- Email: [email protected]
Economic implications of thrombolysis or operation as the initial treatment modality in acute peripheral arterial occlusion
Economic implications of thrombolysis or operation as the initial treatment modality in acute peripheral arterial occlusion Kenneth Ouriel, MD, Mick Kolassa, Phl), James A. DeWeese, MD, and Richard M. Green, MD, Rochester, N.Y., and University, Miss.
Background. Thrombolytic agents have been used to treat arterial occlusion for almost 40 years, l, 2 Recently, an investigation of the costs associated with two treatment options for acute peripheral arterial occlusions, thrombolysis and surgical intervention, was completed. The availability of hospital cost data for patients enrolled in the thrombolytic and operative treatment groups provided a basis with which to accurately assess the financial impact of the different treatment strategies, both from a purely financial standpoint and in relation to outcome. Methods. The patient base was composed of 114 patients with acute limb ischemia of less than 7 days' duration. The patients were randomly assigned to receive urokinase (n = 57) or to undergo an operation (n = 57) as the initial therapeutic intervention. Patients in the thrombolytic group underwent standard intraarterial diagnostic arteriography, and patients with embolic events, in whom complete lysis of all embolic and propagated thrombotic material was achieved, were subsequently treated with heparin and long-term warfarin (Coumadin) therapy. The economic analysis was undertaken after the completion of the trial. Statistical comparisons between groups were made with the Student t test for continuous, normally distributed data. Mortality and limb salvage rates were calculated from Kaplan-Meier curves, appropriate for the censored nature of the data. Results. The total treatment costs did not differ significantly between the two treatment groups ($22,171 +- $4,959 in the thrombolytic group and $19, 775 +- $5,253 in the operative group). The total hospital charges were similar between the two groups. Overall, the total charges were remarkably similar between the two treatment groups, averaging $40,823 +- $8, 764 in the thrombolytic group and $41,930 + $10,398 in the operative group. Conclusions. An economic analysis of the data confirmed that the total economic impact of thrombolysis approximated that of initial operative therapy. The improved clinical outcome in patients treated with thrombolysis suggests that this modality may be appropriate as the initial therapeutic intervention in the select group of patients seen within the first few hours of an acute peripheral arterial occlusive event. (SunGEnr 1995; I 18: 810-4. ) From the Department of Surgery, Rochester General Hospital, The University of Rochester, Rochester, N.Y., and The School of Pharmacy, Research Institute of Pharmaceutical ,Sciences, University of Mississippi, University, Miss.
THROMBOLYPIC AGENTSHAVEBEEN USED in the setting of arterial occlusion for almost 40 years, l' 2 Nevertheless, meaningful data on the differences between thrombolytic a n d operative strategies awaited the c o m p l e t i o n o f scientifically s o u n d studies c o m p a r i n g the two strategies in a randomized, prospective fashion. Recently, two such investigations were completed, one single center study at the University of Rochester ~ a n d a larger, multiinstitutional study entitled the Study o f T h r o m Supported by Abbott Laboratories, Abbott Park. Illinois Accepted for publication March 9, 1995. Reprint requests:Kenneth Ouriel, MD, 919 WesffallRd., Rochester, NY 14618. Copyright 9 1995 by Mosby-YearBook, Inc. 0039-6060/95/$5.00 + 0 11/56/65411 810
SURGERY
bolysis in Ischemia o f the Lower Extremity. 4 An analysis o f the available data suggests that thrombolytic therapy may be beneficial in certain specific subcategories o f patients with p e r i p h e r a l arterial occlusion, with a reduction in mortality rate "~and an increase in the rate o f limb salvage. 4 To date, the e c o n o m i c issues o f thrombolytic intervention have n o t b e e n well analyzed. T h e availability o f hospital cost data on patients enrolled in the thrombolytic a n d operative t r e a t m e n t groups o f the University of Rochester study provided a basis with which to accurately assess the financial impact of the different treatm e n t strategies, both from a purely financial s t a n d p o i n t a n d in relation to outcome. Patients enrolled in the University o f Rochester trial were r a n d o m i z e d to initial treatment with either thrombolytic infusion o r opera-
Surgery Volume 118, Number 5
tion. The ultimate goal of thrombolysis was to unmask an anatomic lesion responsible for the occlusive event; thus thrombolytic therapy was followed by a directed invasive intervention designed to correct any identified etiologic problem. The clinical results of the patients in this series have been reported elsewhere3; the present investigation focuses on the economic implications of therapy.
METHODS The patient base was composed of 114 patients with acute limb ischemia of less than 7 days' duration. Patients were randomly assigned to receive urokinase (Abbokinase; Abbott Laboratories, North Chicago, Ill.) or to undergo an operation as the initial therapeutic intervention. Patient acquisition began in August 1989 and concluded in April 1992. Patients with thrombotic or embolic occlusion of bypass grafts or native arteries were included if the ischemic process was of a limbthreatening nature (class II). The methodology has been described in detail in a previous publication. 3 Briefly, patients meeting defined entry criteria (Table I) were treated with 325 mg aspirin and were randomized to the thrombolytic (57 patients) or operative group (57 patients). Patients in the thrombolytic group underwent standard intraarterial diagnostic arteriography, placing the infusion holes of the catheter into the substance of the thrombus whenever possible. Urokinase was infused at 4000 I U / m i n for 2 hours, 2000 I U / m i n for 2 hours, and then 1000 I U / m i n for the remainder of therapy up to a maximum of 48 hours. Therapy was terminated if successful lysis was documented on arteriogram, if major complications developed, or if clinical deterioration progressed during the period of infusion. After completion of thrombolytic infusion, patients were treated with operative or endovascular correction of residual anatomic lesions unmasked by successful thrombolysis. Patients with embolic events in whom complete lysis of all embolic and propagated thrombotic material was achieved were subsequently treated with heparin and long-term warfarin (Coumadin) therapy. Long-term anticoagulation was also used when successful thrombolysis failed to reveal the presence of a lesion that explained the bypass graft failure or occlusion of a native artery. Patients without a successful thrombolytic result were taken to operation for revascularization, usually in the form of insertion of a new bypass graft. Patients in the operative group underwent diagnostic arteriography unless the procedure was d e e m e d inadvisable on the basis of delays associated with the performance of the test (three patients). Operation was performed as soon as feasible, and the choice of procedure was made by the attending surgeon. Primary amputation was elected when revascularization was not possible.
Ouriel et al.
811
T a b l e I. Eligibility criteria for patient entry Inclusion criteria
Acute symptoms <7 days
Limb-threatening ischemia (class II) Age ->18 years Informed consent
Exclusion criteria
Contraindication to thrombolysis or operation Class I or III ischemia Echocardiographic cardiac thrombus Positive urine pregnancy test
The economic analysis was undertaken after completion of the trial. Hospital cost data were collected from the computerized records of the finance department. The n u m b e r and magnitude of professional services were tabulated from the patients' charts and multiplied by the Medicare Part B allowed fee to determine professional charges. Statistical comparisons between groups were made with the Student t test for continuous, normally distributed data. Mortality and limb salvage rates were calculated from Kaplan-Meier curves, appropriate for the censored nature of the data. The term cost refers to the financial costs borne by the hospital. The term charge refers to the financial outlays by the patient or payer for the provision of services. Both cost and charge data were collected for hospital-related services, whereas only charge data were collected for professional services. Economic data are in 1992 U.S. dollars for all analyses.
RESULTS Comparison o f groups. The two randomized groups were comparable with regard to demographics, comorbid conditions, and cardiac risk index (Table I). The ankle-brachial index averaged 0.04 --+ 0.02 in the two groups. The duration of symptoms did not differ, averaging 39 -+ 6.3 hours in the thrombolytic group and 42 + 6.7 hours in the operative group. Rest pain was present in 89% (51 of 57) of the thrombolytic patients and 88% (50 of 57) of the operative patients. Motor loss was present in 19% (11 of 57) of the thrombolytic group and 25% (14 of 57) of the operative group. The etiologic diagnosis for the occlusive processes did not differ between the two groups. Embolic occlusions occurred in 11 (19%) of the 57 patients randomized to thrombolytic therapy and in 13 (23%) of the 57 patients randomized to operative therapy. Thrombotic occlusions occurred in the remainder of the patients, in native arteries in 13 (23%) of 57 patients randomized to thrombolytic therapy and 14 (25%) of 57 patients randomized to operative therapy. Bypass graft occlusions occurred in 33 (58%) of the 57 patients in the throm-
812
Ouriel et al.
Surgery November 1995
T a b l e II. Comparison of hospital costs
Nonpharmaceutical costs ($) Medication including urokinase ($) Medication, urokinase only ($) TOTALCOSTS
Thrombolytic group (n = 57)
Operative group (n = 57)
17,176 _+4,328
16,365 -+ 4,408
4,995 - 630
3,410 -+ 764
2,653 + 243
0
22,171 --+4,959
19,775 -- 5,253
T a b l e I l L Comparison of charges
Thrombolytic group Total hospital charges ($) Medication with urokinase ($) Urokinase alone ($) No. of primary consultations No. of secondary consultations Total consultation fees ($) Surgeon fees ($) Anesthesiologist fees ($) Radiologist fees ($) TOTAl. FEES ($) TOTAl. CHARGES($)
Operative group
38,377 -+ 8,583 7,259 +- 916
4,790 -+ 1,073
3,854 - 353 1.3
0 1.8
1.4
3.0
186 -+ 76
314 _+ 100
863 -+ 173 787 z 165 609 -+ 21 2,445 + 315 40,823 -- 8,764
1,417 -+ 127 1,356 -+ 227 428 -+ 7 3,517 _+318 41,930 +-- 10,398
bolytic group versus 30 (53%) of the 57 patients in the operative group. Clinical outcome. Thrombolysis was successful in dissolving substantially all (greater than 80%) of the thrombus in 40 (70%) of the patients randomized to the thrombolytic treatment group. The 17 patients with thrombolytic failures underwent a surgical procedure to restore limb viability. O f the 40 patients with successful thrombolysis, an anatomic lesion responsible for the occlusive event was unmasked in 21 (53%). Each underwent correction of the lesion by surgical or endovascular means. Overall, an adjunctive surgical or endovascular procedure was performed in 38 (67%) of the patients in the thrombolytic group. The mean length of hospitalization was 14.2 days in each group. The length of stay was greater in the patients who required a major amputation during the initial hospitalization, irrespective of whether that patient was in the thrombolytic or operative group. Similarly, the hospital costs were substantially greater in pa-
T a b l e IV. Cost-effectiveness of thrombolytic and operative therapy
Life expectancy (yr) Cost per life saved ($) Cost per life-year saved ($)
Thrombolysis
Operation
12.4 _+ 1.3 49,508
9.4 _+ 1.3 70,295
3,980
7,489
tients in whom amputation was necessary. Patients who underwent an open surgical procedure had a significantly increased length of stay and total hospital cost when compared with those patients for whom an endovascular procedure was possible. Substantial differences in the clinical characteristics of the subgroups, however, render meaningful conclusions impossible. As reported previously, the cumulative limb salvage rates were identical in the two treatment groups, approximating 82% at 12 months by Kaplan-Meier analysis. By contrast, the cumulative survival rate was significantly improved in patients randomized to the thrombolysis group, 84% versus 58% at 12 months (p = 0.01). The mortality differences appeared to coincide with an increase in the frequency of in-hospital cardiopulmonary complications in the operative group (49 % versus 16%, p = 0.001 ). Economic analysis. The total treatment costs did not differ significantly between the two treatment groups ($22,171 +-$4,959 in the thrombolytic group and $19,775-+ $5,253 in the operative group), although there were some significant differences in the individual components that comprised the calculation of these parameters (Table II). Pharmacy costs represented 23% of the total hospital cost in the thrombolytic group compared with 17% in the operative group (p< 0.05). The cost of urokinase averaged $2653 -+ $243 in the thrombolytic group, constituting a mean of 53% of the total pharmacy costs for each patient. The total hospital charges were similar in the two groups, averaging just more than $38,000 (Table III). By contrast, the total profession charges were lower in the thrombolytic group, $2,445 + $315 versus $3517 + $318 (p < 0.05). The patients in the thrombolytic group had a mean of 1.3 primary and 1.4 secondary medical consultations during hospitalization (mean charges, $185 -+ $76) compared with 1.8 and 3.0 in the operative patients (mean charges, $314 + $100, p < 0.05). When the professional charges were subgrouped by specialty, there were significantly higher mean charges in the operative group for surgical and anesthesiology charges and significantly higher mean charges in the thrombolytic group for radiologic charges. Overall, the total charges were remarkably similar between the two treatment groups, averaging $40,823 + $8,764 in the throm-
Surgery Volume 118, Number 5
bolytic group and $41,930 -+ $10,398 in the operative group. Cost effectiveness. For the cost-effectiveness analysis, the hospital cost data were added to the professional fee data to approximate the total economic impact of treatment. The life expectancy data were culled from standard life expectancy tables from the Statistical Abstract of the United States (1989). As illustrated in Table IV, operative therapy was associated with an increase of greater than 40% in the cost per life saved and nearly 90% in cost per life-year saved when compared with thrombolytic therapy. The marginal cost-effectiveness of a therapy is defined as that sum required to save one additional life or one additional life-year when using one intervention over another. Thrombolysis was associated with a savings of $20,787 per life saved and $3,509 per life-year saved. Compared with the cost-effectiveness of other commonly used therapeutic strategies, "~ both thrombolysis and operation fared well (Table V), with a relatively low cost to achieve a satisfactory patient outcome. DISCUSSION
Previously, we documented clinical benefits associated with thrombolytic treatment of acute peripheral arterial occlusion. An economic analysis of the data confirmed that these benefits were achieved without an increase in the cost of patient care. A slightly increased hospital cost in the thrombolytic group was offset by lower professional fees, so the total economic impact of thrombolysis approximated that of initial operative therapy. The initial treatment of patients randomized to thrombolytic intervention was urokinase infusion, followed by operative or endovascular repair of defects unmasked after dissolution of thrombus. In the present trial, operative repair was used much more frequently than balloon angioplasty. The length of hospitalization and total hospital costs were significantly shorter in the patients who underwent balloon angioplasty rather than surgical correction of an unmasked lesion. O n e criticism of the study relates to this disparity; the cost of therapy might have been lower if endovascular methods were used more frequently. O u r previous report suggested that success was unlikely if no degree of reperfusion had been accomplished after 18 hours of therapy. ~ The mean cost of urokinase could have been reduced by more than $2600 if patients without any degree of reperfusion after 18 hours of infusion were excluded. Similarly, the average hospital costs were significantly reduced in the patients without reperfusion at 18 hours, suggesting that total cost of patient care might have been significantly reduced if thrombolysis had been abandoned if successful dissolution had not begun by 18 hours.
Ouriel et al.
813
T a b l e V. Comparison of cost per life-year saved for variety of medical interventions
Condition
Intervention
Hypercholesterolemia~ Myocardial infarction 7 Hypertension 6
Cholestyramine rt-PA Captoril and diuretic Operation Thrombolysis (urokinase)
Acute limb ischemia Acute limb ischemia
Cost per life-year saved ($)
190,000 60,478 23,760 7,489 3,980
Janosik et al. s reported their experience with urokinase and streptokinase and surgical thrombectomy in the treatment of patients with peripheral arterial occlusion between 1981 and 1985. The mean length of hospital stay in this study averaged 3 weeks for those receiving streptokinase and surgical thrombectomy compared with only 11.5 days for those treated with urokinase. The total hospital charges, however, did not differ in the three groups, averaging between $20,000 and $30,000 1983 U.S. dollars. The results of this study must be taken in the context of its retrospective design; patients receiving streptokinase were treated exclusively during the early years of the series and patients receiving urokinase were treated during the latter years. A study by van Breda et al. 9 summarized the cost-effectiveness of streptokinase and urokinase in 271 patients undergoing thrombolysis for peripheral arterial occlusions between 1979 and 1987. Successful clot lysis was achieved more frequently in the patients receiving urokinase, at a hospital cost averaging approximately $15,000 per therapeutic success. As in the report by Janosik et al., s however, patients receiving streptokinase were treated during the earlier years of the experience. O u r trial differs from the studies ofJanosik et al. s and van Breda et al. 9 in its prospective, randomized design. In patients assigned to thrombolytic or surgical management, we noted no difference in the length of stay or the hospital charges between the groups. Between 1985, the last year of theJanosik series, and 1991, the last year of our series, the mean length of stay decreased considerably. Despite a shorter length of stay in the more recent series, hospital charges did not differ significantly (corrected for inflation) between 1983 and 1991. In summary, the findings of the present study suggest that the economic impact of thrombolyfic therapy as the initial intervention in patients with acute peripheral arterial occlusion is equivalent to that of immediate operation. The improved clinical outcome in patients treated with thrombolysis suggests that this modality may be the appropriate initial therapeutic intervention
814
Ouriel et al.
in the select g r o u p o f patients s e e n within the first few h o u r s o f an acute p e r i p h e r a l arterial occlusive event. T h e s e results, however, m u s t n o t be g e n e r a l i z e d to the larger p o p u l a t i o n o f patients p r e s e n t i n g with occlusions o f a m o r e c h r o n i c nature. R e c o m m e n d a t i o n s o n the m o s t a p p r o p r i a t e therapy in select subsets o f patients m u s t await the results o f large, well-controlled clinical trials. Until such data b e c o m e available, patients s h o u l d be treated o n an individual basis, relying o n the clinical expertise o f the surgical a n d a n g i o g r a p h i c p a t i e n t care teams. REFERENCES
1. Cliffton EE, Grossi CE. Investigations of intravenous plasmin (fibrinolysin) in humans; physiologic and clinical effects. Circulation 1956;14:919-29. 2. Tillett WS,Johnson AJ, McCartyWR. The intravenous infusion of the streptococcal fibrinolytic principle (Streptokinase) into patients. J Clin Invest 1955;34:169-85. 3. Ouriel K, Shortell CK, Marder VJ, et al. A comparison of throm-
Surgery November 1995 bolytic therapy with operative revascularization in the initial treatment of acute peripheral arterial ischemia. J Vasc Surg 1994;19:1021-30. 4. The STILE Investigators. Results of a prospective, randomized trial evaluating surgery vs. thrombolysis for ischemia of the lower extremity: the "STILE" trial. Ann Surg 1995;222(suppl 3):251-68. 5. Cairns JA, Collins R, Fuster V, Passamani ER. Coronary thrombolysis. Chest 1989;95(suppl):73-87. 6. Martens LL, Rutten FFH, Erkelens DW, Ascoop CAPL. Cost effectiveness of cholesterol lowering therapy in the Netherlands: simvastatin versus cholestyramine. Am J Med 1989;87(suppl 4A) :54s-Ss. 7. Guerci AD. A cost effectiveness analysis of t-PA. Clinical Courier 1989;7:5:5-8. 8. Janosik JE, Bettmann MA, Kaul AF, Souney PF. Therapeutic alternatives for subacute peripheral arterial occlusion: comparison by outcome, length of stay, and hospital charges. Radiology 1991;26:921-5. 9. van Breda A, Graor RA, Katzen BT, Risius B, Gillings D. Relative cost-effectiveness of urokinase versus streptokinase in the treatment of peripheral vascular disease.J Vasc Interv Radio11991;2:7787.
SURGERYis abstracted a n d / o r indexed in Index Medicus, Science Citation Index, Current Contents/ Clinical Medicine, Current Contents~Life Sciences, Excerpta Medica, and MEDLINE. This JOURNAL has been registered with Copyright Clearance Center, Inc., 222 Rosewood Dr., Danvers, MA 01923. Consent is given for the copying of articles for personal or internal use of specific clients. This consent is given on the condition that the copier pay directly to the Center the per-copy fee stated on the first page of each article for copying beyond that permitted by U.S. Copyright Law. This consent does not extend to o t h e r kinds of copying, such as for general distribution, resale, advertising and promotional purposes, or for creating new collective works. All inquiries regarding copyrighted material from this publication o t h e r than those that can be handled through Copyright Clearance Center should be directed to Associate J o u r n a l Publisher, Permissions, Mosby-Year Book, Inc., 11830 Westline Industrial Dr., St. Louis, MO 63146; p h o n e (314)579-2808; fax (314)432-1380.
Background. Thrombolytic agents have been used to treat arterial occlusion for almost 40 years, l, 2 Recently, an investigation of the costs associated with two treatment options for acute peripheral arterial occlusions, thrombolysis and surgical intervention, was completed. The availability of hospital cost data for patients enrolled in the thrombolytic and operative treatment groups provided a basis with which to accurately assess the financial impact of the different treatment strategies, both from a purely financial standpoint and in relation to outcome. Methods. The patient base was composed of 114 patients with acute limb ischemia of less than 7 days' duration. The patients were randomly assigned to receive urokinase (n = 57) or to undergo an operation (n = 57) as the initial therapeutic intervention. Patients in the thrombolytic group underwent standard intraarterial diagnostic arteriography, and patients with embolic events, in whom complete lysis of all embolic and propagated thrombotic material was achieved, were subsequently treated with heparin and long-term warfarin (Coumadin) therapy. The economic analysis was undertaken after the completion of the trial. Statistical comparisons between groups were made with the Student t test for continuous, normally distributed data. Mortality and limb salvage rates were calculated from Kaplan-Meier curves, appropriate for the censored nature of the data. Results. The total treatment costs did not differ significantly between the two treatment groups ($22,171 +- $4,959 in the thrombolytic group and $19, 775 +- $5,253 in the operative group). The total hospital charges were similar between the two groups. Overall, the total charges were remarkably similar between the two treatment groups, averaging $40,823 +- $8, 764 in the thrombolytic group and $41,930 + $10,398 in the operative group. Conclusions. An economic analysis of the data confirmed that the total economic impact of thrombolysis approximated that of initial operative therapy. The improved clinical outcome in patients treated with thrombolysis suggests that this modality may be appropriate as the initial therapeutic intervention in the select group of patients seen within the first few hours of an acute peripheral arterial occlusive event. (SunGEnr 1995; I 18: 810-4. ) From the Department of Surgery, Rochester General Hospital, The University of Rochester, Rochester, N.Y., and The School of Pharmacy, Research Institute of Pharmaceutical ,Sciences, University of Mississippi, University, Miss.
THROMBOLYPIC AGENTSHAVEBEEN USED in the setting of arterial occlusion for almost 40 years, l' 2 Nevertheless, meaningful data on the differences between thrombolytic a n d operative strategies awaited the c o m p l e t i o n o f scientifically s o u n d studies c o m p a r i n g the two strategies in a randomized, prospective fashion. Recently, two such investigations were completed, one single center study at the University of Rochester ~ a n d a larger, multiinstitutional study entitled the Study o f T h r o m Supported by Abbott Laboratories, Abbott Park. Illinois Accepted for publication March 9, 1995. Reprint requests:Kenneth Ouriel, MD, 919 WesffallRd., Rochester, NY 14618. Copyright 9 1995 by Mosby-YearBook, Inc. 0039-6060/95/$5.00 + 0 11/56/65411 810
SURGERY
bolysis in Ischemia o f the Lower Extremity. 4 An analysis o f the available data suggests that thrombolytic therapy may be beneficial in certain specific subcategories o f patients with p e r i p h e r a l arterial occlusion, with a reduction in mortality rate "~and an increase in the rate o f limb salvage. 4 To date, the e c o n o m i c issues o f thrombolytic intervention have n o t b e e n well analyzed. T h e availability o f hospital cost data on patients enrolled in the thrombolytic a n d operative t r e a t m e n t groups o f the University of Rochester study provided a basis with which to accurately assess the financial impact of the different treatm e n t strategies, both from a purely financial s t a n d p o i n t a n d in relation to outcome. Patients enrolled in the University o f Rochester trial were r a n d o m i z e d to initial treatment with either thrombolytic infusion o r opera-
Surgery Volume 118, Number 5
tion. The ultimate goal of thrombolysis was to unmask an anatomic lesion responsible for the occlusive event; thus thrombolytic therapy was followed by a directed invasive intervention designed to correct any identified etiologic problem. The clinical results of the patients in this series have been reported elsewhere3; the present investigation focuses on the economic implications of therapy.
METHODS The patient base was composed of 114 patients with acute limb ischemia of less than 7 days' duration. Patients were randomly assigned to receive urokinase (Abbokinase; Abbott Laboratories, North Chicago, Ill.) or to undergo an operation as the initial therapeutic intervention. Patient acquisition began in August 1989 and concluded in April 1992. Patients with thrombotic or embolic occlusion of bypass grafts or native arteries were included if the ischemic process was of a limbthreatening nature (class II). The methodology has been described in detail in a previous publication. 3 Briefly, patients meeting defined entry criteria (Table I) were treated with 325 mg aspirin and were randomized to the thrombolytic (57 patients) or operative group (57 patients). Patients in the thrombolytic group underwent standard intraarterial diagnostic arteriography, placing the infusion holes of the catheter into the substance of the thrombus whenever possible. Urokinase was infused at 4000 I U / m i n for 2 hours, 2000 I U / m i n for 2 hours, and then 1000 I U / m i n for the remainder of therapy up to a maximum of 48 hours. Therapy was terminated if successful lysis was documented on arteriogram, if major complications developed, or if clinical deterioration progressed during the period of infusion. After completion of thrombolytic infusion, patients were treated with operative or endovascular correction of residual anatomic lesions unmasked by successful thrombolysis. Patients with embolic events in whom complete lysis of all embolic and propagated thrombotic material was achieved were subsequently treated with heparin and long-term warfarin (Coumadin) therapy. Long-term anticoagulation was also used when successful thrombolysis failed to reveal the presence of a lesion that explained the bypass graft failure or occlusion of a native artery. Patients without a successful thrombolytic result were taken to operation for revascularization, usually in the form of insertion of a new bypass graft. Patients in the operative group underwent diagnostic arteriography unless the procedure was d e e m e d inadvisable on the basis of delays associated with the performance of the test (three patients). Operation was performed as soon as feasible, and the choice of procedure was made by the attending surgeon. Primary amputation was elected when revascularization was not possible.
Ouriel et al.
811
T a b l e I. Eligibility criteria for patient entry Inclusion criteria
Acute symptoms <7 days
Limb-threatening ischemia (class II) Age ->18 years Informed consent
Exclusion criteria
Contraindication to thrombolysis or operation Class I or III ischemia Echocardiographic cardiac thrombus Positive urine pregnancy test
The economic analysis was undertaken after completion of the trial. Hospital cost data were collected from the computerized records of the finance department. The n u m b e r and magnitude of professional services were tabulated from the patients' charts and multiplied by the Medicare Part B allowed fee to determine professional charges. Statistical comparisons between groups were made with the Student t test for continuous, normally distributed data. Mortality and limb salvage rates were calculated from Kaplan-Meier curves, appropriate for the censored nature of the data. The term cost refers to the financial costs borne by the hospital. The term charge refers to the financial outlays by the patient or payer for the provision of services. Both cost and charge data were collected for hospital-related services, whereas only charge data were collected for professional services. Economic data are in 1992 U.S. dollars for all analyses.
RESULTS Comparison o f groups. The two randomized groups were comparable with regard to demographics, comorbid conditions, and cardiac risk index (Table I). The ankle-brachial index averaged 0.04 --+ 0.02 in the two groups. The duration of symptoms did not differ, averaging 39 -+ 6.3 hours in the thrombolytic group and 42 + 6.7 hours in the operative group. Rest pain was present in 89% (51 of 57) of the thrombolytic patients and 88% (50 of 57) of the operative patients. Motor loss was present in 19% (11 of 57) of the thrombolytic group and 25% (14 of 57) of the operative group. The etiologic diagnosis for the occlusive processes did not differ between the two groups. Embolic occlusions occurred in 11 (19%) of the 57 patients randomized to thrombolytic therapy and in 13 (23%) of the 57 patients randomized to operative therapy. Thrombotic occlusions occurred in the remainder of the patients, in native arteries in 13 (23%) of 57 patients randomized to thrombolytic therapy and 14 (25%) of 57 patients randomized to operative therapy. Bypass graft occlusions occurred in 33 (58%) of the 57 patients in the throm-
812
Ouriel et al.
Surgery November 1995
T a b l e II. Comparison of hospital costs
Nonpharmaceutical costs ($) Medication including urokinase ($) Medication, urokinase only ($) TOTALCOSTS
Thrombolytic group (n = 57)
Operative group (n = 57)
17,176 _+4,328
16,365 -+ 4,408
4,995 - 630
3,410 -+ 764
2,653 + 243
0
22,171 --+4,959
19,775 -- 5,253
T a b l e I l L Comparison of charges
Thrombolytic group Total hospital charges ($) Medication with urokinase ($) Urokinase alone ($) No. of primary consultations No. of secondary consultations Total consultation fees ($) Surgeon fees ($) Anesthesiologist fees ($) Radiologist fees ($) TOTAl. FEES ($) TOTAl. CHARGES($)
Operative group
38,377 -+ 8,583 7,259 +- 916
4,790 -+ 1,073
3,854 - 353 1.3
0 1.8
1.4
3.0
186 -+ 76
314 _+ 100
863 -+ 173 787 z 165 609 -+ 21 2,445 + 315 40,823 -- 8,764
1,417 -+ 127 1,356 -+ 227 428 -+ 7 3,517 _+318 41,930 +-- 10,398
bolytic group versus 30 (53%) of the 57 patients in the operative group. Clinical outcome. Thrombolysis was successful in dissolving substantially all (greater than 80%) of the thrombus in 40 (70%) of the patients randomized to the thrombolytic treatment group. The 17 patients with thrombolytic failures underwent a surgical procedure to restore limb viability. O f the 40 patients with successful thrombolysis, an anatomic lesion responsible for the occlusive event was unmasked in 21 (53%). Each underwent correction of the lesion by surgical or endovascular means. Overall, an adjunctive surgical or endovascular procedure was performed in 38 (67%) of the patients in the thrombolytic group. The mean length of hospitalization was 14.2 days in each group. The length of stay was greater in the patients who required a major amputation during the initial hospitalization, irrespective of whether that patient was in the thrombolytic or operative group. Similarly, the hospital costs were substantially greater in pa-
T a b l e IV. Cost-effectiveness of thrombolytic and operative therapy
Life expectancy (yr) Cost per life saved ($) Cost per life-year saved ($)
Thrombolysis
Operation
12.4 _+ 1.3 49,508
9.4 _+ 1.3 70,295
3,980
7,489
tients in whom amputation was necessary. Patients who underwent an open surgical procedure had a significantly increased length of stay and total hospital cost when compared with those patients for whom an endovascular procedure was possible. Substantial differences in the clinical characteristics of the subgroups, however, render meaningful conclusions impossible. As reported previously, the cumulative limb salvage rates were identical in the two treatment groups, approximating 82% at 12 months by Kaplan-Meier analysis. By contrast, the cumulative survival rate was significantly improved in patients randomized to the thrombolysis group, 84% versus 58% at 12 months (p = 0.01). The mortality differences appeared to coincide with an increase in the frequency of in-hospital cardiopulmonary complications in the operative group (49 % versus 16%, p = 0.001 ). Economic analysis. The total treatment costs did not differ significantly between the two treatment groups ($22,171 +-$4,959 in the thrombolytic group and $19,775-+ $5,253 in the operative group), although there were some significant differences in the individual components that comprised the calculation of these parameters (Table II). Pharmacy costs represented 23% of the total hospital cost in the thrombolytic group compared with 17% in the operative group (p< 0.05). The cost of urokinase averaged $2653 -+ $243 in the thrombolytic group, constituting a mean of 53% of the total pharmacy costs for each patient. The total hospital charges were similar in the two groups, averaging just more than $38,000 (Table III). By contrast, the total profession charges were lower in the thrombolytic group, $2,445 + $315 versus $3517 + $318 (p < 0.05). The patients in the thrombolytic group had a mean of 1.3 primary and 1.4 secondary medical consultations during hospitalization (mean charges, $185 -+ $76) compared with 1.8 and 3.0 in the operative patients (mean charges, $314 + $100, p < 0.05). When the professional charges were subgrouped by specialty, there were significantly higher mean charges in the operative group for surgical and anesthesiology charges and significantly higher mean charges in the thrombolytic group for radiologic charges. Overall, the total charges were remarkably similar between the two treatment groups, averaging $40,823 + $8,764 in the throm-
Surgery Volume 118, Number 5
bolytic group and $41,930 -+ $10,398 in the operative group. Cost effectiveness. For the cost-effectiveness analysis, the hospital cost data were added to the professional fee data to approximate the total economic impact of treatment. The life expectancy data were culled from standard life expectancy tables from the Statistical Abstract of the United States (1989). As illustrated in Table IV, operative therapy was associated with an increase of greater than 40% in the cost per life saved and nearly 90% in cost per life-year saved when compared with thrombolytic therapy. The marginal cost-effectiveness of a therapy is defined as that sum required to save one additional life or one additional life-year when using one intervention over another. Thrombolysis was associated with a savings of $20,787 per life saved and $3,509 per life-year saved. Compared with the cost-effectiveness of other commonly used therapeutic strategies, "~ both thrombolysis and operation fared well (Table V), with a relatively low cost to achieve a satisfactory patient outcome. DISCUSSION
Previously, we documented clinical benefits associated with thrombolytic treatment of acute peripheral arterial occlusion. An economic analysis of the data confirmed that these benefits were achieved without an increase in the cost of patient care. A slightly increased hospital cost in the thrombolytic group was offset by lower professional fees, so the total economic impact of thrombolysis approximated that of initial operative therapy. The initial treatment of patients randomized to thrombolytic intervention was urokinase infusion, followed by operative or endovascular repair of defects unmasked after dissolution of thrombus. In the present trial, operative repair was used much more frequently than balloon angioplasty. The length of hospitalization and total hospital costs were significantly shorter in the patients who underwent balloon angioplasty rather than surgical correction of an unmasked lesion. O n e criticism of the study relates to this disparity; the cost of therapy might have been lower if endovascular methods were used more frequently. O u r previous report suggested that success was unlikely if no degree of reperfusion had been accomplished after 18 hours of therapy. ~ The mean cost of urokinase could have been reduced by more than $2600 if patients without any degree of reperfusion after 18 hours of infusion were excluded. Similarly, the average hospital costs were significantly reduced in the patients without reperfusion at 18 hours, suggesting that total cost of patient care might have been significantly reduced if thrombolysis had been abandoned if successful dissolution had not begun by 18 hours.
Ouriel et al.
813
T a b l e V. Comparison of cost per life-year saved for variety of medical interventions
Condition
Intervention
Hypercholesterolemia~ Myocardial infarction 7 Hypertension 6
Cholestyramine rt-PA Captoril and diuretic Operation Thrombolysis (urokinase)
Acute limb ischemia Acute limb ischemia
Cost per life-year saved ($)
190,000 60,478 23,760 7,489 3,980
Janosik et al. s reported their experience with urokinase and streptokinase and surgical thrombectomy in the treatment of patients with peripheral arterial occlusion between 1981 and 1985. The mean length of hospital stay in this study averaged 3 weeks for those receiving streptokinase and surgical thrombectomy compared with only 11.5 days for those treated with urokinase. The total hospital charges, however, did not differ in the three groups, averaging between $20,000 and $30,000 1983 U.S. dollars. The results of this study must be taken in the context of its retrospective design; patients receiving streptokinase were treated exclusively during the early years of the series and patients receiving urokinase were treated during the latter years. A study by van Breda et al. 9 summarized the cost-effectiveness of streptokinase and urokinase in 271 patients undergoing thrombolysis for peripheral arterial occlusions between 1979 and 1987. Successful clot lysis was achieved more frequently in the patients receiving urokinase, at a hospital cost averaging approximately $15,000 per therapeutic success. As in the report by Janosik et al., s however, patients receiving streptokinase were treated during the earlier years of the experience. O u r trial differs from the studies ofJanosik et al. s and van Breda et al. 9 in its prospective, randomized design. In patients assigned to thrombolytic or surgical management, we noted no difference in the length of stay or the hospital charges between the groups. Between 1985, the last year of theJanosik series, and 1991, the last year of our series, the mean length of stay decreased considerably. Despite a shorter length of stay in the more recent series, hospital charges did not differ significantly (corrected for inflation) between 1983 and 1991. In summary, the findings of the present study suggest that the economic impact of thrombolyfic therapy as the initial intervention in patients with acute peripheral arterial occlusion is equivalent to that of immediate operation. The improved clinical outcome in patients treated with thrombolysis suggests that this modality may be the appropriate initial therapeutic intervention
814
Ouriel et al.
in the select g r o u p o f patients s e e n within the first few h o u r s o f an acute p e r i p h e r a l arterial occlusive event. T h e s e results, however, m u s t n o t be g e n e r a l i z e d to the larger p o p u l a t i o n o f patients p r e s e n t i n g with occlusions o f a m o r e c h r o n i c nature. R e c o m m e n d a t i o n s o n the m o s t a p p r o p r i a t e therapy in select subsets o f patients m u s t await the results o f large, well-controlled clinical trials. Until such data b e c o m e available, patients s h o u l d be treated o n an individual basis, relying o n the clinical expertise o f the surgical a n d a n g i o g r a p h i c p a t i e n t care teams. REFERENCES
1. Cliffton EE, Grossi CE. Investigations of intravenous plasmin (fibrinolysin) in humans; physiologic and clinical effects. Circulation 1956;14:919-29. 2. Tillett WS,Johnson AJ, McCartyWR. The intravenous infusion of the streptococcal fibrinolytic principle (Streptokinase) into patients. J Clin Invest 1955;34:169-85. 3. Ouriel K, Shortell CK, Marder VJ, et al. A comparison of throm-
Surgery November 1995 bolytic therapy with operative revascularization in the initial treatment of acute peripheral arterial ischemia. J Vasc Surg 1994;19:1021-30. 4. The STILE Investigators. Results of a prospective, randomized trial evaluating surgery vs. thrombolysis for ischemia of the lower extremity: the "STILE" trial. Ann Surg 1995;222(suppl 3):251-68. 5. Cairns JA, Collins R, Fuster V, Passamani ER. Coronary thrombolysis. Chest 1989;95(suppl):73-87. 6. Martens LL, Rutten FFH, Erkelens DW, Ascoop CAPL. Cost effectiveness of cholesterol lowering therapy in the Netherlands: simvastatin versus cholestyramine. Am J Med 1989;87(suppl 4A) :54s-Ss. 7. Guerci AD. A cost effectiveness analysis of t-PA. Clinical Courier 1989;7:5:5-8. 8. Janosik JE, Bettmann MA, Kaul AF, Souney PF. Therapeutic alternatives for subacute peripheral arterial occlusion: comparison by outcome, length of stay, and hospital charges. Radiology 1991;26:921-5. 9. van Breda A, Graor RA, Katzen BT, Risius B, Gillings D. Relative cost-effectiveness of urokinase versus streptokinase in the treatment of peripheral vascular disease.J Vasc Interv Radio11991;2:7787.
SURGERYis abstracted a n d / o r indexed in Index Medicus, Science Citation Index, Current Contents/ Clinical Medicine, Current Contents~Life Sciences, Excerpta Medica, and MEDLINE. This JOURNAL has been registered with Copyright Clearance Center, Inc., 222 Rosewood Dr., Danvers, MA 01923. Consent is given for the copying of articles for personal or internal use of specific clients. This consent is given on the condition that the copier pay directly to the Center the per-copy fee stated on the first page of each article for copying beyond that permitted by U.S. Copyright Law. This consent does not extend to o t h e r kinds of copying, such as for general distribution, resale, advertising and promotional purposes, or for creating new collective works. All inquiries regarding copyrighted material from this publication o t h e r than those that can be handled through Copyright Clearance Center should be directed to Associate J o u r n a l Publisher, Permissions, Mosby-Year Book, Inc., 11830 Westline Industrial Dr., St. Louis, MO 63146; p h o n e (314)579-2808; fax (314)432-1380.