interventions.

The primary outcome measure was the number of interventions made as a result of pharmacy review. A total of 940 questions were asked. At least one intervention was recorded for 87% of patients before counseling (mean 2.7 interventions/patient) and for 97% of patients after (mean 5.3 interventions/patient). Discrepancies were addressed for 63.7% of patients before counseling and 72.5% after. Pharmacotherapy was optimized for 67.2% of patients before counseling and 76.3% after.

Bottom line: Patient engagement in the medication reconciliation process incrementally improves the quality of the history and helps identify clinically meaningful discrepancies at the time of hospital discharge.

Citation: Karapinar-Carkit F, Borgsteede S, Zoer J, Smit HJ, Egberts AC, van den Bemt P. Effect of medication reconciliation with and without patient counseling on the number of pharmaceutical interventions among patients discharged from the hospital. Ann Pharmacother. 2009;43(6):1001-1010.

Clinical question: In patients with symptoms consistent with pulmonary embolism (PE), can evaluation with a clinical risk assessment tool and D-dimer assay identify patients who do not require CT angiography to exclude PE?

Background: D-dimer is a highly sensitive but nonspecific marker of VTE, and studies suggest that VTE can be ruled out without further imaging in patients with low clinical probability of disease and a negative D-dimer test. Nevertheless, this practice has not been adopted uniformly, and CT angiography (CTA) overuse continues.

Study design: Prospective registry cohort.

Setting: A 550-bed community teaching hospital in Chicago.

Synopsis: Consecutive patients presenting to the ED with symptoms suggestive of PE were evaluated with 1) revised Geneva score; 2) D-dimer assay; and 3) CTA. Among the 627 patients who underwent all three components of the evaluation, 44.8% were identified as low probability for PE by revised Geneva score, 52.6% as intermediate probability, and 2.6% as high probability. The overall prevalence of PE (using CTA as the gold standard) was very low (4.5%); just 2.1% of low-risk, 5.2% of intermediate-risk, and 31.2% of high-risk patients were ultimately found to have PE on CTA.

Using a cutoff of 1.2 mg/L, the D-dimer assay accurately detected all low- to intermediate-probability patients with PE (sensitivity and negative predictive value of 100%). One patient in the high probability group did have a PE, even though the patient had a D-dimer value <1.2 mg/L (sensitivity and NPV both 80%). Had diagnostic testing stopped after a negative D-dimer result in the low- to intermediate-probability patients, 172 CTAs (27%) would have been avoided.

Bottom line: In a low-prevalence cohort, no pulmonary emboli were identified by CTA in any patient with a low to intermediate clinical risk assessment and a negative quantitative D-dimer assay result.

Citation: Gupta RT, Kakarla RK, Kirshenbaum KJ, Tapson VF. D-dimers and efficacy of clinical risk estimation algorithms: sensitivity in evaluation of acute pulmonary embolism. AJR Am J Roentgenol. 2009;193(2):425-430.

Clinical question: In patients with symptoms consistent with pulmonary embolism (PE), can evaluation with a clinical risk assessment tool and D-dimer assay identify patients who do not require CT angiography to exclude PE?

Background: D-dimer is a highly sensitive but nonspecific marker of VTE, and studies suggest that VTE can be ruled out without further imaging in patients with low clinical probability of disease and a negative D-dimer test. Nevertheless, this practice has not been adopted uniformly, and CT angiography (CTA) overuse continues.

Study design: Prospective registry cohort.

Setting: A 550-bed community teaching hospital in Chicago.

Synopsis: Consecutive patients presenting to the ED with symptoms suggestive of PE were evaluated with 1) revised Geneva score; 2) D-dimer assay; and 3) CTA. Among the 627 patients who underwent all three components of the evaluation, 44.8% were identified as low probability for PE by revised Geneva score, 52.6% as intermediate probability, and 2.6% as high probability. The overall prevalence of PE (using CTA as the gold standard) was very low (4.5%); just 2.1% of low-risk, 5.2% of intermediate-risk, and 31.2% of high-risk patients were ultimately found to have PE on CTA.

Using a cutoff of 1.2 mg/L, the D-dimer assay accurately detected all low- to intermediate-probability patients with PE (sensitivity and negative predictive value of 100%). One patient in the high probability group did have a PE, even though the patient had a D-dimer value <1.2 mg/L (sensitivity and NPV both 80%). Had diagnostic testing stopped after a negative D-dimer result in the low- to intermediate-probability patients, 172 CTAs (27%) would have been avoided.

Bottom line: In a low-prevalence cohort, no pulmonary emboli were identified by CTA in any patient with a low to intermediate clinical risk assessment and a negative quantitative D-dimer assay result.

Citation: Gupta RT, Kakarla RK, Kirshenbaum KJ, Tapson VF. D-dimers and efficacy of clinical risk estimation algorithms: sensitivity in evaluation of acute pulmonary embolism. AJR Am J Roentgenol. 2009;193(2):425-430.

Clinical question: In patients with symptoms consistent with pulmonary embolism (PE), can evaluation with a clinical risk assessment tool and D-dimer assay identify patients who do not require CT angiography to exclude PE?

Background: D-dimer is a highly sensitive but nonspecific marker of VTE, and studies suggest that VTE can be ruled out without further imaging in patients with low clinical probability of disease and a negative D-dimer test. Nevertheless, this practice has not been adopted uniformly, and CT angiography (CTA) overuse continues.

Study design: Prospective registry cohort.

Setting: A 550-bed community teaching hospital in Chicago.

Synopsis: Consecutive patients presenting to the ED with symptoms suggestive of PE were evaluated with 1) revised Geneva score; 2) D-dimer assay; and 3) CTA. Among the 627 patients who underwent all three components of the evaluation, 44.8% were identified as low probability for PE by revised Geneva score, 52.6% as intermediate probability, and 2.6% as high probability. The overall prevalence of PE (using CTA as the gold standard) was very low (4.5%); just 2.1% of low-risk, 5.2% of intermediate-risk, and 31.2% of high-risk patients were ultimately found to have PE on CTA.

Using a cutoff of 1.2 mg/L, the D-dimer assay accurately detected all low- to intermediate-probability patients with PE (sensitivity and negative predictive value of 100%). One patient in the high probability group did have a PE, even though the patient had a D-dimer value <1.2 mg/L (sensitivity and NPV both 80%). Had diagnostic testing stopped after a negative D-dimer result in the low- to intermediate-probability patients, 172 CTAs (27%) would have been avoided.

Bottom line: In a low-prevalence cohort, no pulmonary emboli were identified by CTA in any patient with a low to intermediate clinical risk assessment and a negative quantitative D-dimer assay result.

Citation: Gupta RT, Kakarla RK, Kirshenbaum KJ, Tapson VF. D-dimers and efficacy of clinical risk estimation algorithms: sensitivity in evaluation of acute pulmonary embolism. AJR Am J Roentgenol. 2009;193(2):425-430.

Clinical question: Do signouts vary in the quality and quantity of information, and what are the various factors affecting signout quality?

Background: Miscommunication during transfers of responsibility for hospitalized patients is common and can result in harm. Recommendations for safe and effective handoffs emphasize key content, clear communication, senior staff supervision, and adequate time for questions. Still, little is known about adherence to these recommendations in clinical practice.

Study design: Prospective, observational cohort.

Setting: Medical unit of an acute-care teaching hospital.

Synopsis: Oral signouts were audiotaped among IM house staff teams and the accompanying written signouts were collected for review of content. Signout sessions (n=88) included eight IM teams at one hospital and contained 503 patient signouts.

The median signout duration was 35 seconds (IQR 19-62) per patient. Key clinical information was present in just 62% of combined written or oral signouts. Most signouts included no questions from the recipient. Factors associated with higher rate of content inclusion included: familiarity with the patient, sense of responsibility (primary team vs. covering team), only one signout per day (as compared to sequential signout), presence of a senior resident, and comprehensive, written signouts.

Study limitations include the Hawthorne effect, as several participants mentioned that the presence of audiotape led to more comprehensive signouts than are typical. Also, the signout quality assessment in this study has not been validated with patient-safety outcomes.

Bottom line: Signouts among internal-medicine residents at this one hospital showed variability in terms of quantitative and qualitative information and often missed crucial information about patient care.

Citation: Horwitz LI, Moin T, Krumholz HM, Wang L, Bradley EH. What are covering doctors told about their patients? Analysis of sign-out among internal medicine house staff. Qual Saf Health Care. 2009;18(4):248-255.

Clinical question: Do signouts vary in the quality and quantity of information, and what are the various factors affecting signout quality?

Background: Miscommunication during transfers of responsibility for hospitalized patients is common and can result in harm. Recommendations for safe and effective handoffs emphasize key content, clear communication, senior staff supervision, and adequate time for questions. Still, little is known about adherence to these recommendations in clinical practice.

Study design: Prospective, observational cohort.

Setting: Medical unit of an acute-care teaching hospital.

Synopsis: Oral signouts were audiotaped among IM house staff teams and the accompanying written signouts were collected for review of content. Signout sessions (n=88) included eight IM teams at one hospital and contained 503 patient signouts.

The median signout duration was 35 seconds (IQR 19-62) per patient. Key clinical information was present in just 62% of combined written or oral signouts. Most signouts included no questions from the recipient. Factors associated with higher rate of content inclusion included: familiarity with the patient, sense of responsibility (primary team vs. covering team), only one signout per day (as compared to sequential signout), presence of a senior resident, and comprehensive, written signouts.

Study limitations include the Hawthorne effect, as several participants mentioned that the presence of audiotape led to more comprehensive signouts than are typical. Also, the signout quality assessment in this study has not been validated with patient-safety outcomes.

Bottom line: Signouts among internal-medicine residents at this one hospital showed variability in terms of quantitative and qualitative information and often missed crucial information about patient care.

Citation: Horwitz LI, Moin T, Krumholz HM, Wang L, Bradley EH. What are covering doctors told about their patients? Analysis of sign-out among internal medicine house staff. Qual Saf Health Care. 2009;18(4):248-255.

Clinical question: Do signouts vary in the quality and quantity of information, and what are the various factors affecting signout quality?

Background: Miscommunication during transfers of responsibility for hospitalized patients is common and can result in harm. Recommendations for safe and effective handoffs emphasize key content, clear communication, senior staff supervision, and adequate time for questions. Still, little is known about adherence to these recommendations in clinical practice.

Study design: Prospective, observational cohort.

Setting: Medical unit of an acute-care teaching hospital.

Synopsis: Oral signouts were audiotaped among IM house staff teams and the accompanying written signouts were collected for review of content. Signout sessions (n=88) included eight IM teams at one hospital and contained 503 patient signouts.

The median signout duration was 35 seconds (IQR 19-62) per patient. Key clinical information was present in just 62% of combined written or oral signouts. Most signouts included no questions from the recipient. Factors associated with higher rate of content inclusion included: familiarity with the patient, sense of responsibility (primary team vs. covering team), only one signout per day (as compared to sequential signout), presence of a senior resident, and comprehensive, written signouts.

Study limitations include the Hawthorne effect, as several participants mentioned that the presence of audiotape led to more comprehensive signouts than are typical. Also, the signout quality assessment in this study has not been validated with patient-safety outcomes.

Bottom line: Signouts among internal-medicine residents at this one hospital showed variability in terms of quantitative and qualitative information and often missed crucial information about patient care.

Citation: Horwitz LI, Moin T, Krumholz HM, Wang L, Bradley EH. What are covering doctors told about their patients? Analysis of sign-out among internal medicine house staff. Qual Saf Health Care. 2009;18(4):248-255.

Clinical question: How does traditional, oral signout from emergency providers to inpatient medicine physicians compare to dictated, voicemail signout?

Background: Communication failures contribute to errors in care transition from ED to inpatient medicine units. Signout between ED providers and internal medicine (IM) physicians is typically oral (“synchronous communication”). It is not known how dictated signout to a voicemail system (“asynchronous communication”) affects the quality and safety of handoff communications.

Study design: Prospective, pre-post analysis.

Setting: A 944-bed urban academic medical center in Connecticut.

Synopsis: Surveys were administered to all IM and ED providers before and after the implementation of a voicemail signout system. In the new system, ED providers dictated signout for stable patients, rather than giving traditional synchronous telephone signout. It was the responsibility of the admitting IM physician to listen to the voicemail after receiving a text notification that a patient was being admitted.

ED providers recorded signouts in 89.5% of medicine admissions. However, voicemails were accessed only 58.5% of the time by receiving physicians. All ED providers and 56% of IM physicians believed signout was easier following the voicemail intervention. Overall, ED providers gave the quality, content, and accuracy of their signout communication higher ratings than IM physicians did; 69% of all providers felt the interaction among participants was worse following the intervention. There was no change in the rate of perceived adverse events or ICU transfers within 24 hours after admission.

This intervention was a QI initiative at a single center. Mixed results and small sample size limit generalizability of the study.

Bottom line: Asynchronous signout by voicemail increased efficiency, particularly among ED providers but decreased perceived quality of interaction between medical providers without obviously affecting patient safety.

Citation: Horwitz LI, Parwani V, Shah NR, et al. Evaluation of an asynchronous physician voicemail sign-out for emergency department admissions. Ann Emerg Med. 2009;54:368-378.

Clinical question: How does traditional, oral signout from emergency providers to inpatient medicine physicians compare to dictated, voicemail signout?

Background: Communication failures contribute to errors in care transition from ED to inpatient medicine units. Signout between ED providers and internal medicine (IM) physicians is typically oral (“synchronous communication”). It is not known how dictated signout to a voicemail system (“asynchronous communication”) affects the quality and safety of handoff communications.

Study design: Prospective, pre-post analysis.

Setting: A 944-bed urban academic medical center in Connecticut.

Synopsis: Surveys were administered to all IM and ED providers before and after the implementation of a voicemail signout system. In the new system, ED providers dictated signout for stable patients, rather than giving traditional synchronous telephone signout. It was the responsibility of the admitting IM physician to listen to the voicemail after receiving a text notification that a patient was being admitted.

ED providers recorded signouts in 89.5% of medicine admissions. However, voicemails were accessed only 58.5% of the time by receiving physicians. All ED providers and 56% of IM physicians believed signout was easier following the voicemail intervention. Overall, ED providers gave the quality, content, and accuracy of their signout communication higher ratings than IM physicians did; 69% of all providers felt the interaction among participants was worse following the intervention. There was no change in the rate of perceived adverse events or ICU transfers within 24 hours after admission.

This intervention was a QI initiative at a single center. Mixed results and small sample size limit generalizability of the study.

Bottom line: Asynchronous signout by voicemail increased efficiency, particularly among ED providers but decreased perceived quality of interaction between medical providers without obviously affecting patient safety.

Citation: Horwitz LI, Parwani V, Shah NR, et al. Evaluation of an asynchronous physician voicemail sign-out for emergency department admissions. Ann Emerg Med. 2009;54:368-378.

Clinical question: How does traditional, oral signout from emergency providers to inpatient medicine physicians compare to dictated, voicemail signout?

Background: Communication failures contribute to errors in care transition from ED to inpatient medicine units. Signout between ED providers and internal medicine (IM) physicians is typically oral (“synchronous communication”). It is not known how dictated signout to a voicemail system (“asynchronous communication”) affects the quality and safety of handoff communications.

Study design: Prospective, pre-post analysis.

Setting: A 944-bed urban academic medical center in Connecticut.

Synopsis: Surveys were administered to all IM and ED providers before and after the implementation of a voicemail signout system. In the new system, ED providers dictated signout for stable patients, rather than giving traditional synchronous telephone signout. It was the responsibility of the admitting IM physician to listen to the voicemail after receiving a text notification that a patient was being admitted.

ED providers recorded signouts in 89.5% of medicine admissions. However, voicemails were accessed only 58.5% of the time by receiving physicians. All ED providers and 56% of IM physicians believed signout was easier following the voicemail intervention. Overall, ED providers gave the quality, content, and accuracy of their signout communication higher ratings than IM physicians did; 69% of all providers felt the interaction among participants was worse following the intervention. There was no change in the rate of perceived adverse events or ICU transfers within 24 hours after admission.

This intervention was a QI initiative at a single center. Mixed results and small sample size limit generalizability of the study.

Bottom line: Asynchronous signout by voicemail increased efficiency, particularly among ED providers but decreased perceived quality of interaction between medical providers without obviously affecting patient safety.

Citation: Horwitz LI, Parwani V, Shah NR, et al. Evaluation of an asynchronous physician voicemail sign-out for emergency department admissions. Ann Emerg Med. 2009;54:368-378.

Clinical question: What is the frequency and nature of undesirable events experienced by patients who “board” in the ED?

Background: Hospital crowding results in patients spending extended amounts of time—also known as “boarding”—in the ED as they wait for an inpatient bed. Prior studies have shown that longer ED boarding times are associated with adverse outcomes. Few studies have examined the nature and frequency of undesirable events that patients experience while boarding.

Study design: Retrospective chart review.

Setting: Urban academic medical center.

Synopsis: In this pilot study, authors reviewed the charts of patients who were treated in the ED and subsequently admitted to the hospital on three different days during the study period (n=151). More than a quarter (27.8%) of patients experienced an undesirable event, such as missing a scheduled medication, while they were boarding. Older patients, those with comorbid illnesses, and those who endured prolonged boarding times (greater than six hours) were more likely to experience an undesirable event. In addition, 3.3% of patients experienced such adverse events as suboptimal blood pressure control, hypotension, hypoxia, or arrhythmia.

This study was performed at a single center and lacks a comparison group (i.e., nonboarded patients). It is intended to serve as an exploratory study for future analysis of adverse events in boarded patients.

Bottom line: Undesirable events are common among boarded patients, although it is unknown whether they are more common than in nonboarded patients.

Citation: Liu SW, Thomas SH, Gordon JA, Hamedani AG, Weissman JS. A pilot study examining undesirable events among emergency-department boarded patients awaiting inpatient beds. Ann Emerg Med. 2009;54(3):381-385.

Clinical question: What is the frequency and nature of undesirable events experienced by patients who “board” in the ED?

Background: Hospital crowding results in patients spending extended amounts of time—also known as “boarding”—in the ED as they wait for an inpatient bed. Prior studies have shown that longer ED boarding times are associated with adverse outcomes. Few studies have examined the nature and frequency of undesirable events that patients experience while boarding.

Study design: Retrospective chart review.

Setting: Urban academic medical center.

Synopsis: In this pilot study, authors reviewed the charts of patients who were treated in the ED and subsequently admitted to the hospital on three different days during the study period (n=151). More than a quarter (27.8%) of patients experienced an undesirable event, such as missing a scheduled medication, while they were boarding. Older patients, those with comorbid illnesses, and those who endured prolonged boarding times (greater than six hours) were more likely to experience an undesirable event. In addition, 3.3% of patients experienced such adverse events as suboptimal blood pressure control, hypotension, hypoxia, or arrhythmia.

This study was performed at a single center and lacks a comparison group (i.e., nonboarded patients). It is intended to serve as an exploratory study for future analysis of adverse events in boarded patients.

Bottom line: Undesirable events are common among boarded patients, although it is unknown whether they are more common than in nonboarded patients.

Citation: Liu SW, Thomas SH, Gordon JA, Hamedani AG, Weissman JS. A pilot study examining undesirable events among emergency-department boarded patients awaiting inpatient beds. Ann Emerg Med. 2009;54(3):381-385.

Clinical question: What is the frequency and nature of undesirable events experienced by patients who “board” in the ED?

Background: Hospital crowding results in patients spending extended amounts of time—also known as “boarding”—in the ED as they wait for an inpatient bed. Prior studies have shown that longer ED boarding times are associated with adverse outcomes. Few studies have examined the nature and frequency of undesirable events that patients experience while boarding.

Study design: Retrospective chart review.

Setting: Urban academic medical center.

Synopsis: In this pilot study, authors reviewed the charts of patients who were treated in the ED and subsequently admitted to the hospital on three different days during the study period (n=151). More than a quarter (27.8%) of patients experienced an undesirable event, such as missing a scheduled medication, while they were boarding. Older patients, those with comorbid illnesses, and those who endured prolonged boarding times (greater than six hours) were more likely to experience an undesirable event. In addition, 3.3% of patients experienced such adverse events as suboptimal blood pressure control, hypotension, hypoxia, or arrhythmia.

This study was performed at a single center and lacks a comparison group (i.e., nonboarded patients). It is intended to serve as an exploratory study for future analysis of adverse events in boarded patients.

Bottom line: Undesirable events are common among boarded patients, although it is unknown whether they are more common than in nonboarded patients.

Citation: Liu SW, Thomas SH, Gordon JA, Hamedani AG, Weissman JS. A pilot study examining undesirable events among emergency-department boarded patients awaiting inpatient beds. Ann Emerg Med. 2009;54(3):381-385.

Clinical question: Does the reduction in work hours for residents affect mortality in medical and surgical ICUs?

Background: A reduction in work hours for residents was enforced in July 2003. Several prior studies using administrative or claims data did not show an association of the reduced work hours for residents with mortality in teaching hospitals when compared with nonteaching hospitals.

Study design: Observational retrospective registry cohort.

Setting: Twelve academic, 12 community, and 16 nonteaching hospitals in the U.S.

Synopsis: Data from 230,151 patients were extracted as post-hoc analysis from a voluntary clinical registry that uses a well-validated severity-of-illness scoring system. The exposure was defined as date of admission to ICU within two years before and after the reform. Hospitals were categorized as academic, community with residents, or nonteaching. Sophisticated statistical analyses were performed, including interaction terms for teaching status and time. To test the effect the reduced work hours had on mortality, the mortality trends of academic hospitals and community hospitals with residents were compared with the baseline trend of nonteaching hospitals. After risk adjustments, all hospitals had improved in-hospital and ICU mortality after the reform. None of the statistical improvements were significantly different.

Study limitations include the selection bias, as only highly motivated hospitals participating in the registry were included, and misclassification bias, as not all hospitals implemented the reform at the same time. Nevertheless, this study supports the consistent literature on the topic and adds a more robust assessment of severity of illness.

Bottom line: The restriction on resident duty hours does not appear to affect patient mortality.

Citation: Prasad M, Iwashyna TJ, Christie JD, et al. Effect of work-hours regulations on intensive care unit mortality in United States teaching hospitals. Crit Care Med. 2009;37(9):2564-2569.

Clinical question: Does the reduction in work hours for residents affect mortality in medical and surgical ICUs?

Background: A reduction in work hours for residents was enforced in July 2003. Several prior studies using administrative or claims data did not show an association of the reduced work hours for residents with mortality in teaching hospitals when compared with nonteaching hospitals.

Study design: Observational retrospective registry cohort.

Setting: Twelve academic, 12 community, and 16 nonteaching hospitals in the U.S.

Synopsis: Data from 230,151 patients were extracted as post-hoc analysis from a voluntary clinical registry that uses a well-validated severity-of-illness scoring system. The exposure was defined as date of admission to ICU within two years before and after the reform. Hospitals were categorized as academic, community with residents, or nonteaching. Sophisticated statistical analyses were performed, including interaction terms for teaching status and time. To test the effect the reduced work hours had on mortality, the mortality trends of academic hospitals and community hospitals with residents were compared with the baseline trend of nonteaching hospitals. After risk adjustments, all hospitals had improved in-hospital and ICU mortality after the reform. None of the statistical improvements were significantly different.

Study limitations include the selection bias, as only highly motivated hospitals participating in the registry were included, and misclassification bias, as not all hospitals implemented the reform at the same time. Nevertheless, this study supports the consistent literature on the topic and adds a more robust assessment of severity of illness.

Bottom line: The restriction on resident duty hours does not appear to affect patient mortality.

Citation: Prasad M, Iwashyna TJ, Christie JD, et al. Effect of work-hours regulations on intensive care unit mortality in United States teaching hospitals. Crit Care Med. 2009;37(9):2564-2569.

Clinical question: Does the reduction in work hours for residents affect mortality in medical and surgical ICUs?

Background: A reduction in work hours for residents was enforced in July 2003. Several prior studies using administrative or claims data did not show an association of the reduced work hours for residents with mortality in teaching hospitals when compared with nonteaching hospitals.

Study design: Observational retrospective registry cohort.

Setting: Twelve academic, 12 community, and 16 nonteaching hospitals in the U.S.

Synopsis: Data from 230,151 patients were extracted as post-hoc analysis from a voluntary clinical registry that uses a well-validated severity-of-illness scoring system. The exposure was defined as date of admission to ICU within two years before and after the reform. Hospitals were categorized as academic, community with residents, or nonteaching. Sophisticated statistical analyses were performed, including interaction terms for teaching status and time. To test the effect the reduced work hours had on mortality, the mortality trends of academic hospitals and community hospitals with residents were compared with the baseline trend of nonteaching hospitals. After risk adjustments, all hospitals had improved in-hospital and ICU mortality after the reform. None of the statistical improvements were significantly different.

Study limitations include the selection bias, as only highly motivated hospitals participating in the registry were included, and misclassification bias, as not all hospitals implemented the reform at the same time. Nevertheless, this study supports the consistent literature on the topic and adds a more robust assessment of severity of illness.

Bottom line: The restriction on resident duty hours does not appear to affect patient mortality.

Citation: Prasad M, Iwashyna TJ, Christie JD, et al. Effect of work-hours regulations on intensive care unit mortality in United States teaching hospitals. Crit Care Med. 2009;37(9):2564-2569.

In This Edition

Literature at a Glance

A guide to this month’s studies

• Resident duty hours and ICU patient outcomes
• Effect of ER boarding on patient outcomes
• ED voicemail signout of patients
• Adequacy of patient signouts
• D-dimer use in patients with low suspicion of pulmonary embolism
• Patient involvement in medication reconciliation
• Effects of on-screen reminders on outcomes

Decreased ICU Duty Hours Does Not Affect Patient Mortality

Clinical question: Does the reduction in work hours for residents affect mortality in medical and surgical ICUs?

Background: A reduction in work hours for residents was enforced in July 2003. Several prior studies using administrative or claims data did not show an association of the reduced work hours for residents with mortality in teaching hospitals when compared with nonteaching hospitals.

Study design: Observational retrospective registry cohort.

Setting: Twelve academic, 12 community, and 16 nonteaching hospitals in the U.S.

Synopsis: Data from 230,151 patients were extracted as post-hoc analysis from a voluntary clinical registry that uses a well-validated severity-of-illness scoring system. The exposure was defined as date of admission to ICU within two years before and after the reform. Hospitals were categorized as academic, community with residents, or nonteaching. Sophisticated statistical analyses were performed, including interaction terms for teaching status and time. To test the effect the reduced work hours had on mortality, the mortality trends of academic hospitals and community hospitals with residents were compared with the baseline trend of nonteaching hospitals. After risk adjustments, all hospitals had improved in-hospital and ICU mortality after the reform. None of the statistical improvements were significantly different.

Study limitations include the selection bias, as only highly motivated hospitals participating in the registry were included, and misclassification bias, as not all hospitals implemented the reform at the same time. Nevertheless, this study supports the consistent literature on the topic and adds a more robust assessment of severity of illness.

Bottom line: The restriction on resident duty hours does not appear to affect patient mortality.

Citation: Prasad M, Iwashyna TJ, Christie JD, et al. Effect of work-hours regulations on intensive care unit mortality in United States teaching hospitals. Crit Care Med. 2009;37(9):2564-2569.

Emergency Department “Boarding” Results in Undesirable Events

Clinical question: What is the frequency and nature of undesirable events experienced by patients who “board” in the ED?

Background: Hospital crowding results in patients spending extended amounts of time—also known as “boarding”—in the ED as they wait for an inpatient bed. Prior studies have shown that longer ED boarding times are associated with adverse outcomes. Few studies have examined the nature and frequency of undesirable events that patients experience while boarding.

Study design: Retrospective chart review.

Setting: Urban academic medical center.

Synopsis: In this pilot study, authors reviewed the charts of patients who were treated in the ED and subsequently admitted to the hospital on three different days during the study period (n=151). More than a quarter (27.8%) of patients experienced an undesirable event, such as missing a scheduled medication, while they were boarding. Older patients, those with comorbid illnesses, and those who endured prolonged boarding times (greater than six hours) were more likely to experience an undesirable event. In addition, 3.3% of patients experienced such adverse events as suboptimal blood pressure control, hypotension, hypoxia, or arrhythmia.

This study was performed at a single center and lacks a comparison group (i.e., nonboarded patients). It is intended to serve as an exploratory study for future analysis of adverse events in boarded patients.

Bottom line: Undesirable events are common among boarded patients, although it is unknown whether they are more common than in nonboarded patients.

Citation: Liu SW, Thomas SH, Gordon JA, Hamedani AG, Weissman JS. A pilot study examining undesirable events among emergency-department boarded patients awaiting inpatient beds. Ann Emerg Med. 2009;54(3):381-385.

Emergency Department Signout via Voicemail Yields Mixed Reviews

Clinical question: How does traditional, oral signout from emergency providers to inpatient medicine physicians compare to dictated, voicemail signout?

Background: Communication failures contribute to errors in care transition from ED to inpatient medicine units. Signout between ED providers and internal medicine (IM) physicians is typically oral (“synchronous communication”). It is not known how dictated signout to a voicemail system (“asynchronous communication”) affects the quality and safety of handoff communications.

Study design: Prospective, pre-post analysis.

Setting: A 944-bed urban academic medical center in Connecticut.

Synopsis: Surveys were administered to all IM and ED providers before and after the implementation of a voicemail signout system. In the new system, ED providers dictated signout for stable patients, rather than giving traditional synchronous telephone signout. It was the responsibility of the admitting IM physician to listen to the voicemail after receiving a text notification that a patient was being admitted.

ED providers recorded signouts in 89.5% of medicine admissions. However, voicemails were accessed only 58.5% of the time by receiving physicians. All ED providers and 56% of IM physicians believed signout was easier following the voicemail intervention. Overall, ED providers gave the quality, content, and accuracy of their signout communication higher ratings than IM physicians did; 69% of all providers felt the interaction among participants was worse following the intervention. There was no change in the rate of perceived adverse events or ICU transfers within 24 hours after admission.

This intervention was a QI initiative at a single center. Mixed results and small sample size limit generalizability of the study.

Bottom line: Asynchronous signout by voicemail increased efficiency, particularly among ED providers but decreased perceived quality of interaction between medical providers without obviously affecting patient safety.

Citation: Horwitz LI, Parwani V, Shah NR, et al. Evaluation of an asynchronous physician voicemail sign-out for emergency department admissions. Ann Emerg Med. 2009;54:368-378.

Patient Signout Is Not Uniformly Comprehensive and Often Lacks Critical Information

Clinical question: Do signouts vary in the quality and quantity of information, and what are the various factors affecting signout quality?

Background: Miscommunication during transfers of responsibility for hospitalized patients is common and can result in harm. Recommendations for safe and effective handoffs emphasize key content, clear communication, senior staff supervision, and adequate time for questions. Still, little is known about adherence to these recommendations in clinical practice.

Study design: Prospective, observational cohort.

Setting: Medical unit of an acute-care teaching hospital.

Synopsis: Oral signouts were audiotaped among IM house staff teams and the accompanying written signouts were collected for review of content. Signout sessions (n=88) included eight IM teams at one hospital and contained 503 patient signouts.

The median signout duration was 35 seconds (IQR 19-62) per patient. Key clinical information was present in just 62% of combined written or oral signouts. Most signouts included no questions from the recipient. Factors associated with higher rate of content inclusion included: familiarity with the patient, sense of responsibility (primary team vs. covering team), only one signout per day (as compared to sequential signout), presence of a senior resident, and comprehensive, written signouts.

Study limitations include the Hawthorne effect, as several participants mentioned that the presence of audiotape led to more comprehensive signouts than are typical. Also, the signout quality assessment in this study has not been validated with patient-safety outcomes.

Bottom line: Signouts among internal-medicine residents at this one hospital showed variability in terms of quantitative and qualitative information and often missed crucial information about patient care.

Citation: Horwitz LI, Moin T, Krumholz HM, Wang L, Bradley EH. What are covering doctors told about their patients? Analysis of sign-out among internal medicine house staff. Qual Saf Health Care. 2009;18(4):248-255.

Negative D-Dimer Test Can Safely Exclude Pulmonary Embolism in Patients at Low To Intermediate Clinical Risk

Clinical question: In patients with symptoms consistent with pulmonary embolism (PE), can evaluation with a clinical risk assessment tool and D-dimer assay identify patients who do not require CT angiography to exclude PE?

Background: D-dimer is a highly sensitive but nonspecific marker of VTE, and studies suggest that VTE can be ruled out without further imaging in patients with low clinical probability of disease and a negative D-dimer test. Nevertheless, this practice has not been adopted uniformly, and CT angiography (CTA) overuse continues.

Study design: Prospective registry cohort.

Setting: A 550-bed community teaching hospital in Chicago.

Synopsis: Consecutive patients presenting to the ED with symptoms suggestive of PE were evaluated with 1) revised Geneva score; 2) D-dimer assay; and 3) CTA. Among the 627 patients who underwent all three components of the evaluation, 44.8% were identified as low probability for PE by revised Geneva score, 52.6% as intermediate probability, and 2.6% as high probability. The overall prevalence of PE (using CTA as the gold standard) was very low (4.5%); just 2.1% of low-risk, 5.2% of intermediate-risk, and 31.2% of high-risk patients were ultimately found to have PE on CTA.

Using a cutoff of 1.2 mg/L, the D-dimer assay accurately detected all low- to intermediate-probability patients with PE (sensitivity and negative predictive value of 100%). One patient in the high probability group did have a PE, even though the patient had a D-dimer value <1.2 mg/L (sensitivity and NPV both 80%). Had diagnostic testing stopped after a negative D-dimer result in the low- to intermediate-probability patients, 172 CTAs (27%) would have been avoided.

Bottom line: In a low-prevalence cohort, no pulmonary emboli were identified by CTA in any patient with a low to intermediate clinical risk assessment and a negative quantitative D-dimer assay result.

Citation: Gupta RT, Kakarla RK, Kirshenbaum KJ, Tapson VF. D-dimers and efficacy of clinical risk estimation algorithms: sensitivity in evaluation of acute pulmonary embolism. AJR Am J Roentgenol. 2009;193(2):425-430.

Patient Participation in Medication Reconciliation at Discharge Helps Detect Prescribing Discrepancies

Clinical question: Does the inclusion of a medication adherence counseling session during a hospital discharge reconciliation process reduce discrepancies in the final medication regimen?

Background: Inadvertent medication prescribing errors are an important cause of preventable adverse drug events and commonly occur at transitions of care. Although medication reconciliation processes can identify errors, the best strategies for implementation remain unclear.

Study design: Prospective, observational cohort.

Setting: A 550-bed teaching hospital in the Netherlands.

Synopsis: Of 437 patients admitted to a pulmonary ward and screened for eligibility, 267 were included in the analysis. A pharmacy specialist reviewed all available community prescription records, inpatient documentation, and discharge medication lists in an effort to identify discrepancies. Potential errors were discussed with the prescriber. Then, the pharmacy specialist interviewed the patient and provided additional counseling. Any new discrepancies were discussed with the prescriber. All questions raised by the pharmacist were recorded, as were all subsequent prescriber interventions.

The primary outcome measure was the number of interventions made as a result of pharmacy review. A total of 940 questions were asked. At least one intervention was recorded for 87% of patients before counseling (mean 2.7 interventions/patient) and for 97% of patients after (mean 5.3 interventions/patient). Discrepancies were addressed for 63.7% of patients before counseling and 72.5% after. Pharmacotherapy was optimized for 67.2% of patients before counseling and 76.3% after.

Bottom line: Patient engagement in the medication reconciliation process incrementally improves the quality of the history and helps identify clinically meaningful discrepancies at the time of hospital discharge.

Citation: Karapinar-Carkit F, Borgsteede S, Zoer J, Smit HJ, Egberts AC, van den Bemt P. Effect of medication reconciliation with and without patient counseling on the number of pharmaceutical interventions among patients discharged from the hospital. Ann Pharmacother. 2009;43(6):1001-1010.

Computer-Based Reminders Have Small to Modest Effect on Care Processes

Clinical question: Do on-screen, computer-based clinical reminders improve adherence to target processes of care or clinical outcomes?

Background: Gaps between practice guidelines and routine care are caused, in part, by the inability of clinicians to access or recall information at the point of care. Although automated reminder systems offer the promise of “just in time” recommendations, studies of electronic reminders have demonstrated mixed results.

Study design: Literature review and meta-analysis.

Setting: Multiple databases and information repositories, including MEDLINE, EMBASE, and CINAHL.

Synopsis: The authors conducted a literature search to identify randomized and quasi-randomized controlled trials measuring the effect of computer-based reminders on process measures or clinical outcomes. To avoid statistical challenges inherent in unit-of-analysis errors, the authors reported median improvement in process adherence or median change in clinical endpoints.

Out of a pool of 2,036 citations, 28 studies detailing 32 comparative analyses were included. Across the 28 studies, reminders resulted in a median improvement in target process adherence of 4.2% (3.3% for prescribing behavior, 2.8% for test ordering). Eight comparisons reported dichotomous clinical endpoints and collectively showed a median absolute improvement of 2.5%.

The greatest contribution to measured treatment effects came from large academic centers with well-established electronic health records and robust informatics departments. No characteristics of the reminder system or the clinical context were associated with the magnitude of impact. A potential limitation in reporting median effects across studies is that all studies were given equal weight.

Bottom line: Electronic reminders appear to have a small, positive effect on clinician adherence to recommended processes, although it is uncertain what contextual or design features are responsible for the greatest treatment effect.

Citation: Shojania K, Jennings A, Mayhew A, Ramsay CR, Eccles MP, Grimshaw J. The effects of on-screen, point of care computer reminders on processes and outcomes of care. Cochrane Database Syst Rev. 2009(3):CD001096. TH

Pediatric HM Literature

Short Course of Oral Antibiotics Effective for Acute Osteomyelitis and Septic Arthritis in Children

By Mark Shen, MD

Reviewed by Pediatric Editor Mark Shen, MD, medical director of hospital medicine at Dell Children’s Medical Center, Austin, Texas.

Clinical question: Is a short course (less than four weeks) of antibiotics effective for the treatment of acute osteomyelitis and septic arthritis?

Background: The optimal duration of treatment for acute bone and joint infections in children has not been assessed adequately in prospectively designed trials. Historically, intravenous (IV) antibiotics in four- to six-week durations have been recommended, although the evidence for this practice is limited. There is widespread variation in both the route of administration (oral vs. IV) and duration of this treatment.

Study design: Prospective cohort study.

Setting: Two children’s hospitals in Australia.

Synopsis: Seventy children ages 17 and under who presented to two tertiary-care children’s hospitals with osteomyelitis or septic arthritis were enrolled. Primary surgical drainage was performed for patients with septic arthritis. Intravenous antibiotics were administered for at least three days, and until clinical symptoms improved and the C-reactive protein levels had stabilized. Patients then were transitioned to oral antibiotics and discharged to complete a minimum of three weeks of therapy.

Fifty-nine percent of patients were converted to oral antibiotics by day three, 86% by day five of therapy. Based on clinical and hematologic assessment, 83% of patients had oral antibiotics stopped at the three-week followup and remained well through the 12-month follow-up period.

This study essentially involved prospective data collection for a cohort of children receiving standardized care. Although the results suggest that a majority of children can be treated with a three-week course of oral antibiotics, the results would have been further strengthened by an explicit protocol with well-defined criteria for the oral to IV transition and cessation of antibiotic therapy. Additional limitations include pathogens and antibiotic choices that might not be applicable to North American populations.

Bottom line: After initial intravenous therapy, a three-week course of oral antibiotics can be effective for acute osteomyelitis and septic arthritis in children.

Citation: Jagodzinski NA, Kanwar R, Graham K, Bache CE. Prospective evaluation of a shortened regimen of treatment for acute osteomyelitis and septic arthritis in children. J Pediatr Orthop. 2009;29(5):518-525.

In This Edition

Literature at a Glance

A guide to this month’s studies

• Resident duty hours and ICU patient outcomes
• Effect of ER boarding on patient outcomes
• ED voicemail signout of patients
• Adequacy of patient signouts
• D-dimer use in patients with low suspicion of pulmonary embolism
• Patient involvement in medication reconciliation
• Effects of on-screen reminders on outcomes

Decreased ICU Duty Hours Does Not Affect Patient Mortality

Clinical question: Does the reduction in work hours for residents affect mortality in medical and surgical ICUs?

Background: A reduction in work hours for residents was enforced in July 2003. Several prior studies using administrative or claims data did not show an association of the reduced work hours for residents with mortality in teaching hospitals when compared with nonteaching hospitals.

Study design: Observational retrospective registry cohort.

Setting: Twelve academic, 12 community, and 16 nonteaching hospitals in the U.S.

Synopsis: Data from 230,151 patients were extracted as post-hoc analysis from a voluntary clinical registry that uses a well-validated severity-of-illness scoring system. The exposure was defined as date of admission to ICU within two years before and after the reform. Hospitals were categorized as academic, community with residents, or nonteaching. Sophisticated statistical analyses were performed, including interaction terms for teaching status and time. To test the effect the reduced work hours had on mortality, the mortality trends of academic hospitals and community hospitals with residents were compared with the baseline trend of nonteaching hospitals. After risk adjustments, all hospitals had improved in-hospital and ICU mortality after the reform. None of the statistical improvements were significantly different.

Study limitations include the selection bias, as only highly motivated hospitals participating in the registry were included, and misclassification bias, as not all hospitals implemented the reform at the same time. Nevertheless, this study supports the consistent literature on the topic and adds a more robust assessment of severity of illness.

Bottom line: The restriction on resident duty hours does not appear to affect patient mortality.

Citation: Prasad M, Iwashyna TJ, Christie JD, et al. Effect of work-hours regulations on intensive care unit mortality in United States teaching hospitals. Crit Care Med. 2009;37(9):2564-2569.

Emergency Department “Boarding” Results in Undesirable Events

Clinical question: What is the frequency and nature of undesirable events experienced by patients who “board” in the ED?

Background: Hospital crowding results in patients spending extended amounts of time—also known as “boarding”—in the ED as they wait for an inpatient bed. Prior studies have shown that longer ED boarding times are associated with adverse outcomes. Few studies have examined the nature and frequency of undesirable events that patients experience while boarding.

Study design: Retrospective chart review.

Setting: Urban academic medical center.

Synopsis: In this pilot study, authors reviewed the charts of patients who were treated in the ED and subsequently admitted to the hospital on three different days during the study period (n=151). More than a quarter (27.8%) of patients experienced an undesirable event, such as missing a scheduled medication, while they were boarding. Older patients, those with comorbid illnesses, and those who endured prolonged boarding times (greater than six hours) were more likely to experience an undesirable event. In addition, 3.3% of patients experienced such adverse events as suboptimal blood pressure control, hypotension, hypoxia, or arrhythmia.

This study was performed at a single center and lacks a comparison group (i.e., nonboarded patients). It is intended to serve as an exploratory study for future analysis of adverse events in boarded patients.

Bottom line: Undesirable events are common among boarded patients, although it is unknown whether they are more common than in nonboarded patients.

Citation: Liu SW, Thomas SH, Gordon JA, Hamedani AG, Weissman JS. A pilot study examining undesirable events among emergency-department boarded patients awaiting inpatient beds. Ann Emerg Med. 2009;54(3):381-385.

Emergency Department Signout via Voicemail Yields Mixed Reviews

Clinical question: How does traditional, oral signout from emergency providers to inpatient medicine physicians compare to dictated, voicemail signout?

Background: Communication failures contribute to errors in care transition from ED to inpatient medicine units. Signout between ED providers and internal medicine (IM) physicians is typically oral (“synchronous communication”). It is not known how dictated signout to a voicemail system (“asynchronous communication”) affects the quality and safety of handoff communications.

Study design: Prospective, pre-post analysis.

Setting: A 944-bed urban academic medical center in Connecticut.

Synopsis: Surveys were administered to all IM and ED providers before and after the implementation of a voicemail signout system. In the new system, ED providers dictated signout for stable patients, rather than giving traditional synchronous telephone signout. It was the responsibility of the admitting IM physician to listen to the voicemail after receiving a text notification that a patient was being admitted.

ED providers recorded signouts in 89.5% of medicine admissions. However, voicemails were accessed only 58.5% of the time by receiving physicians. All ED providers and 56% of IM physicians believed signout was easier following the voicemail intervention. Overall, ED providers gave the quality, content, and accuracy of their signout communication higher ratings than IM physicians did; 69% of all providers felt the interaction among participants was worse following the intervention. There was no change in the rate of perceived adverse events or ICU transfers within 24 hours after admission.

This intervention was a QI initiative at a single center. Mixed results and small sample size limit generalizability of the study.

Bottom line: Asynchronous signout by voicemail increased efficiency, particularly among ED providers but decreased perceived quality of interaction between medical providers without obviously affecting patient safety.

Citation: Horwitz LI, Parwani V, Shah NR, et al. Evaluation of an asynchronous physician voicemail sign-out for emergency department admissions. Ann Emerg Med. 2009;54:368-378.

Patient Signout Is Not Uniformly Comprehensive and Often Lacks Critical Information

Clinical question: Do signouts vary in the quality and quantity of information, and what are the various factors affecting signout quality?

Background: Miscommunication during transfers of responsibility for hospitalized patients is common and can result in harm. Recommendations for safe and effective handoffs emphasize key content, clear communication, senior staff supervision, and adequate time for questions. Still, little is known about adherence to these recommendations in clinical practice.

Study design: Prospective, observational cohort.

Setting: Medical unit of an acute-care teaching hospital.

Synopsis: Oral signouts were audiotaped among IM house staff teams and the accompanying written signouts were collected for review of content. Signout sessions (n=88) included eight IM teams at one hospital and contained 503 patient signouts.

The median signout duration was 35 seconds (IQR 19-62) per patient. Key clinical information was present in just 62% of combined written or oral signouts. Most signouts included no questions from the recipient. Factors associated with higher rate of content inclusion included: familiarity with the patient, sense of responsibility (primary team vs. covering team), only one signout per day (as compared to sequential signout), presence of a senior resident, and comprehensive, written signouts.

Study limitations include the Hawthorne effect, as several participants mentioned that the presence of audiotape led to more comprehensive signouts than are typical. Also, the signout quality assessment in this study has not been validated with patient-safety outcomes.

Bottom line: Signouts among internal-medicine residents at this one hospital showed variability in terms of quantitative and qualitative information and often missed crucial information about patient care.

Citation: Horwitz LI, Moin T, Krumholz HM, Wang L, Bradley EH. What are covering doctors told about their patients? Analysis of sign-out among internal medicine house staff. Qual Saf Health Care. 2009;18(4):248-255.

Negative D-Dimer Test Can Safely Exclude Pulmonary Embolism in Patients at Low To Intermediate Clinical Risk

Clinical question: In patients with symptoms consistent with pulmonary embolism (PE), can evaluation with a clinical risk assessment tool and D-dimer assay identify patients who do not require CT angiography to exclude PE?

Background: D-dimer is a highly sensitive but nonspecific marker of VTE, and studies suggest that VTE can be ruled out without further imaging in patients with low clinical probability of disease and a negative D-dimer test. Nevertheless, this practice has not been adopted uniformly, and CT angiography (CTA) overuse continues.

Study design: Prospective registry cohort.

Setting: A 550-bed community teaching hospital in Chicago.

Synopsis: Consecutive patients presenting to the ED with symptoms suggestive of PE were evaluated with 1) revised Geneva score; 2) D-dimer assay; and 3) CTA. Among the 627 patients who underwent all three components of the evaluation, 44.8% were identified as low probability for PE by revised Geneva score, 52.6% as intermediate probability, and 2.6% as high probability. The overall prevalence of PE (using CTA as the gold standard) was very low (4.5%); just 2.1% of low-risk, 5.2% of intermediate-risk, and 31.2% of high-risk patients were ultimately found to have PE on CTA.

Using a cutoff of 1.2 mg/L, the D-dimer assay accurately detected all low- to intermediate-probability patients with PE (sensitivity and negative predictive value of 100%). One patient in the high probability group did have a PE, even though the patient had a D-dimer value <1.2 mg/L (sensitivity and NPV both 80%). Had diagnostic testing stopped after a negative D-dimer result in the low- to intermediate-probability patients, 172 CTAs (27%) would have been avoided.

Bottom line: In a low-prevalence cohort, no pulmonary emboli were identified by CTA in any patient with a low to intermediate clinical risk assessment and a negative quantitative D-dimer assay result.

Citation: Gupta RT, Kakarla RK, Kirshenbaum KJ, Tapson VF. D-dimers and efficacy of clinical risk estimation algorithms: sensitivity in evaluation of acute pulmonary embolism. AJR Am J Roentgenol. 2009;193(2):425-430.

Patient Participation in Medication Reconciliation at Discharge Helps Detect Prescribing Discrepancies

Clinical question: Does the inclusion of a medication adherence counseling session during a hospital discharge reconciliation process reduce discrepancies in the final medication regimen?

Background: Inadvertent medication prescribing errors are an important cause of preventable adverse drug events and commonly occur at transitions of care. Although medication reconciliation processes can identify errors, the best strategies for implementation remain unclear.

Study design: Prospective, observational cohort.

Setting: A 550-bed teaching hospital in the Netherlands.

Synopsis: Of 437 patients admitted to a pulmonary ward and screened for eligibility, 267 were included in the analysis. A pharmacy specialist reviewed all available community prescription records, inpatient documentation, and discharge medication lists in an effort to identify discrepancies. Potential errors were discussed with the prescriber. Then, the pharmacy specialist interviewed the patient and provided additional counseling. Any new discrepancies were discussed with the prescriber. All questions raised by the pharmacist were recorded, as were all subsequent prescriber interventions.

The primary outcome measure was the number of interventions made as a result of pharmacy review. A total of 940 questions were asked. At least one intervention was recorded for 87% of patients before counseling (mean 2.7 interventions/patient) and for 97% of patients after (mean 5.3 interventions/patient). Discrepancies were addressed for 63.7% of patients before counseling and 72.5% after. Pharmacotherapy was optimized for 67.2% of patients before counseling and 76.3% after.

Bottom line: Patient engagement in the medication reconciliation process incrementally improves the quality of the history and helps identify clinically meaningful discrepancies at the time of hospital discharge.

Citation: Karapinar-Carkit F, Borgsteede S, Zoer J, Smit HJ, Egberts AC, van den Bemt P. Effect of medication reconciliation with and without patient counseling on the number of pharmaceutical interventions among patients discharged from the hospital. Ann Pharmacother. 2009;43(6):1001-1010.

Computer-Based Reminders Have Small to Modest Effect on Care Processes

Clinical question: Do on-screen, computer-based clinical reminders improve adherence to target processes of care or clinical outcomes?

Background: Gaps between practice guidelines and routine care are caused, in part, by the inability of clinicians to access or recall information at the point of care. Although automated reminder systems offer the promise of “just in time” recommendations, studies of electronic reminders have demonstrated mixed results.

Study design: Literature review and meta-analysis.

Setting: Multiple databases and information repositories, including MEDLINE, EMBASE, and CINAHL.

Synopsis: The authors conducted a literature search to identify randomized and quasi-randomized controlled trials measuring the effect of computer-based reminders on process measures or clinical outcomes. To avoid statistical challenges inherent in unit-of-analysis errors, the authors reported median improvement in process adherence or median change in clinical endpoints.

Out of a pool of 2,036 citations, 28 studies detailing 32 comparative analyses were included. Across the 28 studies, reminders resulted in a median improvement in target process adherence of 4.2% (3.3% for prescribing behavior, 2.8% for test ordering). Eight comparisons reported dichotomous clinical endpoints and collectively showed a median absolute improvement of 2.5%.

The greatest contribution to measured treatment effects came from large academic centers with well-established electronic health records and robust informatics departments. No characteristics of the reminder system or the clinical context were associated with the magnitude of impact. A potential limitation in reporting median effects across studies is that all studies were given equal weight.

Bottom line: Electronic reminders appear to have a small, positive effect on clinician adherence to recommended processes, although it is uncertain what contextual or design features are responsible for the greatest treatment effect.

Citation: Shojania K, Jennings A, Mayhew A, Ramsay CR, Eccles MP, Grimshaw J. The effects of on-screen, point of care computer reminders on processes and outcomes of care. Cochrane Database Syst Rev. 2009(3):CD001096. TH

Pediatric HM Literature

Short Course of Oral Antibiotics Effective for Acute Osteomyelitis and Septic Arthritis in Children

By Mark Shen, MD

Reviewed by Pediatric Editor Mark Shen, MD, medical director of hospital medicine at Dell Children’s Medical Center, Austin, Texas.

Clinical question: Is a short course (less than four weeks) of antibiotics effective for the treatment of acute osteomyelitis and septic arthritis?

Background: The optimal duration of treatment for acute bone and joint infections in children has not been assessed adequately in prospectively designed trials. Historically, intravenous (IV) antibiotics in four- to six-week durations have been recommended, although the evidence for this practice is limited. There is widespread variation in both the route of administration (oral vs. IV) and duration of this treatment.

Study design: Prospective cohort study.

Setting: Two children’s hospitals in Australia.

Synopsis: Seventy children ages 17 and under who presented to two tertiary-care children’s hospitals with osteomyelitis or septic arthritis were enrolled. Primary surgical drainage was performed for patients with septic arthritis. Intravenous antibiotics were administered for at least three days, and until clinical symptoms improved and the C-reactive protein levels had stabilized. Patients then were transitioned to oral antibiotics and discharged to complete a minimum of three weeks of therapy.

Fifty-nine percent of patients were converted to oral antibiotics by day three, 86% by day five of therapy. Based on clinical and hematologic assessment, 83% of patients had oral antibiotics stopped at the three-week followup and remained well through the 12-month follow-up period.

This study essentially involved prospective data collection for a cohort of children receiving standardized care. Although the results suggest that a majority of children can be treated with a three-week course of oral antibiotics, the results would have been further strengthened by an explicit protocol with well-defined criteria for the oral to IV transition and cessation of antibiotic therapy. Additional limitations include pathogens and antibiotic choices that might not be applicable to North American populations.

Bottom line: After initial intravenous therapy, a three-week course of oral antibiotics can be effective for acute osteomyelitis and septic arthritis in children.

Citation: Jagodzinski NA, Kanwar R, Graham K, Bache CE. Prospective evaluation of a shortened regimen of treatment for acute osteomyelitis and septic arthritis in children. J Pediatr Orthop. 2009;29(5):518-525.

In This Edition

Literature at a Glance

A guide to this month’s studies

• Resident duty hours and ICU patient outcomes
• Effect of ER boarding on patient outcomes
• ED voicemail signout of patients
• Adequacy of patient signouts
• D-dimer use in patients with low suspicion of pulmonary embolism
• Patient involvement in medication reconciliation
• Effects of on-screen reminders on outcomes

Decreased ICU Duty Hours Does Not Affect Patient Mortality

Clinical question: Does the reduction in work hours for residents affect mortality in medical and surgical ICUs?

Background: A reduction in work hours for residents was enforced in July 2003. Several prior studies using administrative or claims data did not show an association of the reduced work hours for residents with mortality in teaching hospitals when compared with nonteaching hospitals.

Study design: Observational retrospective registry cohort.

Setting: Twelve academic, 12 community, and 16 nonteaching hospitals in the U.S.

Synopsis: Data from 230,151 patients were extracted as post-hoc analysis from a voluntary clinical registry that uses a well-validated severity-of-illness scoring system. The exposure was defined as date of admission to ICU within two years before and after the reform. Hospitals were categorized as academic, community with residents, or nonteaching. Sophisticated statistical analyses were performed, including interaction terms for teaching status and time. To test the effect the reduced work hours had on mortality, the mortality trends of academic hospitals and community hospitals with residents were compared with the baseline trend of nonteaching hospitals. After risk adjustments, all hospitals had improved in-hospital and ICU mortality after the reform. None of the statistical improvements were significantly different.

Study limitations include the selection bias, as only highly motivated hospitals participating in the registry were included, and misclassification bias, as not all hospitals implemented the reform at the same time. Nevertheless, this study supports the consistent literature on the topic and adds a more robust assessment of severity of illness.

Bottom line: The restriction on resident duty hours does not appear to affect patient mortality.

Citation: Prasad M, Iwashyna TJ, Christie JD, et al. Effect of work-hours regulations on intensive care unit mortality in United States teaching hospitals. Crit Care Med. 2009;37(9):2564-2569.

Emergency Department “Boarding” Results in Undesirable Events

Clinical question: What is the frequency and nature of undesirable events experienced by patients who “board” in the ED?

Background: Hospital crowding results in patients spending extended amounts of time—also known as “boarding”—in the ED as they wait for an inpatient bed. Prior studies have shown that longer ED boarding times are associated with adverse outcomes. Few studies have examined the nature and frequency of undesirable events that patients experience while boarding.

Study design: Retrospective chart review.

Setting: Urban academic medical center.

Synopsis: In this pilot study, authors reviewed the charts of patients who were treated in the ED and subsequently admitted to the hospital on three different days during the study period (n=151). More than a quarter (27.8%) of patients experienced an undesirable event, such as missing a scheduled medication, while they were boarding. Older patients, those with comorbid illnesses, and those who endured prolonged boarding times (greater than six hours) were more likely to experience an undesirable event. In addition, 3.3% of patients experienced such adverse events as suboptimal blood pressure control, hypotension, hypoxia, or arrhythmia.

This study was performed at a single center and lacks a comparison group (i.e., nonboarded patients). It is intended to serve as an exploratory study for future analysis of adverse events in boarded patients.

Bottom line: Undesirable events are common among boarded patients, although it is unknown whether they are more common than in nonboarded patients.

Citation: Liu SW, Thomas SH, Gordon JA, Hamedani AG, Weissman JS. A pilot study examining undesirable events among emergency-department boarded patients awaiting inpatient beds. Ann Emerg Med. 2009;54(3):381-385.

Emergency Department Signout via Voicemail Yields Mixed Reviews

Clinical question: How does traditional, oral signout from emergency providers to inpatient medicine physicians compare to dictated, voicemail signout?

Background: Communication failures contribute to errors in care transition from ED to inpatient medicine units. Signout between ED providers and internal medicine (IM) physicians is typically oral (“synchronous communication”). It is not known how dictated signout to a voicemail system (“asynchronous communication”) affects the quality and safety of handoff communications.

Study design: Prospective, pre-post analysis.

Setting: A 944-bed urban academic medical center in Connecticut.

Synopsis: Surveys were administered to all IM and ED providers before and after the implementation of a voicemail signout system. In the new system, ED providers dictated signout for stable patients, rather than giving traditional synchronous telephone signout. It was the responsibility of the admitting IM physician to listen to the voicemail after receiving a text notification that a patient was being admitted.

ED providers recorded signouts in 89.5% of medicine admissions. However, voicemails were accessed only 58.5% of the time by receiving physicians. All ED providers and 56% of IM physicians believed signout was easier following the voicemail intervention. Overall, ED providers gave the quality, content, and accuracy of their signout communication higher ratings than IM physicians did; 69% of all providers felt the interaction among participants was worse following the intervention. There was no change in the rate of perceived adverse events or ICU transfers within 24 hours after admission.

This intervention was a QI initiative at a single center. Mixed results and small sample size limit generalizability of the study.

Bottom line: Asynchronous signout by voicemail increased efficiency, particularly among ED providers but decreased perceived quality of interaction between medical providers without obviously affecting patient safety.

Citation: Horwitz LI, Parwani V, Shah NR, et al. Evaluation of an asynchronous physician voicemail sign-out for emergency department admissions. Ann Emerg Med. 2009;54:368-378.

Patient Signout Is Not Uniformly Comprehensive and Often Lacks Critical Information

Clinical question: Do signouts vary in the quality and quantity of information, and what are the various factors affecting signout quality?

Background: Miscommunication during transfers of responsibility for hospitalized patients is common and can result in harm. Recommendations for safe and effective handoffs emphasize key content, clear communication, senior staff supervision, and adequate time for questions. Still, little is known about adherence to these recommendations in clinical practice.

Study design: Prospective, observational cohort.

Setting: Medical unit of an acute-care teaching hospital.

Synopsis: Oral signouts were audiotaped among IM house staff teams and the accompanying written signouts were collected for review of content. Signout sessions (n=88) included eight IM teams at one hospital and contained 503 patient signouts.

The median signout duration was 35 seconds (IQR 19-62) per patient. Key clinical information was present in just 62% of combined written or oral signouts. Most signouts included no questions from the recipient. Factors associated with higher rate of content inclusion included: familiarity with the patient, sense of responsibility (primary team vs. covering team), only one signout per day (as compared to sequential signout), presence of a senior resident, and comprehensive, written signouts.

Study limitations include the Hawthorne effect, as several participants mentioned that the presence of audiotape led to more comprehensive signouts than are typical. Also, the signout quality assessment in this study has not been validated with patient-safety outcomes.

Bottom line: Signouts among internal-medicine residents at this one hospital showed variability in terms of quantitative and qualitative information and often missed crucial information about patient care.

Citation: Horwitz LI, Moin T, Krumholz HM, Wang L, Bradley EH. What are covering doctors told about their patients? Analysis of sign-out among internal medicine house staff. Qual Saf Health Care. 2009;18(4):248-255.

Negative D-Dimer Test Can Safely Exclude Pulmonary Embolism in Patients at Low To Intermediate Clinical Risk

Clinical question: In patients with symptoms consistent with pulmonary embolism (PE), can evaluation with a clinical risk assessment tool and D-dimer assay identify patients who do not require CT angiography to exclude PE?

Background: D-dimer is a highly sensitive but nonspecific marker of VTE, and studies suggest that VTE can be ruled out without further imaging in patients with low clinical probability of disease and a negative D-dimer test. Nevertheless, this practice has not been adopted uniformly, and CT angiography (CTA) overuse continues.

Study design: Prospective registry cohort.

Setting: A 550-bed community teaching hospital in Chicago.

Synopsis: Consecutive patients presenting to the ED with symptoms suggestive of PE were evaluated with 1) revised Geneva score; 2) D-dimer assay; and 3) CTA. Among the 627 patients who underwent all three components of the evaluation, 44.8% were identified as low probability for PE by revised Geneva score, 52.6% as intermediate probability, and 2.6% as high probability. The overall prevalence of PE (using CTA as the gold standard) was very low (4.5%); just 2.1% of low-risk, 5.2% of intermediate-risk, and 31.2% of high-risk patients were ultimately found to have PE on CTA.

Using a cutoff of 1.2 mg/L, the D-dimer assay accurately detected all low- to intermediate-probability patients with PE (sensitivity and negative predictive value of 100%). One patient in the high probability group did have a PE, even though the patient had a D-dimer value <1.2 mg/L (sensitivity and NPV both 80%). Had diagnostic testing stopped after a negative D-dimer result in the low- to intermediate-probability patients, 172 CTAs (27%) would have been avoided.

Bottom line: In a low-prevalence cohort, no pulmonary emboli were identified by CTA in any patient with a low to intermediate clinical risk assessment and a negative quantitative D-dimer assay result.

Citation: Gupta RT, Kakarla RK, Kirshenbaum KJ, Tapson VF. D-dimers and efficacy of clinical risk estimation algorithms: sensitivity in evaluation of acute pulmonary embolism. AJR Am J Roentgenol. 2009;193(2):425-430.

Patient Participation in Medication Reconciliation at Discharge Helps Detect Prescribing Discrepancies

Clinical question: Does the inclusion of a medication adherence counseling session during a hospital discharge reconciliation process reduce discrepancies in the final medication regimen?

Background: Inadvertent medication prescribing errors are an important cause of preventable adverse drug events and commonly occur at transitions of care. Although medication reconciliation processes can identify errors, the best strategies for implementation remain unclear.

Study design: Prospective, observational cohort.

Setting: A 550-bed teaching hospital in the Netherlands.

Synopsis: Of 437 patients admitted to a pulmonary ward and screened for eligibility, 267 were included in the analysis. A pharmacy specialist reviewed all available community prescription records, inpatient documentation, and discharge medication lists in an effort to identify discrepancies. Potential errors were discussed with the prescriber. Then, the pharmacy specialist interviewed the patient and provided additional counseling. Any new discrepancies were discussed with the prescriber. All questions raised by the pharmacist were recorded, as were all subsequent prescriber interventions.

The primary outcome measure was the number of interventions made as a result of pharmacy review. A total of 940 questions were asked. At least one intervention was recorded for 87% of patients before counseling (mean 2.7 interventions/patient) and for 97% of patients after (mean 5.3 interventions/patient). Discrepancies were addressed for 63.7% of patients before counseling and 72.5% after. Pharmacotherapy was optimized for 67.2% of patients before counseling and 76.3% after.

Bottom line: Patient engagement in the medication reconciliation process incrementally improves the quality of the history and helps identify clinically meaningful discrepancies at the time of hospital discharge.

Citation: Karapinar-Carkit F, Borgsteede S, Zoer J, Smit HJ, Egberts AC, van den Bemt P. Effect of medication reconciliation with and without patient counseling on the number of pharmaceutical interventions among patients discharged from the hospital. Ann Pharmacother. 2009;43(6):1001-1010.

Computer-Based Reminders Have Small to Modest Effect on Care Processes

Clinical question: Do on-screen, computer-based clinical reminders improve adherence to target processes of care or clinical outcomes?

Background: Gaps between practice guidelines and routine care are caused, in part, by the inability of clinicians to access or recall information at the point of care. Although automated reminder systems offer the promise of “just in time” recommendations, studies of electronic reminders have demonstrated mixed results.

Study design: Literature review and meta-analysis.

Setting: Multiple databases and information repositories, including MEDLINE, EMBASE, and CINAHL.

Synopsis: The authors conducted a literature search to identify randomized and quasi-randomized controlled trials measuring the effect of computer-based reminders on process measures or clinical outcomes. To avoid statistical challenges inherent in unit-of-analysis errors, the authors reported median improvement in process adherence or median change in clinical endpoints.

Out of a pool of 2,036 citations, 28 studies detailing 32 comparative analyses were included. Across the 28 studies, reminders resulted in a median improvement in target process adherence of 4.2% (3.3% for prescribing behavior, 2.8% for test ordering). Eight comparisons reported dichotomous clinical endpoints and collectively showed a median absolute improvement of 2.5%.

The greatest contribution to measured treatment effects came from large academic centers with well-established electronic health records and robust informatics departments. No characteristics of the reminder system or the clinical context were associated with the magnitude of impact. A potential limitation in reporting median effects across studies is that all studies were given equal weight.

Bottom line: Electronic reminders appear to have a small, positive effect on clinician adherence to recommended processes, although it is uncertain what contextual or design features are responsible for the greatest treatment effect.

Citation: Shojania K, Jennings A, Mayhew A, Ramsay CR, Eccles MP, Grimshaw J. The effects of on-screen, point of care computer reminders on processes and outcomes of care. Cochrane Database Syst Rev. 2009(3):CD001096. TH

Pediatric HM Literature

Short Course of Oral Antibiotics Effective for Acute Osteomyelitis and Septic Arthritis in Children

By Mark Shen, MD

Reviewed by Pediatric Editor Mark Shen, MD, medical director of hospital medicine at Dell Children’s Medical Center, Austin, Texas.

Clinical question: Is a short course (less than four weeks) of antibiotics effective for the treatment of acute osteomyelitis and septic arthritis?

Background: The optimal duration of treatment for acute bone and joint infections in children has not been assessed adequately in prospectively designed trials. Historically, intravenous (IV) antibiotics in four- to six-week durations have been recommended, although the evidence for this practice is limited. There is widespread variation in both the route of administration (oral vs. IV) and duration of this treatment.

Study design: Prospective cohort study.

Setting: Two children’s hospitals in Australia.

Synopsis: Seventy children ages 17 and under who presented to two tertiary-care children’s hospitals with osteomyelitis or septic arthritis were enrolled. Primary surgical drainage was performed for patients with septic arthritis. Intravenous antibiotics were administered for at least three days, and until clinical symptoms improved and the C-reactive protein levels had stabilized. Patients then were transitioned to oral antibiotics and discharged to complete a minimum of three weeks of therapy.

Fifty-nine percent of patients were converted to oral antibiotics by day three, 86% by day five of therapy. Based on clinical and hematologic assessment, 83% of patients had oral antibiotics stopped at the three-week followup and remained well through the 12-month follow-up period.

This study essentially involved prospective data collection for a cohort of children receiving standardized care. Although the results suggest that a majority of children can be treated with a three-week course of oral antibiotics, the results would have been further strengthened by an explicit protocol with well-defined criteria for the oral to IV transition and cessation of antibiotic therapy. Additional limitations include pathogens and antibiotic choices that might not be applicable to North American populations.

Bottom line: After initial intravenous therapy, a three-week course of oral antibiotics can be effective for acute osteomyelitis and septic arthritis in children.

Citation: Jagodzinski NA, Kanwar R, Graham K, Bache CE. Prospective evaluation of a shortened regimen of treatment for acute osteomyelitis and septic arthritis in children. J Pediatr Orthop. 2009;29(5):518-525.