Clinical Care Conundrums

Clinical question: Does the use of 80% oxygen perioperatively in abdominal surgery decrease the frequency of surgical-site infection within 14 days without increasing the rate of pulmonary complications?

Background: Low oxygen tension in wounds can negatively impact immune response and healing. Increasing inspiratory oxygen fraction during the perioperative period translates into higher wound oxygen tension. However, the benefit of increased oxygen fraction therapy in abdominal surgery healing and complications is not clear, nor is the frequency of pulmonary complications.

Study design: Patient- and observer-blinded clinical trial.

Setting: Fourteen Danish hospitals from October 2006 to October 2008.

Synopsis: Patients were randomized to receive a fraction of inspired oxygen (FIO₂) of 0.80 or 0.30. The primary outcome—surgical-site infection in the superficial or deep wound or intra-abdominal cavity within 14 days of surgery—was defined using Centers for Disease Control and Prevention (CDC) criteria. Secondary outcomes included pulmonary complications within 14 days (pneumonia, atelectasis, or respiratory failure), 30-day mortality, duration of post-op course, ICU stay within 14 days post-op, and any abdominal operation within 14 days. The 1,386 patients were enrolled in the intention-to-treat analysis.

Infection occurred in 19.1% of patients given 0.80 FIO2 and in 20.1% of patients given 0.30 FIO2; odds ratio of 0.94 (95% CI 0.72 to 1.22; P=0.64). Numbers of pulmonary complications were not significantly different between the groups.

This trial included acute and nonacute laparotomies with followup for adverse outcomes. Study limitations included the inability to ensure that both groups received timely antibiotics and prevention for hypothermia. Of patients in the 30% FIO2 group, 7.3% required higher oxygen administration. Additionally, infection might have been underestimated in 11.3% of patients who were not followed up on between days 13 and 30.

Bottom line: High oxygen concentration administered during and after laparotomy did not lead to fewer surgical site infections, nor did it significantly increase the frequency of pulmonary complications or death.

Citation: Meyhoff CS, Wetterslev J, Jorgensen LN, et al. Effect of high perioperative oxygen fraction on surgical site infection and pulmonary complications after abdominal surgery: the PROXI randomized clinical trial. JAMA. 2009;302(14):1543-1550.

Clinical question: Does the use of 80% oxygen perioperatively in abdominal surgery decrease the frequency of surgical-site infection within 14 days without increasing the rate of pulmonary complications?

Background: Low oxygen tension in wounds can negatively impact immune response and healing. Increasing inspiratory oxygen fraction during the perioperative period translates into higher wound oxygen tension. However, the benefit of increased oxygen fraction therapy in abdominal surgery healing and complications is not clear, nor is the frequency of pulmonary complications.

Study design: Patient- and observer-blinded clinical trial.

Setting: Fourteen Danish hospitals from October 2006 to October 2008.

Synopsis: Patients were randomized to receive a fraction of inspired oxygen (FIO₂) of 0.80 or 0.30. The primary outcome—surgical-site infection in the superficial or deep wound or intra-abdominal cavity within 14 days of surgery—was defined using Centers for Disease Control and Prevention (CDC) criteria. Secondary outcomes included pulmonary complications within 14 days (pneumonia, atelectasis, or respiratory failure), 30-day mortality, duration of post-op course, ICU stay within 14 days post-op, and any abdominal operation within 14 days. The 1,386 patients were enrolled in the intention-to-treat analysis.

Infection occurred in 19.1% of patients given 0.80 FIO2 and in 20.1% of patients given 0.30 FIO2; odds ratio of 0.94 (95% CI 0.72 to 1.22; P=0.64). Numbers of pulmonary complications were not significantly different between the groups.

This trial included acute and nonacute laparotomies with followup for adverse outcomes. Study limitations included the inability to ensure that both groups received timely antibiotics and prevention for hypothermia. Of patients in the 30% FIO2 group, 7.3% required higher oxygen administration. Additionally, infection might have been underestimated in 11.3% of patients who were not followed up on between days 13 and 30.

Bottom line: High oxygen concentration administered during and after laparotomy did not lead to fewer surgical site infections, nor did it significantly increase the frequency of pulmonary complications or death.

Citation: Meyhoff CS, Wetterslev J, Jorgensen LN, et al. Effect of high perioperative oxygen fraction on surgical site infection and pulmonary complications after abdominal surgery: the PROXI randomized clinical trial. JAMA. 2009;302(14):1543-1550.

Clinical question: Does the use of 80% oxygen perioperatively in abdominal surgery decrease the frequency of surgical-site infection within 14 days without increasing the rate of pulmonary complications?

Background: Low oxygen tension in wounds can negatively impact immune response and healing. Increasing inspiratory oxygen fraction during the perioperative period translates into higher wound oxygen tension. However, the benefit of increased oxygen fraction therapy in abdominal surgery healing and complications is not clear, nor is the frequency of pulmonary complications.

Study design: Patient- and observer-blinded clinical trial.

Setting: Fourteen Danish hospitals from October 2006 to October 2008.

Synopsis: Patients were randomized to receive a fraction of inspired oxygen (FIO₂) of 0.80 or 0.30. The primary outcome—surgical-site infection in the superficial or deep wound or intra-abdominal cavity within 14 days of surgery—was defined using Centers for Disease Control and Prevention (CDC) criteria. Secondary outcomes included pulmonary complications within 14 days (pneumonia, atelectasis, or respiratory failure), 30-day mortality, duration of post-op course, ICU stay within 14 days post-op, and any abdominal operation within 14 days. The 1,386 patients were enrolled in the intention-to-treat analysis.

Infection occurred in 19.1% of patients given 0.80 FIO2 and in 20.1% of patients given 0.30 FIO2; odds ratio of 0.94 (95% CI 0.72 to 1.22; P=0.64). Numbers of pulmonary complications were not significantly different between the groups.

This trial included acute and nonacute laparotomies with followup for adverse outcomes. Study limitations included the inability to ensure that both groups received timely antibiotics and prevention for hypothermia. Of patients in the 30% FIO2 group, 7.3% required higher oxygen administration. Additionally, infection might have been underestimated in 11.3% of patients who were not followed up on between days 13 and 30.

Bottom line: High oxygen concentration administered during and after laparotomy did not lead to fewer surgical site infections, nor did it significantly increase the frequency of pulmonary complications or death.

Citation: Meyhoff CS, Wetterslev J, Jorgensen LN, et al. Effect of high perioperative oxygen fraction on surgical site infection and pulmonary complications after abdominal surgery: the PROXI randomized clinical trial. JAMA. 2009;302(14):1543-1550.

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.

Case

A 77-year-old woman with a history of stroke five months prior, bileaflet aortic valve prosthesis, hypertension, and insulin-dependent diabetes is admitted for laparoscopy with lysis of adhesions. The patient stopped her warfarin 10 days prior to admission and initiated enoxaparin five days later. When should the enoxaparin be discontinued?

Intra-operatively, the surgeon converted the case to an open laparotomy for a bowel resection with re-anastomosis; post-operatively, when should the hospitalist reinitiate warfarin and enoxaparin?

Background

Many patients receive chronic oral anticoagulant therapy to minimize their long-term risk of thromboembolic disease. Hospitalists and outpatient providers often care for such patients who need to undergo a medical procedure or operation. The risk of bleeding associated with the medical procedure necessitates an interruption in the patient’s chronic oral anticoagulant therapy. In this scenario, providers are faced with several therapeutic decisions:

• How soon before the procedure should patients stop taking oral anticoagulant?
• During the period of time when the patient is not taking chronic oral anticoagulant, should the patient receive parenteral bridging anticoagulant therapy?
• After the procedure, when should patients restart chronic oral anticoagulant therapy?

‘Bridge’ anticoagulant therapy is the administration of a short-acting parenteral anticoagulant during the peri-operative period, when the patient is not taking chronic oral anticoagulant.¹ The intent of bridge anticoagulant therapy is to minimize both the risk of thromboembolic events and the risk of bleeding during the peri-operative period. Bridging anticoagulant therapy is appropriate for some but not all patients undergoing medical procedures.

The Data

When to discontinue warfarin? Warfarin, the most commonly prescribed oral anticoagulant, achieves its therapeutic effects by antagonizing the actions of endogenous vitamin K-dependent coagulation factors. The decision on when to stop warfarin prior to surgery is dependent on the regeneration time of coagulation factors following the discontinuation of warfarin therapy. Although warfarin’s half-life is typically 36-42 hours, its therapeutic effects typically last up to five days in healthy subjects and often longer in elderly patients.²

Current guidelines recommend the discontinuation of warfarin at least five days prior to surgery (Grade 1C recommendation).³ Despite this recommendation, approximately 7% of patients will still have an international normalized ratio (INR) >1.5 after not taking warfarin for five days.⁴ For this reason, the guidelines recommend that all patients have their INR checked on the day of surgery. For those patients with an INR of 1.5 to 1.9 on the day prior to surgery, there is evidence to show that administration of 1 mg of vitamin K will lower the INR to 1.4 in greater than 90% of cases.⁵

Assessment of peri-procedural thrombotic risk. Knowledge of a patient’s past medical history is critical in helping providers stratify the patient’s peri-procedural thrombotic risk. According to the 2012 American College of Chest Physicians (ACCP) guidelines, a history of atrial fibrillation (Afib), mechanical heart valve(s), and previous VTE are independent risk factors for peri-procedural thrombotic events.³ Hospitalists may risk-stratify their patients based on the anticipated annualized rate of thrombosis or embolization: <5%, 5%-10%, or >15% for the respective low, medium, and high-risk groups.⁶

Patients with Afib history. For these patients, the CHADS2 score helps to stratify the risk of peri-procedural thrombosis. Low risk is defined as a CHADS2 score of zero to two, assuming that the two points were not scored for transient ischemic attack (TIA) or cerebrovascular accident (CVA). Any patient with a TIA or CVA within the previous three months is automatically considered high risk. Medium risk is a score of three or four.

In addition to the aforementioned TIA or CVA within the prior three months, high-risk patients also include those with a CHADS2 score of five or six or any patient with a history of rheumatic heart disease.³ Patients with CHADS2 scores less than five but with a TIA or CVA greater than three months in the past are high risk.⁷

Presence of mechanical heart valve(s). For patients with a mechanical heart valve, knowledge of the valve type and location is essential to assist hospitalists in stratifying the risk of peri-procedural thrombosis. The current ACCP guidelines consider patients with bileaflet aortic valve prostheses without additional risk factors for stroke or atrial fibrillation to be low risk.³

The guidelines define the following characteristics as medium risk for patients: the presence of a bileaflet valve with additional risk factors for stroke such as atrial fibrillation, age greater than 75, prior CVA (more than six months prior), hypertension, diabetes mellitus, or congestive heart failure.

Patients at high risk include those with aortic valve prosthesis with a caged-ball or tilting disc, patients with mitral valve prosthesis, and those with a mechanical valve with CVA or TIA during the prior six months.⁷

History of previous VTE. For these patients, the duration of time that has passed since their last VTE event is an important factor in helping to stratify their risk for peri-procedural thrombosis. Hospitalists should consider patients low risk if they had VTE more than one year prior to the procedure.

Medium-risk patients are those with VTE events in the preceding three to twelve months, those with recurrent VTE, those with active cancer who have received cancer therapy within six months, or patients with non-severe thrombophilias (e.g. heterogenous factor V Leiden or prothrombin gene mutation).

Hospitalists should identify high-risk patients as those with VTE that has occurred within three months or those with severe thrombophilias such as Protein C or S deficiency, antithrombin III deficiency, or antiphospholipid antibody syndrome.

Assessment of procedure-related thrombotic risk. The type of anticipated procedure itself conveys peri-procedural thrombotic risk. For example, heart valve replacement, carotid endarterectomy, or other major vascular surgeries automatically stratify patients in the high-risk category, regardless of underlying medication condition.

Assessment of bleeding risk. Hospitalists must identify any preexisting bleeding risk factors (i.e., hemophilias or thrombocytopenia) in addition to the post-procedural bleeding risks. Risk factors for increased post-procedural bleeding include: major surgery with extensive tissue injury, procedures involving highly vascularized organs, removal of large colonic polyps, urological procedures, placement of implantable cardioverter-defibrillator/pacemakers, and procedures at sites where minor bleeding would be clinically devastating, such as the brain or spine.³

Thus, communication with the proceduralist or surgeon regarding the anticipated bleeding risk is vital.

Should the patient receive bridging anticoagulation? Patients considered high risk for peri-procedural thrombosis should receive peri-procedural bridging anticoagulation therapy, while those considered low risk should not. For patients with a moderate peri-procedural risk of thrombosis, hospitalists should base the decision on individual and anticipated pre-surgical/procedural thrombotic risks.

Recent evidence suggests that bridging anticoagulation should be avoided in patients undergoing procedures with high bleeding risk who are not at high thromboembolic risk.⁸

Selection and pre-operative discontinuation of bridging medication. Current ACCP guidelines only support the use of unfractionated heparin (UFH) or low molecular weight heparin (LMWH) as bridging anticoagulants.³ Evidence supports the use of either intravenous UFH (goal aPTT 1.5 to two times control aPTT) or enoxaparin (1 mg/kg BID or 1.5 mg/kg once daily).⁹ UFH is preferred over LMWH in patients with chronic kidney disease stage IV or V due to a more predictable pharmacokinetic profile.

Clinicians should initiate a bridge when a patient’s INR falls to less than 2.0 and discontinue the UFH bridge four to six hours prior to the procedure.¹⁰ The recent update to the guidelines now states that LMWH should be discontinued 24, instead of 12, hours prior to the procedure.³

When to restart UHF or LMWH bridge post-procedure. The type of procedure being performed dictates when bridging anticoagulation should resume. In patients who have undergone surgeries that involve high bleeding risk, LMWH should not be administered until 48-72 hours post-surgery (Grade 2C evidence).³ For those patients undergoing surgeries with low bleeding risk, bridging should be resumed approximately 24 hours after the procedure.

Of note, enoxaparin administered in one single daily dose, as compared to divided doses, is associated with a greater risk of post-operative bleeding. UFH bridging should resume post-operatively without a bolus dose at 24 hours in low-risk bleeding cases or 48-72 hours in high-risk bleeding cases (Grade 2C evidence).³

On occasion, unanticipated adjustments to surgical cases—or complications—change the previously determined post-operative bleeding risk. In these instances, the hospitalist and surgeon/proceduralist should review the case and reassess the bleeding risk prior to employing bridging anticoagulation protocols.

When to restart long-term vitamin K antagonists (VKA) post-procedure. In most instances, regardless of pre-operative bleeding risk stratification, the resumption of VKA may occur once post-operative hemostasis has been achieved and the patient has been instructed to resume eating by the proceduralist or surgeon. This most often occurs on the calendar day following surgery, because it takes approximately five days for an INR to achieve therapeutic levels.

Back to the Case

The patient’s history of prosthetic valve with stroke within the preceding six months stratified her to a high thrombotic risk category. Given the high risk of thrombosis, the decision was made to bridge with LMWH. The hospitalist discontinued LMWH 24 hours prior to surgery, and INR was checked on the morning of the procedure.

Although the patient underwent the operation without significant bleeding, the adjustment from an exploratory laparoscopy to an open laparotomy increased her post-operative bleeding risk from medium to high. Therefore, bridging anticoagulation with LMWH was resumed no sooner than 48 hours after the operation. Her warfarin was restarted on the day following surgery, once she resumed her diet.

Bottom Line

Hospitalists must understand both the pre- and post-procedure thrombotic risks, as well as the pre- and post-procedural bleeding risks, when determining the selection and logistics of initiation and cessation of antithrombotic bridging for inpatients.

Drs. McCormick, Carbo, and Li are hospitalists at Beth Israel Deaconess Medical Center in Boston. Dr. Kerbel is a hospitalist at the University of California Los Angeles.

References

1. BRIDGE Study Investigators. Bridging anticoagulation: is it needed when warfarin is interrupted around the time of a surgery or procedure? Circulation. 2012;125(12):e496-498.
2. Ansell J, Hirsh J, Hylek E, Jacobson A, Crowther M, Palareti G. Pharmacology and management of the vitamin K antagonists: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest. 2008;133(6 Suppl):160S-198S.
3. Douketis JD, Spyropoulos AC, Spencer FA, et al. Perioperative management of antithrombotic therapy: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest. 2012;141(2 Suppl):e326S-350S.
4. Kovacs MJ, Kearon C, Rodger M, et al. Single-arm study of bridging therapy with low-molecular-weight heparin for patients at risk of arterial embolism who require temporary interruption of warfarin. Circulation. 2004;110(12):1658-1663.
5. Woods K, Douketis JD, Kathirgamanathan K, Yi Q, Crowther MA. Low-dose oral vitamin K to normalize the international normalized ratio prior to surgery in patients who require temporary interruption of warfarin. J Thromb Thrombolysis. 2007;24(2):93-97.
6. Ortel TL. Perioperative management of patients on chronic antithrombotic therapy. Blood. 2012;120(24):4699-4705.
7. Kaatz S, Douketis JD, Zhou H, Gage BF, White RH. Risk of stroke after surgery in patients with and without chronic atrial fibrillation. J Thromb Haemost. 2010;8(5):884-890.
8. Siegal D, Yudin J, Kaatz S, Douketis JD, Lim W, Spyropoulos AC. Periprocedural heparin bridging in patients receiving vitamin K antagonists: systematic review and meta-analysis of bleeding and thromboembolic rates. Circulation. 2012;126(13):1630-1639.
9. Lee AY, Levine MN, Baker RI, et al. Low-molecular-weight heparin versus a coumarin for the prevention of recurrent venous thromboembolism in patients with cancer. N Engl J Med. 2003;349(2):146-153.
10. Hirsh J, Raschke R. Heparin and low-molecular-weight heparin: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest. 2004;126(3 Suppl):188S-203S.

Case

A 77-year-old woman with a history of stroke five months prior, bileaflet aortic valve prosthesis, hypertension, and insulin-dependent diabetes is admitted for laparoscopy with lysis of adhesions. The patient stopped her warfarin 10 days prior to admission and initiated enoxaparin five days later. When should the enoxaparin be discontinued?

Intra-operatively, the surgeon converted the case to an open laparotomy for a bowel resection with re-anastomosis; post-operatively, when should the hospitalist reinitiate warfarin and enoxaparin?

Background

Many patients receive chronic oral anticoagulant therapy to minimize their long-term risk of thromboembolic disease. Hospitalists and outpatient providers often care for such patients who need to undergo a medical procedure or operation. The risk of bleeding associated with the medical procedure necessitates an interruption in the patient’s chronic oral anticoagulant therapy. In this scenario, providers are faced with several therapeutic decisions:

• How soon before the procedure should patients stop taking oral anticoagulant?
• During the period of time when the patient is not taking chronic oral anticoagulant, should the patient receive parenteral bridging anticoagulant therapy?
• After the procedure, when should patients restart chronic oral anticoagulant therapy?

‘Bridge’ anticoagulant therapy is the administration of a short-acting parenteral anticoagulant during the peri-operative period, when the patient is not taking chronic oral anticoagulant.¹ The intent of bridge anticoagulant therapy is to minimize both the risk of thromboembolic events and the risk of bleeding during the peri-operative period. Bridging anticoagulant therapy is appropriate for some but not all patients undergoing medical procedures.

The Data

When to discontinue warfarin? Warfarin, the most commonly prescribed oral anticoagulant, achieves its therapeutic effects by antagonizing the actions of endogenous vitamin K-dependent coagulation factors. The decision on when to stop warfarin prior to surgery is dependent on the regeneration time of coagulation factors following the discontinuation of warfarin therapy. Although warfarin’s half-life is typically 36-42 hours, its therapeutic effects typically last up to five days in healthy subjects and often longer in elderly patients.²

Current guidelines recommend the discontinuation of warfarin at least five days prior to surgery (Grade 1C recommendation).³ Despite this recommendation, approximately 7% of patients will still have an international normalized ratio (INR) >1.5 after not taking warfarin for five days.⁴ For this reason, the guidelines recommend that all patients have their INR checked on the day of surgery. For those patients with an INR of 1.5 to 1.9 on the day prior to surgery, there is evidence to show that administration of 1 mg of vitamin K will lower the INR to 1.4 in greater than 90% of cases.⁵

Assessment of peri-procedural thrombotic risk. Knowledge of a patient’s past medical history is critical in helping providers stratify the patient’s peri-procedural thrombotic risk. According to the 2012 American College of Chest Physicians (ACCP) guidelines, a history of atrial fibrillation (Afib), mechanical heart valve(s), and previous VTE are independent risk factors for peri-procedural thrombotic events.³ Hospitalists may risk-stratify their patients based on the anticipated annualized rate of thrombosis or embolization: <5%, 5%-10%, or >15% for the respective low, medium, and high-risk groups.⁶

Patients with Afib history. For these patients, the CHADS2 score helps to stratify the risk of peri-procedural thrombosis. Low risk is defined as a CHADS2 score of zero to two, assuming that the two points were not scored for transient ischemic attack (TIA) or cerebrovascular accident (CVA). Any patient with a TIA or CVA within the previous three months is automatically considered high risk. Medium risk is a score of three or four.

In addition to the aforementioned TIA or CVA within the prior three months, high-risk patients also include those with a CHADS2 score of five or six or any patient with a history of rheumatic heart disease.³ Patients with CHADS2 scores less than five but with a TIA or CVA greater than three months in the past are high risk.⁷

Presence of mechanical heart valve(s). For patients with a mechanical heart valve, knowledge of the valve type and location is essential to assist hospitalists in stratifying the risk of peri-procedural thrombosis. The current ACCP guidelines consider patients with bileaflet aortic valve prostheses without additional risk factors for stroke or atrial fibrillation to be low risk.³

The guidelines define the following characteristics as medium risk for patients: the presence of a bileaflet valve with additional risk factors for stroke such as atrial fibrillation, age greater than 75, prior CVA (more than six months prior), hypertension, diabetes mellitus, or congestive heart failure.

Patients at high risk include those with aortic valve prosthesis with a caged-ball or tilting disc, patients with mitral valve prosthesis, and those with a mechanical valve with CVA or TIA during the prior six months.⁷

History of previous VTE. For these patients, the duration of time that has passed since their last VTE event is an important factor in helping to stratify their risk for peri-procedural thrombosis. Hospitalists should consider patients low risk if they had VTE more than one year prior to the procedure.

Medium-risk patients are those with VTE events in the preceding three to twelve months, those with recurrent VTE, those with active cancer who have received cancer therapy within six months, or patients with non-severe thrombophilias (e.g. heterogenous factor V Leiden or prothrombin gene mutation).

Hospitalists should identify high-risk patients as those with VTE that has occurred within three months or those with severe thrombophilias such as Protein C or S deficiency, antithrombin III deficiency, or antiphospholipid antibody syndrome.

Assessment of procedure-related thrombotic risk. The type of anticipated procedure itself conveys peri-procedural thrombotic risk. For example, heart valve replacement, carotid endarterectomy, or other major vascular surgeries automatically stratify patients in the high-risk category, regardless of underlying medication condition.

Assessment of bleeding risk. Hospitalists must identify any preexisting bleeding risk factors (i.e., hemophilias or thrombocytopenia) in addition to the post-procedural bleeding risks. Risk factors for increased post-procedural bleeding include: major surgery with extensive tissue injury, procedures involving highly vascularized organs, removal of large colonic polyps, urological procedures, placement of implantable cardioverter-defibrillator/pacemakers, and procedures at sites where minor bleeding would be clinically devastating, such as the brain or spine.³

Thus, communication with the proceduralist or surgeon regarding the anticipated bleeding risk is vital.

Should the patient receive bridging anticoagulation? Patients considered high risk for peri-procedural thrombosis should receive peri-procedural bridging anticoagulation therapy, while those considered low risk should not. For patients with a moderate peri-procedural risk of thrombosis, hospitalists should base the decision on individual and anticipated pre-surgical/procedural thrombotic risks.

Recent evidence suggests that bridging anticoagulation should be avoided in patients undergoing procedures with high bleeding risk who are not at high thromboembolic risk.⁸

Selection and pre-operative discontinuation of bridging medication. Current ACCP guidelines only support the use of unfractionated heparin (UFH) or low molecular weight heparin (LMWH) as bridging anticoagulants.³ Evidence supports the use of either intravenous UFH (goal aPTT 1.5 to two times control aPTT) or enoxaparin (1 mg/kg BID or 1.5 mg/kg once daily).⁹ UFH is preferred over LMWH in patients with chronic kidney disease stage IV or V due to a more predictable pharmacokinetic profile.

Clinicians should initiate a bridge when a patient’s INR falls to less than 2.0 and discontinue the UFH bridge four to six hours prior to the procedure.¹⁰ The recent update to the guidelines now states that LMWH should be discontinued 24, instead of 12, hours prior to the procedure.³

When to restart UHF or LMWH bridge post-procedure. The type of procedure being performed dictates when bridging anticoagulation should resume. In patients who have undergone surgeries that involve high bleeding risk, LMWH should not be administered until 48-72 hours post-surgery (Grade 2C evidence).³ For those patients undergoing surgeries with low bleeding risk, bridging should be resumed approximately 24 hours after the procedure.

Of note, enoxaparin administered in one single daily dose, as compared to divided doses, is associated with a greater risk of post-operative bleeding. UFH bridging should resume post-operatively without a bolus dose at 24 hours in low-risk bleeding cases or 48-72 hours in high-risk bleeding cases (Grade 2C evidence).³

On occasion, unanticipated adjustments to surgical cases—or complications—change the previously determined post-operative bleeding risk. In these instances, the hospitalist and surgeon/proceduralist should review the case and reassess the bleeding risk prior to employing bridging anticoagulation protocols.

When to restart long-term vitamin K antagonists (VKA) post-procedure. In most instances, regardless of pre-operative bleeding risk stratification, the resumption of VKA may occur once post-operative hemostasis has been achieved and the patient has been instructed to resume eating by the proceduralist or surgeon. This most often occurs on the calendar day following surgery, because it takes approximately five days for an INR to achieve therapeutic levels.

Back to the Case

The patient’s history of prosthetic valve with stroke within the preceding six months stratified her to a high thrombotic risk category. Given the high risk of thrombosis, the decision was made to bridge with LMWH. The hospitalist discontinued LMWH 24 hours prior to surgery, and INR was checked on the morning of the procedure.

Although the patient underwent the operation without significant bleeding, the adjustment from an exploratory laparoscopy to an open laparotomy increased her post-operative bleeding risk from medium to high. Therefore, bridging anticoagulation with LMWH was resumed no sooner than 48 hours after the operation. Her warfarin was restarted on the day following surgery, once she resumed her diet.

Bottom Line

Hospitalists must understand both the pre- and post-procedure thrombotic risks, as well as the pre- and post-procedural bleeding risks, when determining the selection and logistics of initiation and cessation of antithrombotic bridging for inpatients.

Drs. McCormick, Carbo, and Li are hospitalists at Beth Israel Deaconess Medical Center in Boston. Dr. Kerbel is a hospitalist at the University of California Los Angeles.

References

1. BRIDGE Study Investigators. Bridging anticoagulation: is it needed when warfarin is interrupted around the time of a surgery or procedure? Circulation. 2012;125(12):e496-498.
2. Ansell J, Hirsh J, Hylek E, Jacobson A, Crowther M, Palareti G. Pharmacology and management of the vitamin K antagonists: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest. 2008;133(6 Suppl):160S-198S.
3. Douketis JD, Spyropoulos AC, Spencer FA, et al. Perioperative management of antithrombotic therapy: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest. 2012;141(2 Suppl):e326S-350S.
4. Kovacs MJ, Kearon C, Rodger M, et al. Single-arm study of bridging therapy with low-molecular-weight heparin for patients at risk of arterial embolism who require temporary interruption of warfarin. Circulation. 2004;110(12):1658-1663.
5. Woods K, Douketis JD, Kathirgamanathan K, Yi Q, Crowther MA. Low-dose oral vitamin K to normalize the international normalized ratio prior to surgery in patients who require temporary interruption of warfarin. J Thromb Thrombolysis. 2007;24(2):93-97.
6. Ortel TL. Perioperative management of patients on chronic antithrombotic therapy. Blood. 2012;120(24):4699-4705.
7. Kaatz S, Douketis JD, Zhou H, Gage BF, White RH. Risk of stroke after surgery in patients with and without chronic atrial fibrillation. J Thromb Haemost. 2010;8(5):884-890.
8. Siegal D, Yudin J, Kaatz S, Douketis JD, Lim W, Spyropoulos AC. Periprocedural heparin bridging in patients receiving vitamin K antagonists: systematic review and meta-analysis of bleeding and thromboembolic rates. Circulation. 2012;126(13):1630-1639.
9. Lee AY, Levine MN, Baker RI, et al. Low-molecular-weight heparin versus a coumarin for the prevention of recurrent venous thromboembolism in patients with cancer. N Engl J Med. 2003;349(2):146-153.
10. Hirsh J, Raschke R. Heparin and low-molecular-weight heparin: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest. 2004;126(3 Suppl):188S-203S.

Case

A 77-year-old woman with a history of stroke five months prior, bileaflet aortic valve prosthesis, hypertension, and insulin-dependent diabetes is admitted for laparoscopy with lysis of adhesions. The patient stopped her warfarin 10 days prior to admission and initiated enoxaparin five days later. When should the enoxaparin be discontinued?

Intra-operatively, the surgeon converted the case to an open laparotomy for a bowel resection with re-anastomosis; post-operatively, when should the hospitalist reinitiate warfarin and enoxaparin?

Background

Many patients receive chronic oral anticoagulant therapy to minimize their long-term risk of thromboembolic disease. Hospitalists and outpatient providers often care for such patients who need to undergo a medical procedure or operation. The risk of bleeding associated with the medical procedure necessitates an interruption in the patient’s chronic oral anticoagulant therapy. In this scenario, providers are faced with several therapeutic decisions:

• How soon before the procedure should patients stop taking oral anticoagulant?
• During the period of time when the patient is not taking chronic oral anticoagulant, should the patient receive parenteral bridging anticoagulant therapy?
• After the procedure, when should patients restart chronic oral anticoagulant therapy?

‘Bridge’ anticoagulant therapy is the administration of a short-acting parenteral anticoagulant during the peri-operative period, when the patient is not taking chronic oral anticoagulant.¹ The intent of bridge anticoagulant therapy is to minimize both the risk of thromboembolic events and the risk of bleeding during the peri-operative period. Bridging anticoagulant therapy is appropriate for some but not all patients undergoing medical procedures.

The Data

When to discontinue warfarin? Warfarin, the most commonly prescribed oral anticoagulant, achieves its therapeutic effects by antagonizing the actions of endogenous vitamin K-dependent coagulation factors. The decision on when to stop warfarin prior to surgery is dependent on the regeneration time of coagulation factors following the discontinuation of warfarin therapy. Although warfarin’s half-life is typically 36-42 hours, its therapeutic effects typically last up to five days in healthy subjects and often longer in elderly patients.²

Current guidelines recommend the discontinuation of warfarin at least five days prior to surgery (Grade 1C recommendation).³ Despite this recommendation, approximately 7% of patients will still have an international normalized ratio (INR) >1.5 after not taking warfarin for five days.⁴ For this reason, the guidelines recommend that all patients have their INR checked on the day of surgery. For those patients with an INR of 1.5 to 1.9 on the day prior to surgery, there is evidence to show that administration of 1 mg of vitamin K will lower the INR to 1.4 in greater than 90% of cases.⁵

Assessment of peri-procedural thrombotic risk. Knowledge of a patient’s past medical history is critical in helping providers stratify the patient’s peri-procedural thrombotic risk. According to the 2012 American College of Chest Physicians (ACCP) guidelines, a history of atrial fibrillation (Afib), mechanical heart valve(s), and previous VTE are independent risk factors for peri-procedural thrombotic events.³ Hospitalists may risk-stratify their patients based on the anticipated annualized rate of thrombosis or embolization: <5%, 5%-10%, or >15% for the respective low, medium, and high-risk groups.⁶

Patients with Afib history. For these patients, the CHADS2 score helps to stratify the risk of peri-procedural thrombosis. Low risk is defined as a CHADS2 score of zero to two, assuming that the two points were not scored for transient ischemic attack (TIA) or cerebrovascular accident (CVA). Any patient with a TIA or CVA within the previous three months is automatically considered high risk. Medium risk is a score of three or four.

In addition to the aforementioned TIA or CVA within the prior three months, high-risk patients also include those with a CHADS2 score of five or six or any patient with a history of rheumatic heart disease.³ Patients with CHADS2 scores less than five but with a TIA or CVA greater than three months in the past are high risk.⁷

Presence of mechanical heart valve(s). For patients with a mechanical heart valve, knowledge of the valve type and location is essential to assist hospitalists in stratifying the risk of peri-procedural thrombosis. The current ACCP guidelines consider patients with bileaflet aortic valve prostheses without additional risk factors for stroke or atrial fibrillation to be low risk.³

The guidelines define the following characteristics as medium risk for patients: the presence of a bileaflet valve with additional risk factors for stroke such as atrial fibrillation, age greater than 75, prior CVA (more than six months prior), hypertension, diabetes mellitus, or congestive heart failure.

Patients at high risk include those with aortic valve prosthesis with a caged-ball or tilting disc, patients with mitral valve prosthesis, and those with a mechanical valve with CVA or TIA during the prior six months.⁷

History of previous VTE. For these patients, the duration of time that has passed since their last VTE event is an important factor in helping to stratify their risk for peri-procedural thrombosis. Hospitalists should consider patients low risk if they had VTE more than one year prior to the procedure.

Medium-risk patients are those with VTE events in the preceding three to twelve months, those with recurrent VTE, those with active cancer who have received cancer therapy within six months, or patients with non-severe thrombophilias (e.g. heterogenous factor V Leiden or prothrombin gene mutation).

Hospitalists should identify high-risk patients as those with VTE that has occurred within three months or those with severe thrombophilias such as Protein C or S deficiency, antithrombin III deficiency, or antiphospholipid antibody syndrome.

Assessment of procedure-related thrombotic risk. The type of anticipated procedure itself conveys peri-procedural thrombotic risk. For example, heart valve replacement, carotid endarterectomy, or other major vascular surgeries automatically stratify patients in the high-risk category, regardless of underlying medication condition.

Assessment of bleeding risk. Hospitalists must identify any preexisting bleeding risk factors (i.e., hemophilias or thrombocytopenia) in addition to the post-procedural bleeding risks. Risk factors for increased post-procedural bleeding include: major surgery with extensive tissue injury, procedures involving highly vascularized organs, removal of large colonic polyps, urological procedures, placement of implantable cardioverter-defibrillator/pacemakers, and procedures at sites where minor bleeding would be clinically devastating, such as the brain or spine.³

Thus, communication with the proceduralist or surgeon regarding the anticipated bleeding risk is vital.

Should the patient receive bridging anticoagulation? Patients considered high risk for peri-procedural thrombosis should receive peri-procedural bridging anticoagulation therapy, while those considered low risk should not. For patients with a moderate peri-procedural risk of thrombosis, hospitalists should base the decision on individual and anticipated pre-surgical/procedural thrombotic risks.

Recent evidence suggests that bridging anticoagulation should be avoided in patients undergoing procedures with high bleeding risk who are not at high thromboembolic risk.⁸

Selection and pre-operative discontinuation of bridging medication. Current ACCP guidelines only support the use of unfractionated heparin (UFH) or low molecular weight heparin (LMWH) as bridging anticoagulants.³ Evidence supports the use of either intravenous UFH (goal aPTT 1.5 to two times control aPTT) or enoxaparin (1 mg/kg BID or 1.5 mg/kg once daily).⁹ UFH is preferred over LMWH in patients with chronic kidney disease stage IV or V due to a more predictable pharmacokinetic profile.

Clinicians should initiate a bridge when a patient’s INR falls to less than 2.0 and discontinue the UFH bridge four to six hours prior to the procedure.¹⁰ The recent update to the guidelines now states that LMWH should be discontinued 24, instead of 12, hours prior to the procedure.³

When to restart UHF or LMWH bridge post-procedure. The type of procedure being performed dictates when bridging anticoagulation should resume. In patients who have undergone surgeries that involve high bleeding risk, LMWH should not be administered until 48-72 hours post-surgery (Grade 2C evidence).³ For those patients undergoing surgeries with low bleeding risk, bridging should be resumed approximately 24 hours after the procedure.

Of note, enoxaparin administered in one single daily dose, as compared to divided doses, is associated with a greater risk of post-operative bleeding. UFH bridging should resume post-operatively without a bolus dose at 24 hours in low-risk bleeding cases or 48-72 hours in high-risk bleeding cases (Grade 2C evidence).³

On occasion, unanticipated adjustments to surgical cases—or complications—change the previously determined post-operative bleeding risk. In these instances, the hospitalist and surgeon/proceduralist should review the case and reassess the bleeding risk prior to employing bridging anticoagulation protocols.

When to restart long-term vitamin K antagonists (VKA) post-procedure. In most instances, regardless of pre-operative bleeding risk stratification, the resumption of VKA may occur once post-operative hemostasis has been achieved and the patient has been instructed to resume eating by the proceduralist or surgeon. This most often occurs on the calendar day following surgery, because it takes approximately five days for an INR to achieve therapeutic levels.

Back to the Case

The patient’s history of prosthetic valve with stroke within the preceding six months stratified her to a high thrombotic risk category. Given the high risk of thrombosis, the decision was made to bridge with LMWH. The hospitalist discontinued LMWH 24 hours prior to surgery, and INR was checked on the morning of the procedure.

Although the patient underwent the operation without significant bleeding, the adjustment from an exploratory laparoscopy to an open laparotomy increased her post-operative bleeding risk from medium to high. Therefore, bridging anticoagulation with LMWH was resumed no sooner than 48 hours after the operation. Her warfarin was restarted on the day following surgery, once she resumed her diet.

Bottom Line

Hospitalists must understand both the pre- and post-procedure thrombotic risks, as well as the pre- and post-procedural bleeding risks, when determining the selection and logistics of initiation and cessation of antithrombotic bridging for inpatients.

Drs. McCormick, Carbo, and Li are hospitalists at Beth Israel Deaconess Medical Center in Boston. Dr. Kerbel is a hospitalist at the University of California Los Angeles.

References

1. BRIDGE Study Investigators. Bridging anticoagulation: is it needed when warfarin is interrupted around the time of a surgery or procedure? Circulation. 2012;125(12):e496-498.
2. Ansell J, Hirsh J, Hylek E, Jacobson A, Crowther M, Palareti G. Pharmacology and management of the vitamin K antagonists: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest. 2008;133(6 Suppl):160S-198S.
3. Douketis JD, Spyropoulos AC, Spencer FA, et al. Perioperative management of antithrombotic therapy: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest. 2012;141(2 Suppl):e326S-350S.
4. Kovacs MJ, Kearon C, Rodger M, et al. Single-arm study of bridging therapy with low-molecular-weight heparin for patients at risk of arterial embolism who require temporary interruption of warfarin. Circulation. 2004;110(12):1658-1663.
5. Woods K, Douketis JD, Kathirgamanathan K, Yi Q, Crowther MA. Low-dose oral vitamin K to normalize the international normalized ratio prior to surgery in patients who require temporary interruption of warfarin. J Thromb Thrombolysis. 2007;24(2):93-97.
6. Ortel TL. Perioperative management of patients on chronic antithrombotic therapy. Blood. 2012;120(24):4699-4705.
7. Kaatz S, Douketis JD, Zhou H, Gage BF, White RH. Risk of stroke after surgery in patients with and without chronic atrial fibrillation. J Thromb Haemost. 2010;8(5):884-890.
8. Siegal D, Yudin J, Kaatz S, Douketis JD, Lim W, Spyropoulos AC. Periprocedural heparin bridging in patients receiving vitamin K antagonists: systematic review and meta-analysis of bleeding and thromboembolic rates. Circulation. 2012;126(13):1630-1639.
9. Lee AY, Levine MN, Baker RI, et al. Low-molecular-weight heparin versus a coumarin for the prevention of recurrent venous thromboembolism in patients with cancer. N Engl J Med. 2003;349(2):146-153.
10. Hirsh J, Raschke R. Heparin and low-molecular-weight heparin: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest. 2004;126(3 Suppl):188S-203S.