An Outpatient Anticoagulation Protocol Managed by a Vascular Nurse-Clinician Gary R. Seabrook, MD, Donna Karp, RN, BSN, David D. Schmitt, MD, Dennis F. Bandyk, MD, Jonathan B. Zowne, MD, Milwaukee,Wisconsin
Lifetime anticoagulation has become a therapeutic option for surgical patients with hypercoagulable states or prosthetic arterial bypass grafts. However, physicians may not achieve optimal anticoagulation or may attempt to limit the length of the therapy period because of the perceived morbidity from hemorrhagic complications of Coumadin therapy. A protocol for anticoagulant therapy monitored and regulated by a vascular nurse-clinician was reviewed. Coumadin was prescribed for 1,891 pa, tient-months to 93 patients to maintain their prothrombin time 1.5 to 2 times control (range: 18 to 24 seconds). The mean ('4- SD) prothrombin time for the study population was 19.8 + 1.8 seconds. During follow-up, 472 (14%) of 3,479 prothrombin times measured were below the therapeutic range (n = 232) or prolonged (n = 240), prompting an adjustment in the Coumadin dose in 82 (88%) patients. Four patients developed recurrent vascular graft thrombosis while receiving anticoagulation. There were 6 major and 11 minor hemorrhagic complications. Patients with a chronic risk for arterial or venous thrombosis can have outpatient anticoagulant therapy administered at optimal intensity and regulated safely with a low incidence of hemorrhagic and thrombotic events.
ifetime anticoagulation has become a therapeutic option for patients at risk for arterial or venous thromL bosis due to hypercoagulable states, prosthetic infrainguinal arterial bypasses, arterial emboli, or recurrent thrombophlebitis. Coumadin (crystalline warfarin sodium, USP) is accepted as the optimal form of outpatient anticoagulation for prophylactically protecting the venous circulation in patients with deep venous thrombosis and/or pulmonary emboli [I-3]. There has been a more limited experience with anticoagulation for patients at risk for arterial thrombosis. From the Department of Surgery, Medical College of Wisconsin, Milwaukee, Wisconsin. Requests for reprints should be addressed to Gary R. Seahrook, MD, Department of Surgery, 8700 W. Wisconsin Avenue, Milwaukee, Wisconsin 53226. Manuscript submitted September 26, 1989, and accepted in revised form December 1, 1989.
We have perceived reluctance on the part of our colleagues to prescribe long-term oral anticoagulant therapy. In view of the significant reported morbidity associated with hemorrhagic complications in patients receiving Coumadin therapy, physicians have attempted to limit the length of the therapy period and tend to be conservative in achieving optimal anticoagulation [4,5]. Patients for whom such therapy is often considered contraindicated include those with concurrent medical problems requiring multiple medications, elderly patients, and patients who travel extensively, consume alcohol regularly, or are engaged in strenuous physical activity [6]. To demonstrate that Coumadin can be prescribed safely for extended periods of time, we reviewed a protocol administered by a vascular nurse-clinician and supervised by a physician for monitoring and regulating outpatient anticoagulant therapy on an academic vascular surgery service. Patients maintained on Coumadin included those who had a variety of concomitant medical conditions including hypertension, coronary artery disease, and insulin-dependent diabetes mellitus; age into the ninth decade; and a habit of regular alcohol consumption. The efficacy and safety of the protocol were reviewed with respect to patient selection and maintenance of optimal intensity of anticoagulation versus the incidence of thrombotic or hemorrhagic complications. PATIENTS AND METHODS From 1981 to 1988, 93 patients admitted to an academic peripheral vascular surgery service were maintained on a long-term oral outpatient anticoagulant therapy program administered by a nurse-clinician. There were 56 men and 37 women whose ages ranged from 18 to 89 years (median: 63 years). Indications for therapy included hypercoagulable states in which a coagulation defect was documented, prosthetic infrainguinal arterial grafts, venous thrombosis, arterial emboli, and unexp!ained prosthetic graft thrombosis (Table I). All patients were hospitalized at the time the decision was made to prescribe long-term Coumadin therapy. Prior to discharge, the nurse-clinician reviewed the patient's medical record for demographic data, concurrent medical problems, medications, current Coumadin dose, and coagulation laboratory values and determined if he or she could understand the risks and implications of long-term anticoagulant therapy, report in a systematic fashion for laboratory monitoring of prothrombin time, and be able to adjust the Coumadin dose based on telephoned instructions. Patients excluded from participation in the program included those without a residence and/or telephone, those with uncontrolled mental illness, or those incapable of understanding the arithmetic necessary to alter the Coumadin dose based on verbal instructions.
THE AMERICAN JOURNAL OF SURGERY
VOLUME 160 NOVEMBER 1990
501
SEABROOK ET AL
TABLE I Indications for Long-Term AnUcoagulatlon Patients (n)
Hypercoagulable states Fibrinolytic disorders Antithrombin III deficiency Protein C/protein S deficiency Venous thrombosis Prosthetic infrainguinal arterial bypasses Arterial emboli Unexplained prosthetic graft occlusion
22 6 3 20 20 13 9
Patients accepted into the outpatient program were instructed prior to discharge on the basic pharmacology of anticoagulation and the risks associated with Coumadin therapy. Depending on the patient's level of education, the pathophysiologic process was explained so he or she understood the reason anticoagulation had been prescribed. The patient was taught that the dose of Coumadin might need to be changed from time to time and that monitoring of the prothrombin time would be necessary as long as the therapy was administered. The increased likelihood of minor bleeding (epistaxis, menorrhagia, gingival bleeding, hematuria, hematochezia) was explained, as well as warning signs and symptoms of a major bleeding episode that could occur when on anticoagulant therapy (dizziness, hemoptysis, hematemesis, melena, abdominal pain). The patient was advised not to take aspirin or compounds that contain salicylates while on Coumadin therapy. Excessive restraints on diet, activity, and lifestyle were not required. Patients were instructed to maintain a consistent diet and were permitted to drink alcohol in moderation, but were seriously warned against binge drinking or consuming large quantities of foods containing significant amounts of vitamin K. Patients were advised not to take vitamin supplements that contain vitamin K. The goal of therapy was to maintain a prothrombin time 1.5 to 2 times the control, resulting in a target prothrombin time between 18 and 24 secondS. Coumadin
I PROTHROMBIN TIME PROLONGED
1=
.ow oos
I
DECREASE OR HOLD DOSE
PROLONGED
CHANGE TO
THERAPEUTIC -~
RE
P.T.
1
RANGE
~L STABLE
I
SUBTHERAPEUTIC I
THERAPEUTIC PROTHROMBIN TIME [ T STABLE
INCREASE DOSE
509.
A
I
I PROTHROUSIN T ME
1=
was administered concomitantly with intravenous heparin for 5 days, beginning with a dose of 10 mg daily for 3 days. The dose was reduced to 7.5 mg on the fourth and fifth days to avoid an excessive early prolongation of the prothrombin time. The level of anticoagulation was measured daily using Quick's one-stage prothrombin time, employing a dried rabbit brain thromboplastin with calcium (American Dade, Aguada, Puerto Rico), which gave a standardized control of 12 seconds. Upon discharge from the hospital, the patient returned for measurement of prothrombin times one to three times a week until a stable dose had been achieved. The frequency of prothrombin time determinations was decreased to once monthly after the patient demonstrated a therapeutic prothrombin time on a set dose, although the frequency with which prothrombin times were measured varied depending upon the stability of the anticoagulation. The nurse-clinician contacted the patient by telephone to report that day's laboratory results and altered the daily Coumadin dose or created a dosing schedule in which the patient alternated between two doses on various days of the week. Five milligram tablets were routinely prescribed so the patient could be instructed to take one-half or multiple units of the tablets when dose changes were required. The patient was asked to consistently take the drug in the latter part of the day so that when instructions for dose changes or holding the dose were given, the patient could act on the telephoned message, which usually came in the afternoon. Each telephone contact provided the nurse-clinician with an opportunity to query the patient for evidence of bleeding complications and answer questions regarding the therapy program. Coumadin therapy was regulated during routine follow-up according to an algorithm (Figure 1). When a patient's prothrombin time was prolonged beyond the target range, the nurse-clinician directed the patient to hold or reduce the Coumadin dose and arranged for a follow-up prothrombin time to be obtained. When the prothrombin time returned to the desired level, the daily Coumadin dose was adjusted and monitored frequently until confirmation of a new stable dose was achieved. A prolonged prothrombin time prompted the nurse-clinician to care-
SUB-THERAPEUTIC
REPEAT P.T.
I
THE AMERICAN JOURNAL OF SURGERY
VOLUME 160
Figure 1. Algorithm for regulation of Coumadin therapy. Dose adjustments are increased or decreased in increments of 2.5 mg.
NOVEMBER 1990
OUTPATIENT ANTICOAGULATION PROTOCOL
fully interrogate the patient to ascertain the cause of the prolonged value and to identify any occult bleeding including the presence of petechia, ecchymosis, hematochezia, or hematuria. Patients with prolonged prothrombin times thought to be at risk for a bleeding complication were immediately hospitalized for observation. When the patient's prothrombin time was below the desired therapeutic range, he or she was instructed to increase the daily dose, and repeat prothrombin times were ordered to confirm that the desired level had been achieved. Subtherapeutic prothrombin times prompted the nurse-clinician to investigate the possibility of a venous or arterial thrombotic event. Records of each patient contact and the results of laboratory studies were maintained by the nurse-clinician with the information periodically being transcribed into the patient's permanent medical record. Patients found to be noncompliant with the protocol administered by the nurse-clinician were referred to the primary physician for management, or their therapy was discontinued if the risk of complications was thought to be excessive. Each patient in the nurse-clinician-managed program was seen at least every 6 months by the physician supervising the therapy to formally review the patient's progress, and all significant dose adjustments were reviewed with the physician at the time they were implemented. RESULTS The study reviewed 1,891 patient-months of Coumadin therapy (range: 1 to 88 months). During this time, 3,479 prothrombin times were measured, resulting in a mean of 2.6 prothrombin times per month. The mean 4SD of the prothrombin times for the study population was 19.8 4- 1.8 seconds. Eighty-two (88%) of the 93 patients required periodic alteration in the prescribed Coumadin dose after establishment of an initial stable dose. Sixtytwo (66%) of the patients were prescribed a dosing regimen of Coumadin alternating between two strengths of the drug from day to day. During the course of therapy, 472 (14%) of the 3,479 prothrombin times prompted an adjustment in the Coumadin dose; 232 were below the therapeutic range, requiring an increase in the daily Coumadin dose, and 240 were prolonged, resulting in a decrease in the patient's dose. The mean number of dose changes was 5.0 (range: 0 to 20). Twelve of the patients required more than 10 dose changes, accounting for 169 of 472 (36%) of the total number of dose changes. Eight octogenarians were prescribed 1i3 patient-months of therapy, during which 41 prothrombin times were measured. There were no complications in these elderly patients, and monitoring requirements (2.7 prothrombin times per month) mirrored that of the tota ! population. There were no deaths related to the anticoagulant therapy. There were no venous thrombotic events during this study. Four patients received Coumadin (duration of administration: 6 to 20 months) because antecedent arterial graft occlusion led to prosthetic vascular graft thrombosis while receiving anticoagulation. Three of these patients had therapeutic prothrombin times when their graft occluded. This represents an arterial graft throm-
botic rate of 8.3% for 48 grafts observed (3.7 events per 1,000 patient-months of observation). Etiologic factors contributing to these four thrombotic events included polycythemia, thrombocytosis, antithrombin III deficiency, and an axillo-femoral graft that became mechanically obstructed. Limb salvage was achieved following all of these graft occlusions. Six major hemorrhagic complications (defined as bleeding resulting in the need for hospitalization, blood transfusion, or cessation of anticoagulant therapy), occurring 12 to 42 months from commencement of therapy, resulted in a complication rate of 312 per 1,000 patientmonths of treatment. The indication for anticoagulation therapy was a plasminogen abnormality in three patients, recurrent prosthetic graft occlusion in two patients, and arterial embolization in one patient. The causes of hemorrhage included retroperitoneal bleeding, diverticulosis, ulcerative colitis, a duodenal ulcer, and two gastric ulcers. No gastrointestinal malignancies were discovered [7]. The patient with retroperitoneal bleeding was continued on Coumadin therapy because of a history of recurrent vascular graft thrombosis caused by a coagulation factor defect. Anticoagulation therapy was discontinued in the other five patients. Eleven minor hemorrhagic complications were identified including epistaxis, subcutaneous hematoma, hematuria, hyphema, menorrhagia, and gingival bleeding, representing a rate of 5.3 events per 1,000 patient-months of treatment. Each of these complications was self-limited and resulted in no significant morbidity. There was no evidence of Coumadin-related skin necrosis, which has been associated with high loading doses of the drug [8,9]. At the conclusion of this review, 56 of the 93 patients remained on Coumadin therapy. Eighteen patients were subsequently transferred back to referring physicians. The treatment plan was completed and Coumadin discontinued in 10 patients. Coumadin therapy was discontinued in five patients because of major gastrointestinal bleeding and in another patient diagnosed with a large pulmonary carcinoma. Moreover, in one patient, the drug was converted to subcutaneous heparin therapy during pregnancy. Two patients died of causes unrelated to their indications for oral anticoagulation. COMMENTS Patients requiring long-term, even lifetime, oral anticoagulation for protection against arterial or venous thrombosis can be safely managed in the outpatient setting. Administration of this protocol by a nurse-clinician enabled the patients to have carefully individualized and closely monitored therapy with appropriate education and ongoing surveillance--factors that are important in reducing the risk of complications. In contrast to many long-term medications, Coumadin has the potential for serious complications associated with significant morbidity (i.e., gastrointestinal, retroperitoneal, and intracerebral hemorrhage). Utilization of this protocol allowed patients to receive therapeutic anticoagulation with Coumadin with acceptable risks. To achieve the goals of anticoagulation in this patient
THE A M E R I C A N J O U R N A L O F S U R G E R Y
VOLUME 160
NOVEMBER 1990
503
SEABROOK ET AL
population, the criteria must be liberal for accepting patients into the therapy protocol. The patients in this Series came from an academic peripheral vascular surgery service, and their mean age was higher than that of series of patients receiving anticoagulation for deep venous thrombosis or pulmonary emboli [10]. All of the patients were taking other medications for treatment of diseases including hypertension, coronary artery disease, and diabetes mellitus. Patients receiving Coumadin therapy as prophylaxis against thrombosis of an autogenous or a prosthetic arterial bypass may benefit from anticoagulant therapy for the rest of their lives [11,12]. This study demonstrated that long-term therapy, extending in excess of 7 years, did not result in a higher complication rate for the total patient-months of therapy than in series of patients having considerably shorter courses of therapy [13-16]. Bleeding complications were distributed throughout the follow-up period and were not related to the length of the anticoagulant therapy. Coumadin, which is dose-dependent in its inhibition of factors II, VII, IX, and X, has a half-life of 2 89days; this facilitates use of a dosing schedule in which daily doses of different strengths (e.g., 5.0 mg alternating with 7.5 mg) were prescribed to maintain a stable level of anticoagulation [17]. Some patients were found to be sensitive to Coumadin and required as little as 2.5 mg every other day to achieve a therapeutic level. Heparin therapy has a slight effect on the prothrombin time, and this must be considered while therapy is being changed from heparin to Coumadin. Initial doses of Coumadin were never in excess of 10 mg to avoid the complication of Coumadinassociated skin necrosis. Prothrombin times were measured using standardized rabbit brain thromboplastin, which delivers a control prothrombin time of 12 seconds. The mean prothrombin time calculated for the study population was 19.8 seconds 4- SD 1.8 seconds (1.4 to 2.0 times control), which is an appropriate target range when rabbit brain thromboplastin is used as the reference. When human brain thromboplastin, which is more sensitive than rabbit brain thromboplastin, is used, higher ratios (2.0 to 2.5 times control) have been recommended [18]. The intensity of the degree of anticoagulation was very consistent in the study population. When a patient had a prothrombin time prolonged beyond the target range, a cause could usually be identified and was most frequently related to the use of an over-the-counter medication for a minor illness or an increase in alcohol consumption. A prothrombin time below the therapeutic range could usually be attributed to noncompliance with the dosage schedule or dietary indiscretions. There was considerable variability within the study group with respect to maintaining a stable prothrombin time on a given dose of Coumadin. The number of dose changes correlated with the number of months the patient had been administered Coumadin therapy. The 12 patients requiring more than 10 dose changes received therapy from 10 to 88 months (mean: 40 months). In several patients who were followed for prolonged periods of time, it was noted that after many months of consistent 504
therapy on a single dose regimen, a significant change in the required Coumadin dose would be necessary to maintain the prothrombin time in the therapeutic range. Despite careful questioning for changes in diet, activity, or medication and with the assurances that the patient was reliable and compliant, the cause of the required dose change could not always be established. For the major hemorrhagic complications occurring in this series, the patients' prothrombin times were within the target therapeutic range. No complications occurred when the patient had an excessively prolonged prothrombin time. Of the four patients who had recurrent arterial thrombotic events related to prosthetic arterial bypass grafts, one of the occlusions occurred when the patient's prothrombin time had fallen to a subtherapeutic level. In the other three patients, the prothrombin time was within the desired therapeutic range. However, each of these patients had an etiologic factor to which the occlusion could be ascribed: polycythemia, thrombocytosis, antithrombin III deficiency, and mechanical graft obstruction. Lifetime oral anticoagulation is a viable therapeutic option in treating patients at chronic risk for thrombosis. An outpatient protocol can be managed safely and effectively by a vascular nurse-clinician for an elderly population from an academic surgical service. Complications in this group did not exceed reported complications in other large series of outpatient anticoagulation regimens, many of which were for shorter periods of time in younger patient populations with fewer associated pathologic conditions. Hemorrhagic complications that occurred were not related to anticoagulated states exceeding the target therapeutic range. The nurse-clinician and patient relationships allowed successful management of Coumadin therapy and provided frequent contact and accessibility for the patient to the health care system. Long-term, even lifetime, oral anticoagulant therapy can be managed in the outpatient setting by an experienced registered nurseclinician with excellent patient compliance and a low frequency of hemorrhagic and thrombotic events. Patients with a legitimate chronic risk for arterial or venous thrombosis can have outpatient anticoagulant therapy administered and regulated with a low incidence of hemorrhagic and thrombotic events. REFERENCES 1, Hull R, Hirsh J, Jay R, et al. Differentintensitiesof anticoagula~ tion in the long-term treatment of proximalvenousthrombosis. N Engl J Med 1982; 307: 1676-81. 2. Hull R, Delmore T, Genton E, et al. Warfarin sodiumversus low-doseheparinin the long-termtreatment of venousthromboembolism. N Engl J Med 1979; 301: 855-8. 3. Hull R, Delmore T, Carter C, et at. Adjusted subcutaneous heparin versus warfarin sodiumin the long-termtreatment of venous thrombosis. N Engl J Med 1982; 306: 189-94. 4. MosleyDH, Schatz IJ, BrenemanGM, KeyesJW. Long-term anticoagulant therapy. JAMA 1983; 12: 128-30. 5. PollardJW, HamiltonMJ, ChristensenNA, AchorRWP. Problemsassociatedwith long-termanticoagulanttherapy:observations in 139 cases. Circulation 1962; 25: 311-7. 6. Davis FB, Estruch MT, Samson-CorveraEB, Voigt GC, Tobin
THE AMERICAN JOURNAL OF SURGERY VOLUME 160 NOVEMBER1990
OUTPATIENTANTICOAGULATIONPROTOCOl,
JD. Management of anticoagulation in outpatients. Arch Intern Med 1977; 137: 197-202. 7. Jaffin LW, Bliss M, Lamont T. Significance of occult gastrointestinal bleeding during anticoagulation therapy. Am J Med 1987; 83: 269-72. 8. Kazmier FJ. Thromboembolism, Coumadin necrosis, and protein C. Mayo Clin Proc 1985; 60: 673-4. 9. Peterson CE, Kwaan HC. Current concepts of warfarin therapy. Arch Intern Med 1986; 146: 581-4. 10. Gurwitz JH, Goldberg R J, Holden A, Knapic N, Ansell J. Agerelated risks of long-term oral anticoagulant therapy. Arch Intern Med 1988; 148: 1733-6. 11. Flinn WR, Rohrer M J, Yao JST, McCarthy WJ, Fahey VA, Bergan JJ. Improved long-term patency of infragenicular polytetrafluoroethylene grafts. J Vasc Surg 1988; 7: 685-90. 12. Kretschmer G, Wenzl E, Piza F, et al. The influence of anticoagulant treatment on the probability of function in femoropopliteal
vein bypass surgery: analysis of a clinical series (1970 to 1985) and interim evaluation of a controlled clinical trial. Surgery 1987; 102: 453-9. 13. Coon WW, Willis PW. Hemorrhagic complications of anticoagulant therapy. Arch Intern Med 1974; 133: 386-92. 14. Coon WW, Willis PW. Thromboembolic complications during anticoagulant therapy. Arch Surg 1972; 105: 209-12. 15. Errichetti AM, Holden A, Ansell J. Management of oral anticoagulant therapy: Experience with an anticoagulation clinic. Arch Intern Med 1984; 144: 1966-8. 16. Levine MN, Raskob G, Hirsch J. Hemorrhagic complications of long-term anticoagulant therapy. Arch Intern Med 1986; 146: 446-7. 17. Coon WW. Anticoagulant therapy. Am J Surg. October 8, 1985; 150 (4A): 45-9. 18. Hirsh J, Levine MN. Therapeutic range for the control of oral anticoagulant therapy. Arch Neurol 1986; 43:1162-4.
THE AMERICAN JOURNAL OF SURGERY
VOLUME 160 NOVEMBER 1990 505
Lifetime anticoagulation has become a therapeutic option for surgical patients with hypercoagulable states or prosthetic arterial bypass grafts. However, physicians may not achieve optimal anticoagulation or may attempt to limit the length of the therapy period because of the perceived morbidity from hemorrhagic complications of Coumadin therapy. A protocol for anticoagulant therapy monitored and regulated by a vascular nurse-clinician was reviewed. Coumadin was prescribed for 1,891 pa, tient-months to 93 patients to maintain their prothrombin time 1.5 to 2 times control (range: 18 to 24 seconds). The mean ('4- SD) prothrombin time for the study population was 19.8 + 1.8 seconds. During follow-up, 472 (14%) of 3,479 prothrombin times measured were below the therapeutic range (n = 232) or prolonged (n = 240), prompting an adjustment in the Coumadin dose in 82 (88%) patients. Four patients developed recurrent vascular graft thrombosis while receiving anticoagulation. There were 6 major and 11 minor hemorrhagic complications. Patients with a chronic risk for arterial or venous thrombosis can have outpatient anticoagulant therapy administered at optimal intensity and regulated safely with a low incidence of hemorrhagic and thrombotic events.
ifetime anticoagulation has become a therapeutic option for patients at risk for arterial or venous thromL bosis due to hypercoagulable states, prosthetic infrainguinal arterial bypasses, arterial emboli, or recurrent thrombophlebitis. Coumadin (crystalline warfarin sodium, USP) is accepted as the optimal form of outpatient anticoagulation for prophylactically protecting the venous circulation in patients with deep venous thrombosis and/or pulmonary emboli [I-3]. There has been a more limited experience with anticoagulation for patients at risk for arterial thrombosis. From the Department of Surgery, Medical College of Wisconsin, Milwaukee, Wisconsin. Requests for reprints should be addressed to Gary R. Seahrook, MD, Department of Surgery, 8700 W. Wisconsin Avenue, Milwaukee, Wisconsin 53226. Manuscript submitted September 26, 1989, and accepted in revised form December 1, 1989.
We have perceived reluctance on the part of our colleagues to prescribe long-term oral anticoagulant therapy. In view of the significant reported morbidity associated with hemorrhagic complications in patients receiving Coumadin therapy, physicians have attempted to limit the length of the therapy period and tend to be conservative in achieving optimal anticoagulation [4,5]. Patients for whom such therapy is often considered contraindicated include those with concurrent medical problems requiring multiple medications, elderly patients, and patients who travel extensively, consume alcohol regularly, or are engaged in strenuous physical activity [6]. To demonstrate that Coumadin can be prescribed safely for extended periods of time, we reviewed a protocol administered by a vascular nurse-clinician and supervised by a physician for monitoring and regulating outpatient anticoagulant therapy on an academic vascular surgery service. Patients maintained on Coumadin included those who had a variety of concomitant medical conditions including hypertension, coronary artery disease, and insulin-dependent diabetes mellitus; age into the ninth decade; and a habit of regular alcohol consumption. The efficacy and safety of the protocol were reviewed with respect to patient selection and maintenance of optimal intensity of anticoagulation versus the incidence of thrombotic or hemorrhagic complications. PATIENTS AND METHODS From 1981 to 1988, 93 patients admitted to an academic peripheral vascular surgery service were maintained on a long-term oral outpatient anticoagulant therapy program administered by a nurse-clinician. There were 56 men and 37 women whose ages ranged from 18 to 89 years (median: 63 years). Indications for therapy included hypercoagulable states in which a coagulation defect was documented, prosthetic infrainguinal arterial grafts, venous thrombosis, arterial emboli, and unexp!ained prosthetic graft thrombosis (Table I). All patients were hospitalized at the time the decision was made to prescribe long-term Coumadin therapy. Prior to discharge, the nurse-clinician reviewed the patient's medical record for demographic data, concurrent medical problems, medications, current Coumadin dose, and coagulation laboratory values and determined if he or she could understand the risks and implications of long-term anticoagulant therapy, report in a systematic fashion for laboratory monitoring of prothrombin time, and be able to adjust the Coumadin dose based on telephoned instructions. Patients excluded from participation in the program included those without a residence and/or telephone, those with uncontrolled mental illness, or those incapable of understanding the arithmetic necessary to alter the Coumadin dose based on verbal instructions.
THE AMERICAN JOURNAL OF SURGERY
VOLUME 160 NOVEMBER 1990
501
SEABROOK ET AL
TABLE I Indications for Long-Term AnUcoagulatlon Patients (n)
Hypercoagulable states Fibrinolytic disorders Antithrombin III deficiency Protein C/protein S deficiency Venous thrombosis Prosthetic infrainguinal arterial bypasses Arterial emboli Unexplained prosthetic graft occlusion
22 6 3 20 20 13 9
Patients accepted into the outpatient program were instructed prior to discharge on the basic pharmacology of anticoagulation and the risks associated with Coumadin therapy. Depending on the patient's level of education, the pathophysiologic process was explained so he or she understood the reason anticoagulation had been prescribed. The patient was taught that the dose of Coumadin might need to be changed from time to time and that monitoring of the prothrombin time would be necessary as long as the therapy was administered. The increased likelihood of minor bleeding (epistaxis, menorrhagia, gingival bleeding, hematuria, hematochezia) was explained, as well as warning signs and symptoms of a major bleeding episode that could occur when on anticoagulant therapy (dizziness, hemoptysis, hematemesis, melena, abdominal pain). The patient was advised not to take aspirin or compounds that contain salicylates while on Coumadin therapy. Excessive restraints on diet, activity, and lifestyle were not required. Patients were instructed to maintain a consistent diet and were permitted to drink alcohol in moderation, but were seriously warned against binge drinking or consuming large quantities of foods containing significant amounts of vitamin K. Patients were advised not to take vitamin supplements that contain vitamin K. The goal of therapy was to maintain a prothrombin time 1.5 to 2 times the control, resulting in a target prothrombin time between 18 and 24 secondS. Coumadin
I PROTHROMBIN TIME PROLONGED
1=
.ow oos
I
DECREASE OR HOLD DOSE
PROLONGED
CHANGE TO
THERAPEUTIC -~
RE
P.T.
1
RANGE
~L STABLE
I
SUBTHERAPEUTIC I
THERAPEUTIC PROTHROMBIN TIME [ T STABLE
INCREASE DOSE
509.
A
I
I PROTHROUSIN T ME
1=
was administered concomitantly with intravenous heparin for 5 days, beginning with a dose of 10 mg daily for 3 days. The dose was reduced to 7.5 mg on the fourth and fifth days to avoid an excessive early prolongation of the prothrombin time. The level of anticoagulation was measured daily using Quick's one-stage prothrombin time, employing a dried rabbit brain thromboplastin with calcium (American Dade, Aguada, Puerto Rico), which gave a standardized control of 12 seconds. Upon discharge from the hospital, the patient returned for measurement of prothrombin times one to three times a week until a stable dose had been achieved. The frequency of prothrombin time determinations was decreased to once monthly after the patient demonstrated a therapeutic prothrombin time on a set dose, although the frequency with which prothrombin times were measured varied depending upon the stability of the anticoagulation. The nurse-clinician contacted the patient by telephone to report that day's laboratory results and altered the daily Coumadin dose or created a dosing schedule in which the patient alternated between two doses on various days of the week. Five milligram tablets were routinely prescribed so the patient could be instructed to take one-half or multiple units of the tablets when dose changes were required. The patient was asked to consistently take the drug in the latter part of the day so that when instructions for dose changes or holding the dose were given, the patient could act on the telephoned message, which usually came in the afternoon. Each telephone contact provided the nurse-clinician with an opportunity to query the patient for evidence of bleeding complications and answer questions regarding the therapy program. Coumadin therapy was regulated during routine follow-up according to an algorithm (Figure 1). When a patient's prothrombin time was prolonged beyond the target range, the nurse-clinician directed the patient to hold or reduce the Coumadin dose and arranged for a follow-up prothrombin time to be obtained. When the prothrombin time returned to the desired level, the daily Coumadin dose was adjusted and monitored frequently until confirmation of a new stable dose was achieved. A prolonged prothrombin time prompted the nurse-clinician to care-
SUB-THERAPEUTIC
REPEAT P.T.
I
THE AMERICAN JOURNAL OF SURGERY
VOLUME 160
Figure 1. Algorithm for regulation of Coumadin therapy. Dose adjustments are increased or decreased in increments of 2.5 mg.
NOVEMBER 1990
OUTPATIENT ANTICOAGULATION PROTOCOL
fully interrogate the patient to ascertain the cause of the prolonged value and to identify any occult bleeding including the presence of petechia, ecchymosis, hematochezia, or hematuria. Patients with prolonged prothrombin times thought to be at risk for a bleeding complication were immediately hospitalized for observation. When the patient's prothrombin time was below the desired therapeutic range, he or she was instructed to increase the daily dose, and repeat prothrombin times were ordered to confirm that the desired level had been achieved. Subtherapeutic prothrombin times prompted the nurse-clinician to investigate the possibility of a venous or arterial thrombotic event. Records of each patient contact and the results of laboratory studies were maintained by the nurse-clinician with the information periodically being transcribed into the patient's permanent medical record. Patients found to be noncompliant with the protocol administered by the nurse-clinician were referred to the primary physician for management, or their therapy was discontinued if the risk of complications was thought to be excessive. Each patient in the nurse-clinician-managed program was seen at least every 6 months by the physician supervising the therapy to formally review the patient's progress, and all significant dose adjustments were reviewed with the physician at the time they were implemented. RESULTS The study reviewed 1,891 patient-months of Coumadin therapy (range: 1 to 88 months). During this time, 3,479 prothrombin times were measured, resulting in a mean of 2.6 prothrombin times per month. The mean 4SD of the prothrombin times for the study population was 19.8 4- 1.8 seconds. Eighty-two (88%) of the 93 patients required periodic alteration in the prescribed Coumadin dose after establishment of an initial stable dose. Sixtytwo (66%) of the patients were prescribed a dosing regimen of Coumadin alternating between two strengths of the drug from day to day. During the course of therapy, 472 (14%) of the 3,479 prothrombin times prompted an adjustment in the Coumadin dose; 232 were below the therapeutic range, requiring an increase in the daily Coumadin dose, and 240 were prolonged, resulting in a decrease in the patient's dose. The mean number of dose changes was 5.0 (range: 0 to 20). Twelve of the patients required more than 10 dose changes, accounting for 169 of 472 (36%) of the total number of dose changes. Eight octogenarians were prescribed 1i3 patient-months of therapy, during which 41 prothrombin times were measured. There were no complications in these elderly patients, and monitoring requirements (2.7 prothrombin times per month) mirrored that of the tota ! population. There were no deaths related to the anticoagulant therapy. There were no venous thrombotic events during this study. Four patients received Coumadin (duration of administration: 6 to 20 months) because antecedent arterial graft occlusion led to prosthetic vascular graft thrombosis while receiving anticoagulation. Three of these patients had therapeutic prothrombin times when their graft occluded. This represents an arterial graft throm-
botic rate of 8.3% for 48 grafts observed (3.7 events per 1,000 patient-months of observation). Etiologic factors contributing to these four thrombotic events included polycythemia, thrombocytosis, antithrombin III deficiency, and an axillo-femoral graft that became mechanically obstructed. Limb salvage was achieved following all of these graft occlusions. Six major hemorrhagic complications (defined as bleeding resulting in the need for hospitalization, blood transfusion, or cessation of anticoagulant therapy), occurring 12 to 42 months from commencement of therapy, resulted in a complication rate of 312 per 1,000 patientmonths of treatment. The indication for anticoagulation therapy was a plasminogen abnormality in three patients, recurrent prosthetic graft occlusion in two patients, and arterial embolization in one patient. The causes of hemorrhage included retroperitoneal bleeding, diverticulosis, ulcerative colitis, a duodenal ulcer, and two gastric ulcers. No gastrointestinal malignancies were discovered [7]. The patient with retroperitoneal bleeding was continued on Coumadin therapy because of a history of recurrent vascular graft thrombosis caused by a coagulation factor defect. Anticoagulation therapy was discontinued in the other five patients. Eleven minor hemorrhagic complications were identified including epistaxis, subcutaneous hematoma, hematuria, hyphema, menorrhagia, and gingival bleeding, representing a rate of 5.3 events per 1,000 patient-months of treatment. Each of these complications was self-limited and resulted in no significant morbidity. There was no evidence of Coumadin-related skin necrosis, which has been associated with high loading doses of the drug [8,9]. At the conclusion of this review, 56 of the 93 patients remained on Coumadin therapy. Eighteen patients were subsequently transferred back to referring physicians. The treatment plan was completed and Coumadin discontinued in 10 patients. Coumadin therapy was discontinued in five patients because of major gastrointestinal bleeding and in another patient diagnosed with a large pulmonary carcinoma. Moreover, in one patient, the drug was converted to subcutaneous heparin therapy during pregnancy. Two patients died of causes unrelated to their indications for oral anticoagulation. COMMENTS Patients requiring long-term, even lifetime, oral anticoagulation for protection against arterial or venous thrombosis can be safely managed in the outpatient setting. Administration of this protocol by a nurse-clinician enabled the patients to have carefully individualized and closely monitored therapy with appropriate education and ongoing surveillance--factors that are important in reducing the risk of complications. In contrast to many long-term medications, Coumadin has the potential for serious complications associated with significant morbidity (i.e., gastrointestinal, retroperitoneal, and intracerebral hemorrhage). Utilization of this protocol allowed patients to receive therapeutic anticoagulation with Coumadin with acceptable risks. To achieve the goals of anticoagulation in this patient
THE A M E R I C A N J O U R N A L O F S U R G E R Y
VOLUME 160
NOVEMBER 1990
503
SEABROOK ET AL
population, the criteria must be liberal for accepting patients into the therapy protocol. The patients in this Series came from an academic peripheral vascular surgery service, and their mean age was higher than that of series of patients receiving anticoagulation for deep venous thrombosis or pulmonary emboli [10]. All of the patients were taking other medications for treatment of diseases including hypertension, coronary artery disease, and diabetes mellitus. Patients receiving Coumadin therapy as prophylaxis against thrombosis of an autogenous or a prosthetic arterial bypass may benefit from anticoagulant therapy for the rest of their lives [11,12]. This study demonstrated that long-term therapy, extending in excess of 7 years, did not result in a higher complication rate for the total patient-months of therapy than in series of patients having considerably shorter courses of therapy [13-16]. Bleeding complications were distributed throughout the follow-up period and were not related to the length of the anticoagulant therapy. Coumadin, which is dose-dependent in its inhibition of factors II, VII, IX, and X, has a half-life of 2 89days; this facilitates use of a dosing schedule in which daily doses of different strengths (e.g., 5.0 mg alternating with 7.5 mg) were prescribed to maintain a stable level of anticoagulation [17]. Some patients were found to be sensitive to Coumadin and required as little as 2.5 mg every other day to achieve a therapeutic level. Heparin therapy has a slight effect on the prothrombin time, and this must be considered while therapy is being changed from heparin to Coumadin. Initial doses of Coumadin were never in excess of 10 mg to avoid the complication of Coumadinassociated skin necrosis. Prothrombin times were measured using standardized rabbit brain thromboplastin, which delivers a control prothrombin time of 12 seconds. The mean prothrombin time calculated for the study population was 19.8 seconds 4- SD 1.8 seconds (1.4 to 2.0 times control), which is an appropriate target range when rabbit brain thromboplastin is used as the reference. When human brain thromboplastin, which is more sensitive than rabbit brain thromboplastin, is used, higher ratios (2.0 to 2.5 times control) have been recommended [18]. The intensity of the degree of anticoagulation was very consistent in the study population. When a patient had a prothrombin time prolonged beyond the target range, a cause could usually be identified and was most frequently related to the use of an over-the-counter medication for a minor illness or an increase in alcohol consumption. A prothrombin time below the therapeutic range could usually be attributed to noncompliance with the dosage schedule or dietary indiscretions. There was considerable variability within the study group with respect to maintaining a stable prothrombin time on a given dose of Coumadin. The number of dose changes correlated with the number of months the patient had been administered Coumadin therapy. The 12 patients requiring more than 10 dose changes received therapy from 10 to 88 months (mean: 40 months). In several patients who were followed for prolonged periods of time, it was noted that after many months of consistent 504
therapy on a single dose regimen, a significant change in the required Coumadin dose would be necessary to maintain the prothrombin time in the therapeutic range. Despite careful questioning for changes in diet, activity, or medication and with the assurances that the patient was reliable and compliant, the cause of the required dose change could not always be established. For the major hemorrhagic complications occurring in this series, the patients' prothrombin times were within the target therapeutic range. No complications occurred when the patient had an excessively prolonged prothrombin time. Of the four patients who had recurrent arterial thrombotic events related to prosthetic arterial bypass grafts, one of the occlusions occurred when the patient's prothrombin time had fallen to a subtherapeutic level. In the other three patients, the prothrombin time was within the desired therapeutic range. However, each of these patients had an etiologic factor to which the occlusion could be ascribed: polycythemia, thrombocytosis, antithrombin III deficiency, and mechanical graft obstruction. Lifetime oral anticoagulation is a viable therapeutic option in treating patients at chronic risk for thrombosis. An outpatient protocol can be managed safely and effectively by a vascular nurse-clinician for an elderly population from an academic surgical service. Complications in this group did not exceed reported complications in other large series of outpatient anticoagulation regimens, many of which were for shorter periods of time in younger patient populations with fewer associated pathologic conditions. Hemorrhagic complications that occurred were not related to anticoagulated states exceeding the target therapeutic range. The nurse-clinician and patient relationships allowed successful management of Coumadin therapy and provided frequent contact and accessibility for the patient to the health care system. Long-term, even lifetime, oral anticoagulant therapy can be managed in the outpatient setting by an experienced registered nurseclinician with excellent patient compliance and a low frequency of hemorrhagic and thrombotic events. Patients with a legitimate chronic risk for arterial or venous thrombosis can have outpatient anticoagulant therapy administered and regulated with a low incidence of hemorrhagic and thrombotic events. REFERENCES 1, Hull R, Hirsh J, Jay R, et al. Differentintensitiesof anticoagula~ tion in the long-term treatment of proximalvenousthrombosis. N Engl J Med 1982; 307: 1676-81. 2. Hull R, Delmore T, Genton E, et al. Warfarin sodiumversus low-doseheparinin the long-termtreatment of venousthromboembolism. N Engl J Med 1979; 301: 855-8. 3. Hull R, Delmore T, Carter C, et at. Adjusted subcutaneous heparin versus warfarin sodiumin the long-termtreatment of venous thrombosis. N Engl J Med 1982; 306: 189-94. 4. MosleyDH, Schatz IJ, BrenemanGM, KeyesJW. Long-term anticoagulant therapy. JAMA 1983; 12: 128-30. 5. PollardJW, HamiltonMJ, ChristensenNA, AchorRWP. Problemsassociatedwith long-termanticoagulanttherapy:observations in 139 cases. Circulation 1962; 25: 311-7. 6. Davis FB, Estruch MT, Samson-CorveraEB, Voigt GC, Tobin
THE AMERICAN JOURNAL OF SURGERY VOLUME 160 NOVEMBER1990
OUTPATIENTANTICOAGULATIONPROTOCOl,
JD. Management of anticoagulation in outpatients. Arch Intern Med 1977; 137: 197-202. 7. Jaffin LW, Bliss M, Lamont T. Significance of occult gastrointestinal bleeding during anticoagulation therapy. Am J Med 1987; 83: 269-72. 8. Kazmier FJ. Thromboembolism, Coumadin necrosis, and protein C. Mayo Clin Proc 1985; 60: 673-4. 9. Peterson CE, Kwaan HC. Current concepts of warfarin therapy. Arch Intern Med 1986; 146: 581-4. 10. Gurwitz JH, Goldberg R J, Holden A, Knapic N, Ansell J. Agerelated risks of long-term oral anticoagulant therapy. Arch Intern Med 1988; 148: 1733-6. 11. Flinn WR, Rohrer M J, Yao JST, McCarthy WJ, Fahey VA, Bergan JJ. Improved long-term patency of infragenicular polytetrafluoroethylene grafts. J Vasc Surg 1988; 7: 685-90. 12. Kretschmer G, Wenzl E, Piza F, et al. The influence of anticoagulant treatment on the probability of function in femoropopliteal
vein bypass surgery: analysis of a clinical series (1970 to 1985) and interim evaluation of a controlled clinical trial. Surgery 1987; 102: 453-9. 13. Coon WW, Willis PW. Hemorrhagic complications of anticoagulant therapy. Arch Intern Med 1974; 133: 386-92. 14. Coon WW, Willis PW. Thromboembolic complications during anticoagulant therapy. Arch Surg 1972; 105: 209-12. 15. Errichetti AM, Holden A, Ansell J. Management of oral anticoagulant therapy: Experience with an anticoagulation clinic. Arch Intern Med 1984; 144: 1966-8. 16. Levine MN, Raskob G, Hirsch J. Hemorrhagic complications of long-term anticoagulant therapy. Arch Intern Med 1986; 146: 446-7. 17. Coon WW. Anticoagulant therapy. Am J Surg. October 8, 1985; 150 (4A): 45-9. 18. Hirsh J, Levine MN. Therapeutic range for the control of oral anticoagulant therapy. Arch Neurol 1986; 43:1162-4.
THE AMERICAN JOURNAL OF SURGERY
VOLUME 160 NOVEMBER 1990 505
