must read
must read
Al mentioned this article back in April 2005. Here are the details. In brief, over a 12 month period 197 patients presented with cardioembolic cerebral infarction at Cedars-Sinai Medical Center and UCLA Medical Center. 14 of the 197 strokes were in patients that had there anticoagulation stopped for low bleed risk procedures. Way to high don't you think.
I had an email conversation with Dr. Cohen and he really understands the problem.
Original Paper
Withdrawal of Warfarin prior to a Surgical Procedure: Time to Follow the Guidelines?
Sergey E. Akopova, Shuichi Suzukib, Andre Fredieub, Chelsea S. Kidwellb, Jeffrey L. Saverb, Stanley N. Cohen
Division of Neurology, Cedars-Sinai Medical Center Los Angeles,
Department of Neurology, University of California, Los Angeles, Calif., USA
Introduction
Continuous oral anticoagulant therapy with warfarin has been used for over 40 years to decrease the risk for thromboembolism. Currently more than 1 million patients in the United States are taking daily warfarin. Patients on warfarin frequently need modifications of their anticoagulation regimen prior to, during, and following invasive procedures. In such situations, clinicians must assess the risk of bleeding from a procedure if anticoagulation is continued versus the risk of thromboembolism if anticoagulation therapy is discontinued.
To avoid these potential complications, several alternative periprocedural anticoagulation strategies have been proposed; however, each of these techniques may be problematic [1]. Use of intravenous unfractionated heparin (UH) as a bridging therapy is an option but gener ally requires increased length of hospitalization. Use of subcutaneous low-molecular-weight heparin (LMWH), which may be administered on an outpatient basis, is an additional potential bridging strategy but is associated with increased cost and logistic difficulties in arranging twice daily injections. Some physicians have attempted to shorten the period of modified anticoagulation regimens by adding a small dose of vitamin K to accelerate cessation of warfarin effects. However, this approach may carry an increased risk for transient under- or over-anticoagulation and may lead to postprocedural temporary warfarin resistance.
Given these limitations of bridging parenteral anticoagulation strategies, the option to hold warfarin 4-5 days before the procedure and not use any transitional anticoagulant until the postoperative period continues to warrant consideration, given the relatively brief duration of the period during which the patient will be unprotected by active anticoagulation. In fact, current recommendations from the American Heart Association and the American College of Chest Physicians [2] suggest this option for patients with low risk for thromboembolism, e.g. patients without venous thrombosis for >3 months or patients with atrial fibrillation who do not have a history of cerebral infarction. These national guidelines do recommend bridging anticoagulation therapies for patients at high thromboembolic risk. These national recommendations are not based on prospective randomized trials, as reflected in the low grade given for the strength of the evidence on which they are advanced. In the absence of unequivocal recommendations, many clinicians continue routinely to withdraw warfarin without substituting an alternative anticoagulant therapy in patients requiring an invasive procedure. The purpose of this study was to determine frequency of cardioembolic cerebral infarction during periprocedural warfarin withdrawal to evaluate the risk of temporary warfarin withdrawal.
Methods
The study was designed as a retrospective analysis of prospectively gathered data on patients admitted with a presumed diagnosis of cerebral infarction evaluated by the Cerebral Infarction Services at Cedars-Sinai Medical Center and UCLA Medical Center within the 12-month epoch. Patients suffering cardioembolic cerebral infarction were identified based on history, neurological examination, and imaging studies and application of a modified version of the TOAST cerebral infarction subtype criteria [3]. Charts of all cardioembolic cerebral infarction patients were reviewed to identify those who developed cerebral infarction after temporary discontinuation of warfarin for a medical procedure.
Results
During the 12-month study period, 197 patients presented with cardioembolic cerebral infarction. In 14 (7.1%), the cerebrovascular event occurred after discontinuation of warfarin for a medical procedure. The mean age of these 14 patients was 75 years (range 54-91), 8 patients were male. Each had been on chronic anticoagulation with warfarin for more than 1 year. Table 1 outlines characteristics of patients who experienced a cerebrovascular event during temporary warfarin discontinuation. The reasons for anticoagulation included chronic atri al fibrillation (6 cases), paroxysmal atrial fibrillation (6 cases), patent foramen ovale (1), and mitral valve endocardial fibroelastosis (1). The majority of the patients (78%) were at high risk for thromboembolic events because of previous history of cardioembolic cerebral infarctions. The admission international normalized ratio (INR) was subtherapeutic in 93%, with average INR 1.4, range 1.0-2.0 (with therapeutic range for INR considered in the range 2-3). In 3 cases, the focal neurologic deficit totally resolved. In 11, the cerebral infarction resulted in residual deficits, with disabled status (Rankin 3-5) at discharge. No patient suffered a fatal cerebral infarction. Representative images from a case are shown in figure 1.
Warfarin was stopped because of dental procedures (1 case), dermatological procedures (2 cases), cardiologic procedures (2 cases), hernia repair (2 cases), endoscopy (2 cases), laminectomy (2 cases), lithotripsy (1 case), cataract surgery (1 case), and epidural injections (1 case). Patients had generally been instructed to temporarily discontinue their warfarin 3-7 days prior to their procedure. No substitutive anticoagulation therapy was used in any of these patients during the planned warfarin holiday. Across all patients, cerebral infarctions developed an average of 5.4 days after the last dose of warfarin (range 3-8; fig. 2). Only in 1 case did the cerebral infarction occur prior to the planned procedure, in the 13 other cases it occurred following the operative procedure, an average of 1.9 days after the procedure itself (range 0-4). In 1 case warfarin was stopped prior to cataract surgery and not restarted after the procedure because the patient was scheduled for a tooth extraction. In 3 patients, warfarin had already been restarted after the procedure, 1-2 days prior to cerebral infarction, but without therapeutic anticoagulation achieved at the time of the event.
Discussion
This case series suggests that cardiogenic cerebral infarction related to elective discontinuation of oral anticoagulation therapy for surgical procedures is a frequent event and that the majority of these cerebral infarctions are debilitating. In our experience, approximately 7% of cardioembolic cerebral infarctions were iatrogenic, due to discontinuation of warfarin in patients with an established potential cardiac source of embolism. Cerebral infarction associated with temporary discontinuation of warfarin for a medical procedure is a potentially preventable cause of cerebral infarction if recommendations regarding appropriate management of anticoagulation during invasive procedures are carefully followed.
Decisions about discontinuation of warfarin for a procedure should be based on rational evaluation of the risks of hemorrhage during a procedure if warfarin is continued compared to the risk of emboli as a result of warfarin withdrawal. Among our patients, the majority had warfarin discontinued because of dental, dermatological, ophthalmologic and endoscopic procedures rather than major cardiothoracic, abdominal, or neurosurgical interventions.
Recent analyses of the rate of bleeding complications associated with these procedures when performed in the presence of therapeutic warfarin anticoagulation suggest very low risks for hemorrhages. Evaluations of bleeding risks as a result of dental procedures in patients on continuous anticoagulation therapy demonstrated a very low hemorrhagic complication rate for routine dental procedures and uncomplicated extractions [4,5,6]. The adverse consequences of the mostly theoretical risk of hemorrhage after a dental procedure, which may be controlled with local measures should it occur at all, are much lower than those of a potentially life-threatening thromboembolic event. Even more complex oral surgical procedures may be performed without changes in intensity of anticoagulation therapy [7]. If a selected dental procedure is considered too dangerous to perform on a full dose of warfarin, the dose may be lowered or stopped for no more than 2 days with minimal INR decreases [8]. In our patient, warfarin was stopped 4 days before procedure and restarted 1 day after, when the patient had already developed a small cerebral infarction.
Analogously, simple skin operations carry very low risk for hemorrhagic complications when performed in chronically anticoagulated patients. Overall hemorrhagic complications occur in 1.6% of cases and are not significantly more frequent compared with complications in control subjects [9]. Furthermore, there is no statistically significant reduction in the rates of severe complications in patients who have their medications held preoperatively. Recent reviews documented the extremely low risk for bleeding complications after dermatological procedures in chronically anticoagulated patients [10]. The surveying physicians, who perform dermatologic surgery, came to the conclusion that there is a 'compelling argument to maintain patients on medically necessary blood thinners during a cutaneous operation' [11]. In a recent detailed review [12], low risk for hemorrhagic complications was confirmed for dermatologic procedures, cataract surgery and endoscopy performed in patients on chronic anticoagulation without warfarin withdrawal.
In situations when bleeding risk during invasive procedures is estimated as too high without the withdrawal of anticoagulants, other anticoagulation options should be explored. The most current recommendations of the American College of Chest Physicians [2] suggest use of substitutive therapy in patients with high risk of thromboembolism (e.g. patients with a recent history of venous thromboembolism, a mechanical cardiac valve in the mitral position or an old model of cardiac valve, or patients with atrial fibrillation and a history of embolic cerebral infarctions) during temporary warfarin withdrawal, including the options of intravenous UH or subcutaneous LMWH. In several small prospective studies, bridging LMWH, including in the outpatient setting, has been shown to prevent thromboembolic events in all patients with a very low (<1%) incidence of hemorrhagic complications [12,13,14]. The majority of our patients were in a high-risk group because of a history of previous cardioembolic cerebral infarctions; however, no substitutive therapy was used. Bridging LMHW is a generally safe and effective strategy in patients with cerebral infarction, event when the infarcts are of recent onset [15].
There are several limitations to this study. Due to its retrospective nature, data comparing cerebral infarction rates in patients treated with bridging therapy to those without are not available. It is also possible that rates of thromboembolic complications in patients not receiving bridging therapy is higher than recorded in our cohort, as we identified only patients with cerebrovascular complications. It is possible that a substantial number of patients may develop other systemic embolic events as well in this scenario.
In conclusion, temporary discontinuation of warfarin for invasive procedures in patients with established high-risk for cardioembolic cerebral infarction may lead to a devastating cerebral infarction. These events constitute an unacceptably large percentage of hospital admissions for cardioembolic cerebral infarctions. Treatment following existing guidelines for the management of anticoagulation therapy in high-risk patients prior to an invasive procedure would have prevented cerebral infarction in virtually all our cases. Stricter adherence to these guidelines is recommended to avoid potentially catastrophic consequences of cardioembolic cerebral infarctions provoked by warfarin withdrawal in chronically anticoagulated patients.
THE INR LEVELS FOR THE 14 PATIENTS WERE AS FOLLOWS WHEN ADMITTED WITH STROKE SYMPTOMS
1. 2.0
2. 1.5
3. 1.2
4. 1.6
5. 1.0
6. 1.3
7. 1.1
8. 1.2
9. 1.5
10. 1.1
11. 1.3
12. 1.6
13. 1.1
14. 1.1
must read
Al mentioned this article back in April 2005. Here are the details. In brief, over a 12 month period 197 patients presented with cardioembolic cerebral infarction at Cedars-Sinai Medical Center and UCLA Medical Center. 14 of the 197 strokes were in patients that had there anticoagulation stopped for low bleed risk procedures. Way to high don't you think.
I had an email conversation with Dr. Cohen and he really understands the problem.
Original Paper
Withdrawal of Warfarin prior to a Surgical Procedure: Time to Follow the Guidelines?
Sergey E. Akopova, Shuichi Suzukib, Andre Fredieub, Chelsea S. Kidwellb, Jeffrey L. Saverb, Stanley N. Cohen
Division of Neurology, Cedars-Sinai Medical Center Los Angeles,
Department of Neurology, University of California, Los Angeles, Calif., USA
Introduction
Continuous oral anticoagulant therapy with warfarin has been used for over 40 years to decrease the risk for thromboembolism. Currently more than 1 million patients in the United States are taking daily warfarin. Patients on warfarin frequently need modifications of their anticoagulation regimen prior to, during, and following invasive procedures. In such situations, clinicians must assess the risk of bleeding from a procedure if anticoagulation is continued versus the risk of thromboembolism if anticoagulation therapy is discontinued.
To avoid these potential complications, several alternative periprocedural anticoagulation strategies have been proposed; however, each of these techniques may be problematic [1]. Use of intravenous unfractionated heparin (UH) as a bridging therapy is an option but gener ally requires increased length of hospitalization. Use of subcutaneous low-molecular-weight heparin (LMWH), which may be administered on an outpatient basis, is an additional potential bridging strategy but is associated with increased cost and logistic difficulties in arranging twice daily injections. Some physicians have attempted to shorten the period of modified anticoagulation regimens by adding a small dose of vitamin K to accelerate cessation of warfarin effects. However, this approach may carry an increased risk for transient under- or over-anticoagulation and may lead to postprocedural temporary warfarin resistance.
Given these limitations of bridging parenteral anticoagulation strategies, the option to hold warfarin 4-5 days before the procedure and not use any transitional anticoagulant until the postoperative period continues to warrant consideration, given the relatively brief duration of the period during which the patient will be unprotected by active anticoagulation. In fact, current recommendations from the American Heart Association and the American College of Chest Physicians [2] suggest this option for patients with low risk for thromboembolism, e.g. patients without venous thrombosis for >3 months or patients with atrial fibrillation who do not have a history of cerebral infarction. These national guidelines do recommend bridging anticoagulation therapies for patients at high thromboembolic risk. These national recommendations are not based on prospective randomized trials, as reflected in the low grade given for the strength of the evidence on which they are advanced. In the absence of unequivocal recommendations, many clinicians continue routinely to withdraw warfarin without substituting an alternative anticoagulant therapy in patients requiring an invasive procedure. The purpose of this study was to determine frequency of cardioembolic cerebral infarction during periprocedural warfarin withdrawal to evaluate the risk of temporary warfarin withdrawal.
Methods
The study was designed as a retrospective analysis of prospectively gathered data on patients admitted with a presumed diagnosis of cerebral infarction evaluated by the Cerebral Infarction Services at Cedars-Sinai Medical Center and UCLA Medical Center within the 12-month epoch. Patients suffering cardioembolic cerebral infarction were identified based on history, neurological examination, and imaging studies and application of a modified version of the TOAST cerebral infarction subtype criteria [3]. Charts of all cardioembolic cerebral infarction patients were reviewed to identify those who developed cerebral infarction after temporary discontinuation of warfarin for a medical procedure.
Results
During the 12-month study period, 197 patients presented with cardioembolic cerebral infarction. In 14 (7.1%), the cerebrovascular event occurred after discontinuation of warfarin for a medical procedure. The mean age of these 14 patients was 75 years (range 54-91), 8 patients were male. Each had been on chronic anticoagulation with warfarin for more than 1 year. Table 1 outlines characteristics of patients who experienced a cerebrovascular event during temporary warfarin discontinuation. The reasons for anticoagulation included chronic atri al fibrillation (6 cases), paroxysmal atrial fibrillation (6 cases), patent foramen ovale (1), and mitral valve endocardial fibroelastosis (1). The majority of the patients (78%) were at high risk for thromboembolic events because of previous history of cardioembolic cerebral infarctions. The admission international normalized ratio (INR) was subtherapeutic in 93%, with average INR 1.4, range 1.0-2.0 (with therapeutic range for INR considered in the range 2-3). In 3 cases, the focal neurologic deficit totally resolved. In 11, the cerebral infarction resulted in residual deficits, with disabled status (Rankin 3-5) at discharge. No patient suffered a fatal cerebral infarction. Representative images from a case are shown in figure 1.
Warfarin was stopped because of dental procedures (1 case), dermatological procedures (2 cases), cardiologic procedures (2 cases), hernia repair (2 cases), endoscopy (2 cases), laminectomy (2 cases), lithotripsy (1 case), cataract surgery (1 case), and epidural injections (1 case). Patients had generally been instructed to temporarily discontinue their warfarin 3-7 days prior to their procedure. No substitutive anticoagulation therapy was used in any of these patients during the planned warfarin holiday. Across all patients, cerebral infarctions developed an average of 5.4 days after the last dose of warfarin (range 3-8; fig. 2). Only in 1 case did the cerebral infarction occur prior to the planned procedure, in the 13 other cases it occurred following the operative procedure, an average of 1.9 days after the procedure itself (range 0-4). In 1 case warfarin was stopped prior to cataract surgery and not restarted after the procedure because the patient was scheduled for a tooth extraction. In 3 patients, warfarin had already been restarted after the procedure, 1-2 days prior to cerebral infarction, but without therapeutic anticoagulation achieved at the time of the event.
Discussion
This case series suggests that cardiogenic cerebral infarction related to elective discontinuation of oral anticoagulation therapy for surgical procedures is a frequent event and that the majority of these cerebral infarctions are debilitating. In our experience, approximately 7% of cardioembolic cerebral infarctions were iatrogenic, due to discontinuation of warfarin in patients with an established potential cardiac source of embolism. Cerebral infarction associated with temporary discontinuation of warfarin for a medical procedure is a potentially preventable cause of cerebral infarction if recommendations regarding appropriate management of anticoagulation during invasive procedures are carefully followed.
Decisions about discontinuation of warfarin for a procedure should be based on rational evaluation of the risks of hemorrhage during a procedure if warfarin is continued compared to the risk of emboli as a result of warfarin withdrawal. Among our patients, the majority had warfarin discontinued because of dental, dermatological, ophthalmologic and endoscopic procedures rather than major cardiothoracic, abdominal, or neurosurgical interventions.
Recent analyses of the rate of bleeding complications associated with these procedures when performed in the presence of therapeutic warfarin anticoagulation suggest very low risks for hemorrhages. Evaluations of bleeding risks as a result of dental procedures in patients on continuous anticoagulation therapy demonstrated a very low hemorrhagic complication rate for routine dental procedures and uncomplicated extractions [4,5,6]. The adverse consequences of the mostly theoretical risk of hemorrhage after a dental procedure, which may be controlled with local measures should it occur at all, are much lower than those of a potentially life-threatening thromboembolic event. Even more complex oral surgical procedures may be performed without changes in intensity of anticoagulation therapy [7]. If a selected dental procedure is considered too dangerous to perform on a full dose of warfarin, the dose may be lowered or stopped for no more than 2 days with minimal INR decreases [8]. In our patient, warfarin was stopped 4 days before procedure and restarted 1 day after, when the patient had already developed a small cerebral infarction.
Analogously, simple skin operations carry very low risk for hemorrhagic complications when performed in chronically anticoagulated patients. Overall hemorrhagic complications occur in 1.6% of cases and are not significantly more frequent compared with complications in control subjects [9]. Furthermore, there is no statistically significant reduction in the rates of severe complications in patients who have their medications held preoperatively. Recent reviews documented the extremely low risk for bleeding complications after dermatological procedures in chronically anticoagulated patients [10]. The surveying physicians, who perform dermatologic surgery, came to the conclusion that there is a 'compelling argument to maintain patients on medically necessary blood thinners during a cutaneous operation' [11]. In a recent detailed review [12], low risk for hemorrhagic complications was confirmed for dermatologic procedures, cataract surgery and endoscopy performed in patients on chronic anticoagulation without warfarin withdrawal.
In situations when bleeding risk during invasive procedures is estimated as too high without the withdrawal of anticoagulants, other anticoagulation options should be explored. The most current recommendations of the American College of Chest Physicians [2] suggest use of substitutive therapy in patients with high risk of thromboembolism (e.g. patients with a recent history of venous thromboembolism, a mechanical cardiac valve in the mitral position or an old model of cardiac valve, or patients with atrial fibrillation and a history of embolic cerebral infarctions) during temporary warfarin withdrawal, including the options of intravenous UH or subcutaneous LMWH. In several small prospective studies, bridging LMWH, including in the outpatient setting, has been shown to prevent thromboembolic events in all patients with a very low (<1%) incidence of hemorrhagic complications [12,13,14]. The majority of our patients were in a high-risk group because of a history of previous cardioembolic cerebral infarctions; however, no substitutive therapy was used. Bridging LMHW is a generally safe and effective strategy in patients with cerebral infarction, event when the infarcts are of recent onset [15].
There are several limitations to this study. Due to its retrospective nature, data comparing cerebral infarction rates in patients treated with bridging therapy to those without are not available. It is also possible that rates of thromboembolic complications in patients not receiving bridging therapy is higher than recorded in our cohort, as we identified only patients with cerebrovascular complications. It is possible that a substantial number of patients may develop other systemic embolic events as well in this scenario.
In conclusion, temporary discontinuation of warfarin for invasive procedures in patients with established high-risk for cardioembolic cerebral infarction may lead to a devastating cerebral infarction. These events constitute an unacceptably large percentage of hospital admissions for cardioembolic cerebral infarctions. Treatment following existing guidelines for the management of anticoagulation therapy in high-risk patients prior to an invasive procedure would have prevented cerebral infarction in virtually all our cases. Stricter adherence to these guidelines is recommended to avoid potentially catastrophic consequences of cardioembolic cerebral infarctions provoked by warfarin withdrawal in chronically anticoagulated patients.
THE INR LEVELS FOR THE 14 PATIENTS WERE AS FOLLOWS WHEN ADMITTED WITH STROKE SYMPTOMS
1. 2.0
2. 1.5
3. 1.2
4. 1.6
5. 1.0
6. 1.3
7. 1.1
8. 1.2
9. 1.5
10. 1.1
11. 1.3
12. 1.6
13. 1.1
14. 1.1
