Ian McCormick, MD

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.

Clinical question: Is exogenous glucocorticoid administration associated with an increased risk of VTE?

Background: Endogenous hypercortisolism is linked to increased VTE rates, and pathophysiologic data exist to suggest glucocorticoids increase clotting, but few studies have measured the clinical link between glucocorticoid administration and VTE events.

Study design: Case-control study.

Setting: Denmark.

Synopsis: The authors analyzed Danish national registries, which include information on diagnoses and prescriptions. The study selection period was Jan. 1, 2005, to Dec. 31, 2011. During this period, 38,675 cases of VTE (both DVT and pulmonary embolism) were identified in the population of Denmark. These cases were matched with 387,650 controls. Three routes of glucocorticoid use were studied: systemic (oral and intravenous), inhaled, and intestinal-acting. Cases were classified as present (within 90 days of VTE event), recent (within 91 to 365 days), or former (more than 365 days prior) users of glucocorticoids. Categories were also created for new versus continuous users.

Glucocorticoid use was associated with a significant increase in VTE occurrence. The strongest link was in new and recent users, and the effect diminished over time. Key limitations of the study included its reliance on registry data, as well as the fact that cases had more comorbid conditions than controls (e.g. recent infection, chronic illnesses).

Bottom line: Recipients of glucocorticoids had an increased risk of VTE; the effect was strongest in new and recent users.

Citation: Johannesdottir SA, Horvath-Puho E, Dekkers OM, et al. Use of glucocorticoids and risk of venous thromboembolism. JAMA Intern Med. 2013;173(9):743-752.

Visit our website for more physician reviews of recent HM-relevant literature.

Clinical question: Is exogenous glucocorticoid administration associated with an increased risk of VTE?

Background: Endogenous hypercortisolism is linked to increased VTE rates, and pathophysiologic data exist to suggest glucocorticoids increase clotting, but few studies have measured the clinical link between glucocorticoid administration and VTE events.

Study design: Case-control study.

Setting: Denmark.

Synopsis: The authors analyzed Danish national registries, which include information on diagnoses and prescriptions. The study selection period was Jan. 1, 2005, to Dec. 31, 2011. During this period, 38,675 cases of VTE (both DVT and pulmonary embolism) were identified in the population of Denmark. These cases were matched with 387,650 controls. Three routes of glucocorticoid use were studied: systemic (oral and intravenous), inhaled, and intestinal-acting. Cases were classified as present (within 90 days of VTE event), recent (within 91 to 365 days), or former (more than 365 days prior) users of glucocorticoids. Categories were also created for new versus continuous users.

Glucocorticoid use was associated with a significant increase in VTE occurrence. The strongest link was in new and recent users, and the effect diminished over time. Key limitations of the study included its reliance on registry data, as well as the fact that cases had more comorbid conditions than controls (e.g. recent infection, chronic illnesses).

Bottom line: Recipients of glucocorticoids had an increased risk of VTE; the effect was strongest in new and recent users.

Citation: Johannesdottir SA, Horvath-Puho E, Dekkers OM, et al. Use of glucocorticoids and risk of venous thromboembolism. JAMA Intern Med. 2013;173(9):743-752.

Visit our website for more physician reviews of recent HM-relevant literature.

Clinical question: Is exogenous glucocorticoid administration associated with an increased risk of VTE?

Background: Endogenous hypercortisolism is linked to increased VTE rates, and pathophysiologic data exist to suggest glucocorticoids increase clotting, but few studies have measured the clinical link between glucocorticoid administration and VTE events.

Study design: Case-control study.

Setting: Denmark.

Synopsis: The authors analyzed Danish national registries, which include information on diagnoses and prescriptions. The study selection period was Jan. 1, 2005, to Dec. 31, 2011. During this period, 38,675 cases of VTE (both DVT and pulmonary embolism) were identified in the population of Denmark. These cases were matched with 387,650 controls. Three routes of glucocorticoid use were studied: systemic (oral and intravenous), inhaled, and intestinal-acting. Cases were classified as present (within 90 days of VTE event), recent (within 91 to 365 days), or former (more than 365 days prior) users of glucocorticoids. Categories were also created for new versus continuous users.

Glucocorticoid use was associated with a significant increase in VTE occurrence. The strongest link was in new and recent users, and the effect diminished over time. Key limitations of the study included its reliance on registry data, as well as the fact that cases had more comorbid conditions than controls (e.g. recent infection, chronic illnesses).

Bottom line: Recipients of glucocorticoids had an increased risk of VTE; the effect was strongest in new and recent users.

Citation: Johannesdottir SA, Horvath-Puho E, Dekkers OM, et al. Use of glucocorticoids and risk of venous thromboembolism. JAMA Intern Med. 2013;173(9):743-752.

Visit our website for more physician reviews of recent HM-relevant literature.

In This Edition

Literature At A Glance

A guide to this month’s studies

1. Steroids may increase VTE risk
2. Mortality rates rise at critical-care hospitals
3. Physicians don’t discuss advance-care planning with elderly, families
4. Ultrasound imaging for lumbar puncture, epidural catheterization
5. Financial impact of surgical complications on hospitals
6. Lab test fees and frequency of test ordering
7. Inpatient elderly multidisciplinary teams reduce readmissions
8. Use of beta-blockers in high-risk patients during noncardiac surgery
9. Patient-centered decision-making and health-care outcomes
10. Adverse surgical outcomes in patients with schizophrenia

Steroids May Increase VTE Risk

Clinical question: Is exogenous glucocorticoid administration associated with an increased risk of VTE?

Background: Endogenous hypercortisolism is linked to increased VTE rates, and pathophysiologic data exist to suggest glucocorticoids increase clotting, but few studies have measured the clinical link between glucocorticoid administration and VTE events.

Study design: Case-control study.

Setting: Denmark.

Synopsis: The authors analyzed Danish national registries, which include information on diagnoses and prescriptions. The study selection period was Jan. 1, 2005, to Dec. 31, 2011. During this period, 38,675 cases of VTE (both DVT and pulmonary embolism) were identified in the population of Denmark. These cases were matched with 387,650 controls. Three routes of glucocorticoid use were studied: systemic (oral and intravenous), inhaled, and intestinal-acting. Cases were classified as present (within 90 days of VTE event), recent (91 to 365 days), or former (over 365 days) users of glucocorticoids. Categories were also created for new versus continuous users.

Glucocorticoid use was associated with a significant increase in VTE occurrence. The strongest link was in new and recent users, and the effect diminished over time. Key limitations of the study included its reliance on registry data, as well as the fact that cases had more comorbid conditions than controls (e.g. recent infection, chronic illnesses).

Bottom line: Recipients of glucocorticoids had an increased risk of VTE; the effect was strongest in new and recent users.

Citation: Johannesdottir SA, Horvath-Puho E, Dekkers OM, et al. Use of glucocorticoids and risk of venous thromboembolism. JAMA Intern Med. 2013;173(9):743-752.

Mortality Rates Rise at Critical-Access Hospitals

Clinical question: How have trends in mortality changed in the past decade at critical-access hospitals when compared to other hospitals?

Background: Hospitals are designated as critical-access hospitals (CAH) by meeting certain requirements—namely, rural setting, small number of beds, and minimum distance from the nearest hospital. Because of the intrinsic challenges they face, CAHs are exempt from certain quality measures. Little data exist on patient outcomes at CAHs.

Study design: Retrospective observational study.

Setting: All nonfederal hospitals in the U.S. that provide acute care to Medicare beneficiaries.

Synopsis: Using Medicare data, risk-adjusted 30-day mortality rates were calculated at critical-access hospitals and non-critical-access hospitals from 2002 to 2010 for three conditions: acute myocardial infarction, congestive heart failure, and pneumonia. Mortality trends across these conditions were compared at baseline and over time. In 2002, CAHs had mortality rates comparable with non-CAHs. From 2002 to 2010, mortality rates increased by 0.1% per year at CAHs and decreased by 0.2% per year at non-CAHs, resulting in a mortality difference between these two hospital groups at the end of the study period. The authors postulate that differences in policy initiatives, enforcement of quality measures, and access to technology may play a role.

This study is limited in its ability to explain the findings. Notably, within the CAHs, the authors found that 48% did improve from 2002 to 2010; there were no significant differences in hospital characteristics between the CAHs that did and did not improve. The reasons for the overall widening gap between CAHs and non-CAHs, therefore, merit further investigation.

Bottom line: From 2002 to 2010, mortality rates at U.S. critical-access hospitals rose while rates fell at non-critical access hospitals.

Citation: Joynt KE, Orav EJ, Jha AK. Mortality rates for Medicare beneficiaries admitted to critical access and non-critical access hospitals, 2002-2010. JAMA. 2013;309(13):1379-1387.

Physicians Fail to Discuss Advance-Care Planning with Hospitalized Elderly, Relatives

Clinical question: Do elderly patients who are at high risk for death have documented advance-care plans, and do their physicians discuss goals of care with them?

Background: While studies indicate that most sick, elderly patients prefer a better quality of life to life-sustaining treatment, many patients still die in the hospital, often in the ICU. It is unclear whether this is due to the absence of advance-care planning or the lack of communicating goals of care.

Study design: Multicenter prospective cohort study.

Setting: Canada.

Synopsis: A validated questionnaire was used to evaluate advance-care planning both before hospitalization and on admission. Patients and their families were enrolled if they were considered at high risk for dying, including patients with advanced disease, or aged >80 years. Of the 278 patients enrolled, 76.3% had thought about advance-care planning. Approximately 47.9% of patients had written an advance-care plan, and 73.3% had formally documented a surrogate health-care decision-maker. Only a quarter of patients reported that they had been asked about advance-care planning on admission to the hospital. Patients’ stated preferences for end-of-life care were notably incongruent with goals-of-care orders documented in the hospital records. While 28% of patients preferred comfort-only care, this was documented in the hospital records for only 4.5% of patients.

In this study of mostly white, English-speaking hospitalized patients, many had considered and made advance-care plans but few had discussed this with their health-care providers. Not surprisingly, the goals-of-care orders that were documented did not match patients’ previously stated end-of-life preferences.

Bottom line: Physicians routinely fail to discuss patients’ advance-care planning, which may have profound effects on their inpatient care.

Citation: Heyland DK, Barwich D, Pichora D, et al. Failure to engage hospitalized elderly patients and their families in advance care planning. JAMA Intern Med. 2013;173(9):778-787.

Ultrasound Use Reduces Lumbar Puncture and Epidural Catheterization Failure

Clinical question: Does ultrasound imaging for lumbar puncture and epidural catheterization reduce risk of procedure failures?

Background: Numerous studies have compared ultrasound-guided lumbar puncture and epidural catheterization with standard techniques, but they have been underpowered to evaluate whether ultrasound use reduces lumbar puncture and epidural catheterization failure.

Study design: Systematic review and meta-analysis of randomized controlled trials.

Setting: Hospitals in North America, Europe, and Asia.

Synopsis: Among 14 trials, a total of 1,334 patients (including one pediatric study group) were randomly assigned either to receive ultrasound imaging or to a control group (using manual palpation or loss of resistance). Ultrasound imaging comprised a preprocedure marking approach, real-time visual guidance, or both. In the 12 trials in which the primary outcome was available, ultrasound imaging significantly reduced the risk of failed procedures with a risk ratio of 0.21 (95% confidence interval 0.10 to 0.43). A total of 16 ultrasound-guided procedures were needed to avoid one procedure failure. Ultrasound guidance also reduced the number of traumatic procedures, needle reinsertions, and needle redirections. These results were consistent across multiple subgroup analyses.

Only one of the studies included in the meta-analysis was double-blind, and in all of the studies, ultrasound imaging was performed by a clinician with high-level ultrasound experience. While this study suggests that ultrasound does reduce the frequency of procedure failure, the investigators did not include cost-effectiveness analyses.

Bottom line: Ultrasound imaging likely reduces the failure of lumbar puncture and epidural catheterization, but future research will be necessary to evaluate the cost-effectiveness of ultrasound use for these procedures.

Citation: Shaikh F, Brzezinski J, Alexander S, et al. Ultrasound imaging for lumbar punctures and epidural catheterisations: systematic review and meta-analysis. BMJ. 2013;346:f1720.

Surgical Complications might Be Financially Advantageous for Hospitals

Clinical question: What is the impact of surgical complications on hospital finances?

Background: Surgical complications are common and lead to longer lengths of stay and higher costs. Strategies are available to reduce postsurgical complications but have not been universally adopted.

Study design: Observational study.

Setting: Twelve hospitals in one nonprofit Southern hospital system, which includes academic, nonacademic, and rural settings.

Synopsis: Researchers identified 34,526 patients who underwent surgery in 2010, excluding patients undergoing Caesarean section. Of those, 1,820 procedures (5.3%) were associated with at least one complication. The most frequent complications were surgical-site infection, other infections, pneumonia, and thromboembolic disease. The mortality rate for patients with complications was 12.3% compared with 0.6% for those without. Length of stay was four times longer for patients with complications.

Complications were associated with a higher total cost of hospitalization, with a differential of $37,917. This translated into a higher contribution to the margin. The cost differential varied by insurance type, with higher contributions under Medicare and private insurance but not with Medicaid.

The study had the benefit of using a large administrative database; however, this may have underestimated the actual rate of postoperative complications. The study supports the paradox in which quality-improvement (QI) programs that reduce surgical complications and improve postoperative mortality may negatively affect a hospital’s financial performance.

Bottom line: Surgical complications lead to higher mortality for patients but a financial benefit for hospitals.

Citation: Eappen S, Lane BH, Rosenberg B, et al. Relationship between occurrence of surgical complications and hospital finances. JAMA. 2013;309(5):1509-1606.

Showing Lab Test Fees May Decrease Frequency Of Test Ordering

Clinical question: Does information on the cost of a lab test lead to lowered ordering frequency among internal-medicine residents?

Background: Lab test overuse is common; some studies estimate that 70% of lab tests do not affect care. Strategies to reduce frequency of unnecessary lab tests are needed.

Study design: Randomized controlled trial.

Setting: Johns Hopkins Hospital, a 1,051-bed academic medical center in Baltimore.

Synopsis: Researchers used an administrative database to identify the 35 most frequently ordered and the 35 most expensive tests (each ordered at least 50 times). They randomized tests to an active arm, which displayed the Medicare allowable fee at the time of order entry within the computerized physician order entry, and a control arm. A total of 1,166,753 tests were ordered during the baseline and intervention period. Many more tests were ordered in the active group relative to the control group, a consequence of the randomization process. Relative to a six-month baseline period, tests in the active group were ordered 9.1% less frequently; control-group tests were ordered 5.1% more frequently. Charges decreased by $3.79 per patient-day in the active group and increased by $0.52 per patient-day in the control group.

This study reflects a low-cost strategy to reduce lab testing and associated costs. It is unknown whether only unnecessary tests were averted, or if there was any effect on the quality of care. The durability of the intervention and its applicability to other settings and with other types of providers is unclear.

Bottom line: Showing the fee associated with lab tests may decrease the frequency of ordering these tests and the resultant costs.

Citation: Feldman LS, Shihab HM, Thiemann D, et al. Impact of providing fee data on laboratory test ordering: a controlled clinical trial. JAMA Intern Med. 2013;173(10):903-908.

Inpatient Elderly Multidisciplinary-Team-Based Unit Reduces Readmissions

Clinical question: Can an inpatient unit that uses interdisciplinary rounds and a team approach reduce 30-day readmissions and lower cost compared with usual care?

Background: Elderly patients are at risk for adverse outcomes including delirium, deconditioning, and undernutrition. The Acute Care for Elders (ACE) model is designed as a team-based approach to improve communication between clinicians and to begin discharge planning soon after admission. ACE units have been shown to improve specific outcomes in older adults, but the impact on cost is less clear.

Study design: Retrospective cohort study.

Setting: University of Alabama at Birmingham Acute Care for Elders unit.

Synopsis: The analysis of cost data extrapolated from administrative sources indicated an average of $371 in variable cost savings per patient, and therefore an expected $148,400 savings for 400 patients admitted to this unit. The rate of 30-day readmissions was significantly lower in the ACE model: 7.9% versus 12.8% in patients receiving usual care.

Limitations of this study relate both to its design and single-center location. It is unclear which aspect of this ACE unit was helpful in the studied outcomes, and how this specific program would be reproduced at another institution. There were also some costs associated with the ACE-unit staffing, which may have resulted in an overestimation of the cost savings.

Bottom line: ACE units appear to improve outcomes and may lower cost, but further investigation is needed.

Citation: Flood KL, MacLennan PA, McGrew D, Green D, Dodd C, Brown CJ. Effects of an acute care for elders unit on costs and 30-day readmissions. JAMA. 2013 April 22 [Epub ahead of print].

Continue Beta-Blockers in High-Risk Patients During Noncardiac Surgery

Clinical question: Does exposure to a beta-blocker at the time of noncardiac surgery have an association with mortality, cardiac arrest, or Q-wave myocardial infarction (MI)?

Background: Exposure to beta-blockers perioperatively is controversial. The authors suggest that a randomized trial would be difficult, as one might need to withhold beta-blockers from stable patients. The American Heart Association and American College of Cardiology currently recommend continuation of pre-existing beta-blockades perioperatively.

Study design: Retrospective, propensity-matched cohort analysis.

Setting: One hundred four Veterans Affairs (VA) medical centers from January 2005 to August 2010.

Synopsis: Exposure for any reason to beta-blockers around nonvascular surgery appears to lower the risk of mortality and cardiac arrest, particularly in patients with higher revised cardiac risk index factors. Stopping beta-blockers in this same period was associated with increased mortality. The reasons for beta-blocker use in the exposed cohort were not well-established or -stratified.

Limitations include that the exposed cohort had a higher rate of cardiovascular disease and comorbidities and were generally older. Duration of exposure to beta-blockers was mixed, although more than 75% had a prescription for more than three months, with less than 2% for several days. Some of the matched cohort appears to have had beta-blocker exposure, thus diminishing the potential impact.

Bottom line: In this retrospective analysis, there is an association between beta-blocker use during noncardiac, nonvascular surgery and lower mortality, as well as lower rates of cardiac arrest and Q-wave MI, particularly among higher-risk patients.

Citation: London MJ, Hur K, Schwartz G, Henderson WG. Association of perioperative beta-blockade with mortality and cardiovascular morbidity following major noncardiac surgery. JAMA. 2013;309(16):1704-1713.

Patient-Centered Decision-Making and Health-Care Outcomes

Clinical question: Does recognition of contextual domains in the care plan lead to resolution of patients’ presenting problems?

Background: Patient-centered decision-making, or contextualization of care, adapts best evidence to the care of the individual patient. Examples of contextual domains include access to care, social support, and financial situation. Contextual errors (e.g. unrecognized domains) are, on average, more costly than errors in evidenced-based medicine.

Study design: Observational study.

Setting: VA ambulatory-care centers.

Synopsis: A total of 548 patient-derived audio recordings of physician encounters that included contextual red flags (e.g. missed appointments, HbA1c >8%) were reviewed at two VA hospital ambulatory-care centers using the content coding for contextualization of care (4C) method. Prospectively determined good and poor outcomes were derived from specific red flags (e.g. keeping next appointment, decrease in HbA1c). Of the 548 red flags, 208 were associated with contextual domains using the 4C method. Some 59% of physicians recognized contextual domains in care-plan development, leading to good outcomes in 71% of red flags. As many as 41% of physicians did not recognize contextual domains, leading to poor outcomes in 54% of red-flag instances.

Hospitalists should be aware of contextual domains and red flags (e.g. readmissions), and this study provides a method of evaluating patient-centered decision-making in the hospital setting. However, the inherently subjective 4C method may underestimate the number of contextual domains.

Bottom line: Recognition and incorporation of contextual domains in care-plan development in the ambulatory setting are associated with improved contextual red flag outcomes.

Citation: Weiner SJ, Schwartz A, Sharma G, et al. Patient-centered decision making and health care outcomes: an observational study. Ann Intern Med. 2013;158:573-579.

Adverse Surgical Outcomes in Schizophrenia Patients

Clinical question: What is the full spectrum of postoperative complications and mortality in schizophrenic patients receiving in-hospital major surgery?

Background: Schizophrenia affects an estimated 4 to 7 per 1,000 persons worldwide, and patients with schizophrenia are at increased risk of diabetes, cardiovascular disease, and stroke with subsequent increased mortality risk. The correlation between severity of mental illness and postoperative in-hospital mortality has not been validated in previous studies.

Study design: Population-based, cross-sectional study.

Setting: Taiwan hospitals.

Synopsis: Using the Taiwan National Health Insurance Research Database, researchers examined claims from 2004 to 2007 and retrospectively identified 8,967 schizophrenic patients who underwent major inpatient surgery and were hospitalized for more than one day. Primary outcomes included acute myocardial infarction, acute renal failure, stroke, and in-hospital mortality within 30 days. Postoperative complications and mortality rates were compared between schizophrenic patients and patients without mental illness.

Schizophrenic patients had higher rates of many primary outcomes, including 30-day postoperative mortality, compared with patients without mental illness, after adjusting for sex, age, surgery type, and hospital setting. The risk for 30-day mortality rose with the number of preoperative schizophrenia-related services provided. Limitations include the nature of retrospective analysis and generalizability.

Bottom line: Compared to patients without mental illness, schizophrenic patients have an increased risk of acute renal failure, pneumonia, septicemia, and 30-day mortality in the postoperative setting, with higher mortality rates in schizophrenic patients with more severe disease.

Citation: Liao CC, Shen WW, Chang CC, Chang H, Chen T. Surgical adverse outcomes in patients with schizophrenia. Ann Surg. 2013;257:433-438.

In This Edition

Literature At A Glance

A guide to this month’s studies

1. Steroids may increase VTE risk
2. Mortality rates rise at critical-care hospitals
3. Physicians don’t discuss advance-care planning with elderly, families
4. Ultrasound imaging for lumbar puncture, epidural catheterization
5. Financial impact of surgical complications on hospitals
6. Lab test fees and frequency of test ordering
7. Inpatient elderly multidisciplinary teams reduce readmissions
8. Use of beta-blockers in high-risk patients during noncardiac surgery
9. Patient-centered decision-making and health-care outcomes
10. Adverse surgical outcomes in patients with schizophrenia

Steroids May Increase VTE Risk

Clinical question: Is exogenous glucocorticoid administration associated with an increased risk of VTE?

Background: Endogenous hypercortisolism is linked to increased VTE rates, and pathophysiologic data exist to suggest glucocorticoids increase clotting, but few studies have measured the clinical link between glucocorticoid administration and VTE events.

Study design: Case-control study.

Setting: Denmark.

Synopsis: The authors analyzed Danish national registries, which include information on diagnoses and prescriptions. The study selection period was Jan. 1, 2005, to Dec. 31, 2011. During this period, 38,675 cases of VTE (both DVT and pulmonary embolism) were identified in the population of Denmark. These cases were matched with 387,650 controls. Three routes of glucocorticoid use were studied: systemic (oral and intravenous), inhaled, and intestinal-acting. Cases were classified as present (within 90 days of VTE event), recent (91 to 365 days), or former (over 365 days) users of glucocorticoids. Categories were also created for new versus continuous users.

Glucocorticoid use was associated with a significant increase in VTE occurrence. The strongest link was in new and recent users, and the effect diminished over time. Key limitations of the study included its reliance on registry data, as well as the fact that cases had more comorbid conditions than controls (e.g. recent infection, chronic illnesses).

Bottom line: Recipients of glucocorticoids had an increased risk of VTE; the effect was strongest in new and recent users.

Citation: Johannesdottir SA, Horvath-Puho E, Dekkers OM, et al. Use of glucocorticoids and risk of venous thromboembolism. JAMA Intern Med. 2013;173(9):743-752.

Mortality Rates Rise at Critical-Access Hospitals

Clinical question: How have trends in mortality changed in the past decade at critical-access hospitals when compared to other hospitals?

Background: Hospitals are designated as critical-access hospitals (CAH) by meeting certain requirements—namely, rural setting, small number of beds, and minimum distance from the nearest hospital. Because of the intrinsic challenges they face, CAHs are exempt from certain quality measures. Little data exist on patient outcomes at CAHs.

Study design: Retrospective observational study.

Setting: All nonfederal hospitals in the U.S. that provide acute care to Medicare beneficiaries.

Synopsis: Using Medicare data, risk-adjusted 30-day mortality rates were calculated at critical-access hospitals and non-critical-access hospitals from 2002 to 2010 for three conditions: acute myocardial infarction, congestive heart failure, and pneumonia. Mortality trends across these conditions were compared at baseline and over time. In 2002, CAHs had mortality rates comparable with non-CAHs. From 2002 to 2010, mortality rates increased by 0.1% per year at CAHs and decreased by 0.2% per year at non-CAHs, resulting in a mortality difference between these two hospital groups at the end of the study period. The authors postulate that differences in policy initiatives, enforcement of quality measures, and access to technology may play a role.

This study is limited in its ability to explain the findings. Notably, within the CAHs, the authors found that 48% did improve from 2002 to 2010; there were no significant differences in hospital characteristics between the CAHs that did and did not improve. The reasons for the overall widening gap between CAHs and non-CAHs, therefore, merit further investigation.

Bottom line: From 2002 to 2010, mortality rates at U.S. critical-access hospitals rose while rates fell at non-critical access hospitals.

Citation: Joynt KE, Orav EJ, Jha AK. Mortality rates for Medicare beneficiaries admitted to critical access and non-critical access hospitals, 2002-2010. JAMA. 2013;309(13):1379-1387.

Physicians Fail to Discuss Advance-Care Planning with Hospitalized Elderly, Relatives

Clinical question: Do elderly patients who are at high risk for death have documented advance-care plans, and do their physicians discuss goals of care with them?

Background: While studies indicate that most sick, elderly patients prefer a better quality of life to life-sustaining treatment, many patients still die in the hospital, often in the ICU. It is unclear whether this is due to the absence of advance-care planning or the lack of communicating goals of care.

Study design: Multicenter prospective cohort study.

Setting: Canada.

Synopsis: A validated questionnaire was used to evaluate advance-care planning both before hospitalization and on admission. Patients and their families were enrolled if they were considered at high risk for dying, including patients with advanced disease, or aged >80 years. Of the 278 patients enrolled, 76.3% had thought about advance-care planning. Approximately 47.9% of patients had written an advance-care plan, and 73.3% had formally documented a surrogate health-care decision-maker. Only a quarter of patients reported that they had been asked about advance-care planning on admission to the hospital. Patients’ stated preferences for end-of-life care were notably incongruent with goals-of-care orders documented in the hospital records. While 28% of patients preferred comfort-only care, this was documented in the hospital records for only 4.5% of patients.

In this study of mostly white, English-speaking hospitalized patients, many had considered and made advance-care plans but few had discussed this with their health-care providers. Not surprisingly, the goals-of-care orders that were documented did not match patients’ previously stated end-of-life preferences.

Bottom line: Physicians routinely fail to discuss patients’ advance-care planning, which may have profound effects on their inpatient care.

Citation: Heyland DK, Barwich D, Pichora D, et al. Failure to engage hospitalized elderly patients and their families in advance care planning. JAMA Intern Med. 2013;173(9):778-787.

Ultrasound Use Reduces Lumbar Puncture and Epidural Catheterization Failure

Clinical question: Does ultrasound imaging for lumbar puncture and epidural catheterization reduce risk of procedure failures?

Background: Numerous studies have compared ultrasound-guided lumbar puncture and epidural catheterization with standard techniques, but they have been underpowered to evaluate whether ultrasound use reduces lumbar puncture and epidural catheterization failure.

Study design: Systematic review and meta-analysis of randomized controlled trials.

Setting: Hospitals in North America, Europe, and Asia.

Synopsis: Among 14 trials, a total of 1,334 patients (including one pediatric study group) were randomly assigned either to receive ultrasound imaging or to a control group (using manual palpation or loss of resistance). Ultrasound imaging comprised a preprocedure marking approach, real-time visual guidance, or both. In the 12 trials in which the primary outcome was available, ultrasound imaging significantly reduced the risk of failed procedures with a risk ratio of 0.21 (95% confidence interval 0.10 to 0.43). A total of 16 ultrasound-guided procedures were needed to avoid one procedure failure. Ultrasound guidance also reduced the number of traumatic procedures, needle reinsertions, and needle redirections. These results were consistent across multiple subgroup analyses.

Only one of the studies included in the meta-analysis was double-blind, and in all of the studies, ultrasound imaging was performed by a clinician with high-level ultrasound experience. While this study suggests that ultrasound does reduce the frequency of procedure failure, the investigators did not include cost-effectiveness analyses.

Bottom line: Ultrasound imaging likely reduces the failure of lumbar puncture and epidural catheterization, but future research will be necessary to evaluate the cost-effectiveness of ultrasound use for these procedures.

Citation: Shaikh F, Brzezinski J, Alexander S, et al. Ultrasound imaging for lumbar punctures and epidural catheterisations: systematic review and meta-analysis. BMJ. 2013;346:f1720.

Surgical Complications might Be Financially Advantageous for Hospitals

Clinical question: What is the impact of surgical complications on hospital finances?

Background: Surgical complications are common and lead to longer lengths of stay and higher costs. Strategies are available to reduce postsurgical complications but have not been universally adopted.

Study design: Observational study.

Setting: Twelve hospitals in one nonprofit Southern hospital system, which includes academic, nonacademic, and rural settings.

Synopsis: Researchers identified 34,526 patients who underwent surgery in 2010, excluding patients undergoing Caesarean section. Of those, 1,820 procedures (5.3%) were associated with at least one complication. The most frequent complications were surgical-site infection, other infections, pneumonia, and thromboembolic disease. The mortality rate for patients with complications was 12.3% compared with 0.6% for those without. Length of stay was four times longer for patients with complications.

Complications were associated with a higher total cost of hospitalization, with a differential of $37,917. This translated into a higher contribution to the margin. The cost differential varied by insurance type, with higher contributions under Medicare and private insurance but not with Medicaid.

The study had the benefit of using a large administrative database; however, this may have underestimated the actual rate of postoperative complications. The study supports the paradox in which quality-improvement (QI) programs that reduce surgical complications and improve postoperative mortality may negatively affect a hospital’s financial performance.

Bottom line: Surgical complications lead to higher mortality for patients but a financial benefit for hospitals.

Citation: Eappen S, Lane BH, Rosenberg B, et al. Relationship between occurrence of surgical complications and hospital finances. JAMA. 2013;309(5):1509-1606.

Showing Lab Test Fees May Decrease Frequency Of Test Ordering

Clinical question: Does information on the cost of a lab test lead to lowered ordering frequency among internal-medicine residents?

Background: Lab test overuse is common; some studies estimate that 70% of lab tests do not affect care. Strategies to reduce frequency of unnecessary lab tests are needed.

Study design: Randomized controlled trial.

Setting: Johns Hopkins Hospital, a 1,051-bed academic medical center in Baltimore.

Synopsis: Researchers used an administrative database to identify the 35 most frequently ordered and the 35 most expensive tests (each ordered at least 50 times). They randomized tests to an active arm, which displayed the Medicare allowable fee at the time of order entry within the computerized physician order entry, and a control arm. A total of 1,166,753 tests were ordered during the baseline and intervention period. Many more tests were ordered in the active group relative to the control group, a consequence of the randomization process. Relative to a six-month baseline period, tests in the active group were ordered 9.1% less frequently; control-group tests were ordered 5.1% more frequently. Charges decreased by $3.79 per patient-day in the active group and increased by $0.52 per patient-day in the control group.

This study reflects a low-cost strategy to reduce lab testing and associated costs. It is unknown whether only unnecessary tests were averted, or if there was any effect on the quality of care. The durability of the intervention and its applicability to other settings and with other types of providers is unclear.

Bottom line: Showing the fee associated with lab tests may decrease the frequency of ordering these tests and the resultant costs.

Citation: Feldman LS, Shihab HM, Thiemann D, et al. Impact of providing fee data on laboratory test ordering: a controlled clinical trial. JAMA Intern Med. 2013;173(10):903-908.

Inpatient Elderly Multidisciplinary-Team-Based Unit Reduces Readmissions

Clinical question: Can an inpatient unit that uses interdisciplinary rounds and a team approach reduce 30-day readmissions and lower cost compared with usual care?

Background: Elderly patients are at risk for adverse outcomes including delirium, deconditioning, and undernutrition. The Acute Care for Elders (ACE) model is designed as a team-based approach to improve communication between clinicians and to begin discharge planning soon after admission. ACE units have been shown to improve specific outcomes in older adults, but the impact on cost is less clear.

Study design: Retrospective cohort study.

Setting: University of Alabama at Birmingham Acute Care for Elders unit.

Synopsis: The analysis of cost data extrapolated from administrative sources indicated an average of $371 in variable cost savings per patient, and therefore an expected $148,400 savings for 400 patients admitted to this unit. The rate of 30-day readmissions was significantly lower in the ACE model: 7.9% versus 12.8% in patients receiving usual care.

Limitations of this study relate both to its design and single-center location. It is unclear which aspect of this ACE unit was helpful in the studied outcomes, and how this specific program would be reproduced at another institution. There were also some costs associated with the ACE-unit staffing, which may have resulted in an overestimation of the cost savings.

Bottom line: ACE units appear to improve outcomes and may lower cost, but further investigation is needed.

Citation: Flood KL, MacLennan PA, McGrew D, Green D, Dodd C, Brown CJ. Effects of an acute care for elders unit on costs and 30-day readmissions. JAMA. 2013 April 22 [Epub ahead of print].

Continue Beta-Blockers in High-Risk Patients During Noncardiac Surgery

Clinical question: Does exposure to a beta-blocker at the time of noncardiac surgery have an association with mortality, cardiac arrest, or Q-wave myocardial infarction (MI)?

Background: Exposure to beta-blockers perioperatively is controversial. The authors suggest that a randomized trial would be difficult, as one might need to withhold beta-blockers from stable patients. The American Heart Association and American College of Cardiology currently recommend continuation of pre-existing beta-blockades perioperatively.

Study design: Retrospective, propensity-matched cohort analysis.

Setting: One hundred four Veterans Affairs (VA) medical centers from January 2005 to August 2010.

Synopsis: Exposure for any reason to beta-blockers around nonvascular surgery appears to lower the risk of mortality and cardiac arrest, particularly in patients with higher revised cardiac risk index factors. Stopping beta-blockers in this same period was associated with increased mortality. The reasons for beta-blocker use in the exposed cohort were not well-established or -stratified.

Limitations include that the exposed cohort had a higher rate of cardiovascular disease and comorbidities and were generally older. Duration of exposure to beta-blockers was mixed, although more than 75% had a prescription for more than three months, with less than 2% for several days. Some of the matched cohort appears to have had beta-blocker exposure, thus diminishing the potential impact.

Bottom line: In this retrospective analysis, there is an association between beta-blocker use during noncardiac, nonvascular surgery and lower mortality, as well as lower rates of cardiac arrest and Q-wave MI, particularly among higher-risk patients.

Citation: London MJ, Hur K, Schwartz G, Henderson WG. Association of perioperative beta-blockade with mortality and cardiovascular morbidity following major noncardiac surgery. JAMA. 2013;309(16):1704-1713.

Patient-Centered Decision-Making and Health-Care Outcomes

Clinical question: Does recognition of contextual domains in the care plan lead to resolution of patients’ presenting problems?

Background: Patient-centered decision-making, or contextualization of care, adapts best evidence to the care of the individual patient. Examples of contextual domains include access to care, social support, and financial situation. Contextual errors (e.g. unrecognized domains) are, on average, more costly than errors in evidenced-based medicine.

Study design: Observational study.

Setting: VA ambulatory-care centers.

Synopsis: A total of 548 patient-derived audio recordings of physician encounters that included contextual red flags (e.g. missed appointments, HbA1c >8%) were reviewed at two VA hospital ambulatory-care centers using the content coding for contextualization of care (4C) method. Prospectively determined good and poor outcomes were derived from specific red flags (e.g. keeping next appointment, decrease in HbA1c). Of the 548 red flags, 208 were associated with contextual domains using the 4C method. Some 59% of physicians recognized contextual domains in care-plan development, leading to good outcomes in 71% of red flags. As many as 41% of physicians did not recognize contextual domains, leading to poor outcomes in 54% of red-flag instances.

Hospitalists should be aware of contextual domains and red flags (e.g. readmissions), and this study provides a method of evaluating patient-centered decision-making in the hospital setting. However, the inherently subjective 4C method may underestimate the number of contextual domains.

Bottom line: Recognition and incorporation of contextual domains in care-plan development in the ambulatory setting are associated with improved contextual red flag outcomes.

Citation: Weiner SJ, Schwartz A, Sharma G, et al. Patient-centered decision making and health care outcomes: an observational study. Ann Intern Med. 2013;158:573-579.

Adverse Surgical Outcomes in Schizophrenia Patients

Clinical question: What is the full spectrum of postoperative complications and mortality in schizophrenic patients receiving in-hospital major surgery?

Background: Schizophrenia affects an estimated 4 to 7 per 1,000 persons worldwide, and patients with schizophrenia are at increased risk of diabetes, cardiovascular disease, and stroke with subsequent increased mortality risk. The correlation between severity of mental illness and postoperative in-hospital mortality has not been validated in previous studies.

Study design: Population-based, cross-sectional study.

Setting: Taiwan hospitals.

Synopsis: Using the Taiwan National Health Insurance Research Database, researchers examined claims from 2004 to 2007 and retrospectively identified 8,967 schizophrenic patients who underwent major inpatient surgery and were hospitalized for more than one day. Primary outcomes included acute myocardial infarction, acute renal failure, stroke, and in-hospital mortality within 30 days. Postoperative complications and mortality rates were compared between schizophrenic patients and patients without mental illness.

Schizophrenic patients had higher rates of many primary outcomes, including 30-day postoperative mortality, compared with patients without mental illness, after adjusting for sex, age, surgery type, and hospital setting. The risk for 30-day mortality rose with the number of preoperative schizophrenia-related services provided. Limitations include the nature of retrospective analysis and generalizability.

Bottom line: Compared to patients without mental illness, schizophrenic patients have an increased risk of acute renal failure, pneumonia, septicemia, and 30-day mortality in the postoperative setting, with higher mortality rates in schizophrenic patients with more severe disease.

Citation: Liao CC, Shen WW, Chang CC, Chang H, Chen T. Surgical adverse outcomes in patients with schizophrenia. Ann Surg. 2013;257:433-438.

In This Edition

Literature At A Glance

A guide to this month’s studies

1. Steroids may increase VTE risk
2. Mortality rates rise at critical-care hospitals
3. Physicians don’t discuss advance-care planning with elderly, families
4. Ultrasound imaging for lumbar puncture, epidural catheterization
5. Financial impact of surgical complications on hospitals
6. Lab test fees and frequency of test ordering
7. Inpatient elderly multidisciplinary teams reduce readmissions
8. Use of beta-blockers in high-risk patients during noncardiac surgery
9. Patient-centered decision-making and health-care outcomes
10. Adverse surgical outcomes in patients with schizophrenia

Steroids May Increase VTE Risk

Clinical question: Is exogenous glucocorticoid administration associated with an increased risk of VTE?

Background: Endogenous hypercortisolism is linked to increased VTE rates, and pathophysiologic data exist to suggest glucocorticoids increase clotting, but few studies have measured the clinical link between glucocorticoid administration and VTE events.

Study design: Case-control study.

Setting: Denmark.

Synopsis: The authors analyzed Danish national registries, which include information on diagnoses and prescriptions. The study selection period was Jan. 1, 2005, to Dec. 31, 2011. During this period, 38,675 cases of VTE (both DVT and pulmonary embolism) were identified in the population of Denmark. These cases were matched with 387,650 controls. Three routes of glucocorticoid use were studied: systemic (oral and intravenous), inhaled, and intestinal-acting. Cases were classified as present (within 90 days of VTE event), recent (91 to 365 days), or former (over 365 days) users of glucocorticoids. Categories were also created for new versus continuous users.

Glucocorticoid use was associated with a significant increase in VTE occurrence. The strongest link was in new and recent users, and the effect diminished over time. Key limitations of the study included its reliance on registry data, as well as the fact that cases had more comorbid conditions than controls (e.g. recent infection, chronic illnesses).

Bottom line: Recipients of glucocorticoids had an increased risk of VTE; the effect was strongest in new and recent users.

Citation: Johannesdottir SA, Horvath-Puho E, Dekkers OM, et al. Use of glucocorticoids and risk of venous thromboembolism. JAMA Intern Med. 2013;173(9):743-752.

Mortality Rates Rise at Critical-Access Hospitals

Clinical question: How have trends in mortality changed in the past decade at critical-access hospitals when compared to other hospitals?

Background: Hospitals are designated as critical-access hospitals (CAH) by meeting certain requirements—namely, rural setting, small number of beds, and minimum distance from the nearest hospital. Because of the intrinsic challenges they face, CAHs are exempt from certain quality measures. Little data exist on patient outcomes at CAHs.

Study design: Retrospective observational study.

Setting: All nonfederal hospitals in the U.S. that provide acute care to Medicare beneficiaries.

Synopsis: Using Medicare data, risk-adjusted 30-day mortality rates were calculated at critical-access hospitals and non-critical-access hospitals from 2002 to 2010 for three conditions: acute myocardial infarction, congestive heart failure, and pneumonia. Mortality trends across these conditions were compared at baseline and over time. In 2002, CAHs had mortality rates comparable with non-CAHs. From 2002 to 2010, mortality rates increased by 0.1% per year at CAHs and decreased by 0.2% per year at non-CAHs, resulting in a mortality difference between these two hospital groups at the end of the study period. The authors postulate that differences in policy initiatives, enforcement of quality measures, and access to technology may play a role.

This study is limited in its ability to explain the findings. Notably, within the CAHs, the authors found that 48% did improve from 2002 to 2010; there were no significant differences in hospital characteristics between the CAHs that did and did not improve. The reasons for the overall widening gap between CAHs and non-CAHs, therefore, merit further investigation.

Bottom line: From 2002 to 2010, mortality rates at U.S. critical-access hospitals rose while rates fell at non-critical access hospitals.

Citation: Joynt KE, Orav EJ, Jha AK. Mortality rates for Medicare beneficiaries admitted to critical access and non-critical access hospitals, 2002-2010. JAMA. 2013;309(13):1379-1387.

Physicians Fail to Discuss Advance-Care Planning with Hospitalized Elderly, Relatives

Clinical question: Do elderly patients who are at high risk for death have documented advance-care plans, and do their physicians discuss goals of care with them?

Background: While studies indicate that most sick, elderly patients prefer a better quality of life to life-sustaining treatment, many patients still die in the hospital, often in the ICU. It is unclear whether this is due to the absence of advance-care planning or the lack of communicating goals of care.

Study design: Multicenter prospective cohort study.

Setting: Canada.

Synopsis: A validated questionnaire was used to evaluate advance-care planning both before hospitalization and on admission. Patients and their families were enrolled if they were considered at high risk for dying, including patients with advanced disease, or aged >80 years. Of the 278 patients enrolled, 76.3% had thought about advance-care planning. Approximately 47.9% of patients had written an advance-care plan, and 73.3% had formally documented a surrogate health-care decision-maker. Only a quarter of patients reported that they had been asked about advance-care planning on admission to the hospital. Patients’ stated preferences for end-of-life care were notably incongruent with goals-of-care orders documented in the hospital records. While 28% of patients preferred comfort-only care, this was documented in the hospital records for only 4.5% of patients.

In this study of mostly white, English-speaking hospitalized patients, many had considered and made advance-care plans but few had discussed this with their health-care providers. Not surprisingly, the goals-of-care orders that were documented did not match patients’ previously stated end-of-life preferences.

Bottom line: Physicians routinely fail to discuss patients’ advance-care planning, which may have profound effects on their inpatient care.

Citation: Heyland DK, Barwich D, Pichora D, et al. Failure to engage hospitalized elderly patients and their families in advance care planning. JAMA Intern Med. 2013;173(9):778-787.

Ultrasound Use Reduces Lumbar Puncture and Epidural Catheterization Failure

Clinical question: Does ultrasound imaging for lumbar puncture and epidural catheterization reduce risk of procedure failures?

Background: Numerous studies have compared ultrasound-guided lumbar puncture and epidural catheterization with standard techniques, but they have been underpowered to evaluate whether ultrasound use reduces lumbar puncture and epidural catheterization failure.

Study design: Systematic review and meta-analysis of randomized controlled trials.

Setting: Hospitals in North America, Europe, and Asia.

Synopsis: Among 14 trials, a total of 1,334 patients (including one pediatric study group) were randomly assigned either to receive ultrasound imaging or to a control group (using manual palpation or loss of resistance). Ultrasound imaging comprised a preprocedure marking approach, real-time visual guidance, or both. In the 12 trials in which the primary outcome was available, ultrasound imaging significantly reduced the risk of failed procedures with a risk ratio of 0.21 (95% confidence interval 0.10 to 0.43). A total of 16 ultrasound-guided procedures were needed to avoid one procedure failure. Ultrasound guidance also reduced the number of traumatic procedures, needle reinsertions, and needle redirections. These results were consistent across multiple subgroup analyses.

Only one of the studies included in the meta-analysis was double-blind, and in all of the studies, ultrasound imaging was performed by a clinician with high-level ultrasound experience. While this study suggests that ultrasound does reduce the frequency of procedure failure, the investigators did not include cost-effectiveness analyses.

Bottom line: Ultrasound imaging likely reduces the failure of lumbar puncture and epidural catheterization, but future research will be necessary to evaluate the cost-effectiveness of ultrasound use for these procedures.

Citation: Shaikh F, Brzezinski J, Alexander S, et al. Ultrasound imaging for lumbar punctures and epidural catheterisations: systematic review and meta-analysis. BMJ. 2013;346:f1720.

Surgical Complications might Be Financially Advantageous for Hospitals

Clinical question: What is the impact of surgical complications on hospital finances?

Background: Surgical complications are common and lead to longer lengths of stay and higher costs. Strategies are available to reduce postsurgical complications but have not been universally adopted.

Study design: Observational study.

Setting: Twelve hospitals in one nonprofit Southern hospital system, which includes academic, nonacademic, and rural settings.

Synopsis: Researchers identified 34,526 patients who underwent surgery in 2010, excluding patients undergoing Caesarean section. Of those, 1,820 procedures (5.3%) were associated with at least one complication. The most frequent complications were surgical-site infection, other infections, pneumonia, and thromboembolic disease. The mortality rate for patients with complications was 12.3% compared with 0.6% for those without. Length of stay was four times longer for patients with complications.

Complications were associated with a higher total cost of hospitalization, with a differential of $37,917. This translated into a higher contribution to the margin. The cost differential varied by insurance type, with higher contributions under Medicare and private insurance but not with Medicaid.

The study had the benefit of using a large administrative database; however, this may have underestimated the actual rate of postoperative complications. The study supports the paradox in which quality-improvement (QI) programs that reduce surgical complications and improve postoperative mortality may negatively affect a hospital’s financial performance.

Bottom line: Surgical complications lead to higher mortality for patients but a financial benefit for hospitals.

Citation: Eappen S, Lane BH, Rosenberg B, et al. Relationship between occurrence of surgical complications and hospital finances. JAMA. 2013;309(5):1509-1606.

Showing Lab Test Fees May Decrease Frequency Of Test Ordering

Clinical question: Does information on the cost of a lab test lead to lowered ordering frequency among internal-medicine residents?

Background: Lab test overuse is common; some studies estimate that 70% of lab tests do not affect care. Strategies to reduce frequency of unnecessary lab tests are needed.

Study design: Randomized controlled trial.

Setting: Johns Hopkins Hospital, a 1,051-bed academic medical center in Baltimore.

Synopsis: Researchers used an administrative database to identify the 35 most frequently ordered and the 35 most expensive tests (each ordered at least 50 times). They randomized tests to an active arm, which displayed the Medicare allowable fee at the time of order entry within the computerized physician order entry, and a control arm. A total of 1,166,753 tests were ordered during the baseline and intervention period. Many more tests were ordered in the active group relative to the control group, a consequence of the randomization process. Relative to a six-month baseline period, tests in the active group were ordered 9.1% less frequently; control-group tests were ordered 5.1% more frequently. Charges decreased by $3.79 per patient-day in the active group and increased by $0.52 per patient-day in the control group.

This study reflects a low-cost strategy to reduce lab testing and associated costs. It is unknown whether only unnecessary tests were averted, or if there was any effect on the quality of care. The durability of the intervention and its applicability to other settings and with other types of providers is unclear.

Bottom line: Showing the fee associated with lab tests may decrease the frequency of ordering these tests and the resultant costs.

Citation: Feldman LS, Shihab HM, Thiemann D, et al. Impact of providing fee data on laboratory test ordering: a controlled clinical trial. JAMA Intern Med. 2013;173(10):903-908.

Inpatient Elderly Multidisciplinary-Team-Based Unit Reduces Readmissions

Clinical question: Can an inpatient unit that uses interdisciplinary rounds and a team approach reduce 30-day readmissions and lower cost compared with usual care?

Background: Elderly patients are at risk for adverse outcomes including delirium, deconditioning, and undernutrition. The Acute Care for Elders (ACE) model is designed as a team-based approach to improve communication between clinicians and to begin discharge planning soon after admission. ACE units have been shown to improve specific outcomes in older adults, but the impact on cost is less clear.

Study design: Retrospective cohort study.

Setting: University of Alabama at Birmingham Acute Care for Elders unit.

Synopsis: The analysis of cost data extrapolated from administrative sources indicated an average of $371 in variable cost savings per patient, and therefore an expected $148,400 savings for 400 patients admitted to this unit. The rate of 30-day readmissions was significantly lower in the ACE model: 7.9% versus 12.8% in patients receiving usual care.

Limitations of this study relate both to its design and single-center location. It is unclear which aspect of this ACE unit was helpful in the studied outcomes, and how this specific program would be reproduced at another institution. There were also some costs associated with the ACE-unit staffing, which may have resulted in an overestimation of the cost savings.

Bottom line: ACE units appear to improve outcomes and may lower cost, but further investigation is needed.

Citation: Flood KL, MacLennan PA, McGrew D, Green D, Dodd C, Brown CJ. Effects of an acute care for elders unit on costs and 30-day readmissions. JAMA. 2013 April 22 [Epub ahead of print].

Continue Beta-Blockers in High-Risk Patients During Noncardiac Surgery

Clinical question: Does exposure to a beta-blocker at the time of noncardiac surgery have an association with mortality, cardiac arrest, or Q-wave myocardial infarction (MI)?

Background: Exposure to beta-blockers perioperatively is controversial. The authors suggest that a randomized trial would be difficult, as one might need to withhold beta-blockers from stable patients. The American Heart Association and American College of Cardiology currently recommend continuation of pre-existing beta-blockades perioperatively.

Study design: Retrospective, propensity-matched cohort analysis.

Setting: One hundred four Veterans Affairs (VA) medical centers from January 2005 to August 2010.

Synopsis: Exposure for any reason to beta-blockers around nonvascular surgery appears to lower the risk of mortality and cardiac arrest, particularly in patients with higher revised cardiac risk index factors. Stopping beta-blockers in this same period was associated with increased mortality. The reasons for beta-blocker use in the exposed cohort were not well-established or -stratified.

Limitations include that the exposed cohort had a higher rate of cardiovascular disease and comorbidities and were generally older. Duration of exposure to beta-blockers was mixed, although more than 75% had a prescription for more than three months, with less than 2% for several days. Some of the matched cohort appears to have had beta-blocker exposure, thus diminishing the potential impact.

Bottom line: In this retrospective analysis, there is an association between beta-blocker use during noncardiac, nonvascular surgery and lower mortality, as well as lower rates of cardiac arrest and Q-wave MI, particularly among higher-risk patients.

Citation: London MJ, Hur K, Schwartz G, Henderson WG. Association of perioperative beta-blockade with mortality and cardiovascular morbidity following major noncardiac surgery. JAMA. 2013;309(16):1704-1713.

Patient-Centered Decision-Making and Health-Care Outcomes

Clinical question: Does recognition of contextual domains in the care plan lead to resolution of patients’ presenting problems?

Background: Patient-centered decision-making, or contextualization of care, adapts best evidence to the care of the individual patient. Examples of contextual domains include access to care, social support, and financial situation. Contextual errors (e.g. unrecognized domains) are, on average, more costly than errors in evidenced-based medicine.

Study design: Observational study.

Setting: VA ambulatory-care centers.

Synopsis: A total of 548 patient-derived audio recordings of physician encounters that included contextual red flags (e.g. missed appointments, HbA1c >8%) were reviewed at two VA hospital ambulatory-care centers using the content coding for contextualization of care (4C) method. Prospectively determined good and poor outcomes were derived from specific red flags (e.g. keeping next appointment, decrease in HbA1c). Of the 548 red flags, 208 were associated with contextual domains using the 4C method. Some 59% of physicians recognized contextual domains in care-plan development, leading to good outcomes in 71% of red flags. As many as 41% of physicians did not recognize contextual domains, leading to poor outcomes in 54% of red-flag instances.

Hospitalists should be aware of contextual domains and red flags (e.g. readmissions), and this study provides a method of evaluating patient-centered decision-making in the hospital setting. However, the inherently subjective 4C method may underestimate the number of contextual domains.

Bottom line: Recognition and incorporation of contextual domains in care-plan development in the ambulatory setting are associated with improved contextual red flag outcomes.

Citation: Weiner SJ, Schwartz A, Sharma G, et al. Patient-centered decision making and health care outcomes: an observational study. Ann Intern Med. 2013;158:573-579.

Adverse Surgical Outcomes in Schizophrenia Patients

Clinical question: What is the full spectrum of postoperative complications and mortality in schizophrenic patients receiving in-hospital major surgery?

Background: Schizophrenia affects an estimated 4 to 7 per 1,000 persons worldwide, and patients with schizophrenia are at increased risk of diabetes, cardiovascular disease, and stroke with subsequent increased mortality risk. The correlation between severity of mental illness and postoperative in-hospital mortality has not been validated in previous studies.

Study design: Population-based, cross-sectional study.

Setting: Taiwan hospitals.

Synopsis: Using the Taiwan National Health Insurance Research Database, researchers examined claims from 2004 to 2007 and retrospectively identified 8,967 schizophrenic patients who underwent major inpatient surgery and were hospitalized for more than one day. Primary outcomes included acute myocardial infarction, acute renal failure, stroke, and in-hospital mortality within 30 days. Postoperative complications and mortality rates were compared between schizophrenic patients and patients without mental illness.

Schizophrenic patients had higher rates of many primary outcomes, including 30-day postoperative mortality, compared with patients without mental illness, after adjusting for sex, age, surgery type, and hospital setting. The risk for 30-day mortality rose with the number of preoperative schizophrenia-related services provided. Limitations include the nature of retrospective analysis and generalizability.

Bottom line: Compared to patients without mental illness, schizophrenic patients have an increased risk of acute renal failure, pneumonia, septicemia, and 30-day mortality in the postoperative setting, with higher mortality rates in schizophrenic patients with more severe disease.

Citation: Liao CC, Shen WW, Chang CC, Chang H, Chen T. Surgical adverse outcomes in patients with schizophrenia. Ann Surg. 2013;257:433-438.