Journal of Vascular Surgery
Volume 46, Issue 4 , Pages 701-708.e2, October 2007

Prospective decision analysis for peripheral vascular disease predicts future quality of life

Presented at the Thirty-first Annual Meeting of the Southern Association for Vascular Surgery, Rio Mar, Puerto Rico, Jan 18, 2007.

Medical University of South Carolina and the Ralph H. Johnson Department of Veterans Affairs Medical Center, Charleston, SC.

Received 13 January 2007; accepted 31 May 2007. published online 03 September 2007.

Article Outline

Objective

Decision making for peripheral vascular disease can be quite complex as a result of pre-existing compromise of patient functional status, anatomic considerations, uncertainty of favorable outcome, medical comorbidities, and limitations in life expectancy. The ability of prospective decision-analysis models to predict individual quality of life in patients with lower extremity arterial occlusive disease was tested.

Methods

This was a prospective cohort study. The settings were university and Veterans Administration vascular surgery practices. All 214 patients referred with symptomatic lower extremity arterial disease of any severity over a 2-year period were screened, and 206 were enrolled. A Markov model was compared with standard clinical decision-making. Utility assessment and generalized (Short Form-36; SF-36) and disease-specific (Walking Impairment Questionnaire; WIQ) quality of life were derived before treatment. Estimates of treatment outcome probabilities and intended and actual treatment plans were provided by attending vascular surgeons. The main outcome measures were generalized (SF-36) and disease-specific (WIQ) variables at study entry and at 4 and 12 months.

Results

Primary intervention consisted of amputation for 9, bypass for 42, angioplasty for 8, and medical treatment for 147 patients. Considering all patients, no improvement in mean overall patient quality of life measured by the SF-36 Physical Component Score (27 ± 8 vs 28 ± 8; P = .87) or WIQ (39 ± 22 vs 39 ± 23; P = .13) was noted 12 months after counseling and treatment by the vascular surgeons. Individually considered SF-36 categories were improved only for Bodily Pain (40 ± 23 vs 49 ± 25; P = .03), with the most significant improvement observed among patients with the most severe pain (68 ± 25 vs 37 ± 23; P = .02) and among those undergoing bypass (60 ± 29 vs 31 ± 22; P = .02). It is noteworthy that when the treatment chosen was incongruent with the Markov model, patients were more likely to report a poorer quality of life at 1 year (Physical Component Score, 25 ± 8 vs 29 ± 8; P < .001). The quality of life predicted at baseline by the Markov model correlated positively with the Physical Component Score (r = 0.23), Bodily Pain (r = 0.33), and Fatigue (r = 0.44) and negatively with WIQ (r = −0.08) observed 1 year later.

Conclusions

Prospective application of an individualized decision Markov model in patients with vascular disease was predictive of patient quality of life at 1 year. The patient’s outcome was worse when the treatment received did not follow the model’s recommendation. This decision analysis model may be useful to identify patients at risk for poor outcomes with standard clinical decision making.

 

Graft or arterial occlusion and progressive compromise of distal arterial outflow threaten patient quality of life after intervention for peripheral arterial occlusive disease of the lower extremities.1, 2, 3 Wound breakdown, infection, or extensive gangrene may mandate amputation even when the reconstruction remains patent. Few patients who are nonambulatory before intervention resume their ambulatory status.4, 5 Pre-existing compromise of patient functional status, anatomic considerations, uncertainty of favorable outcome, medical comorbidities, and limitations in patient life expectancy all serve to make the decision to intervene quite complex. Typically, surgeons advise their patients on the basis of published reports of the outcomes of intervention melded with their own subjective assessment of the patient’s needs. Unfortunately, this strategy is vulnerable to improper interpretation and application of the available data by the surgeon and unrealistic expectations by the patient.

Various methods to codify vascular treatment decisions to optimize patient quality of life have been reported.6, 7 Decision analysis is one set of tools developed by health care researchers to evaluate complex medical decisions.8, 9, 10, 11 We have previously examined the use of individualized decision analysis in patients with chronic peripheral vascular occlusive disease scheduled to undergo bypass operation.12 Among this group, the chance of a favorable long-term outcome was much greater when the decision analysis model agreed with bypass as the best therapy compared with when it favored a different intervention (80% vs 50% “good”), thus identifying a subgroup of patients less likely to benefit from bypass operation. We subsequently expanded the use of this model to include a broader spectrum of severity of peripheral arterial occlusive disease in patients treated only medically, those eligible for endovascular procedures, and those deemed to require primary amputation.13 Preliminary experience with this Markov model favored more use of angioplasty and amputations and less use of conservative medical therapy than proposed by the vascular surgeons, although the actual treatment provided frequently did not follow the preference of the decision analysis model. This study represents an extension of that preliminary experience, with specific objectives of determining whether the Markov model could accurately predict the actual quality of life experienced by these patients according to their type of intervention and whether this intervention resulted in an improvement in patient quality of life.

Back to Article Outline

Methods 

Study entry 

All patients referred for evaluation of signs and symptoms of chronic lower extremity ischemia to the vascular surgery services at the Medical University of South Carolina and the Ralph H. Johnson Department of Veterans Affairs Medical Center over a 28-month period were screened for participation in this prospective study. Inclusion criteria included the presence of symptomatic arterial occlusive disease as manifested by intermittent claudication, resting ischemia pain, or ischemic tissue loss with ulcers or gangrene. Exclusion criteria included the inability of the patient, by his or her own admission, to adequately comprehend the study. The experimental protocol was approved by the institutional review board and performed in accordance with institutional guidelines, which are in accordance with the Helsinki Declaration of 1975. Informed consent was obtained from all patients after being seen in consultation by one of the three participating vascular surgeons. After enrollment, patients were randomized by computerized random-number generation for their attending surgeon to be made aware or remain unaware of the results of decision analysis with the Markov model. This article describes the follow-up and quality-of-life data from this group of patients previously reported.13 Enrollment occurred between July 1999 and November 2001, with the final 12-month evaluation completed in November 2002.

Decision-analysis model 

Patient-specific utility values were assessed before surgery and at 4 and 12 months after surgery by using visual analog scales and time trade-off methods.10 The probabilities of success for the various potential outcomes in the model were estimated before surgery for each patient by the responsible attending surgeon. The mean, range, and standard deviation for each of the utility values and probability estimates used in the decision analysis model are included in Appendix Table I (online only), Appendix Table II (online only). Individualized decision analysis consisted of a Markov model constructed using the DATA 3.5 software package (TreeAge Software Inc, Williamston, Mass). Four therapeutic interventions were considered by the computer model: primary amputation, bypass operation, percutaneous balloon angioplasty, and medical management. This outcome was expressed in quality-adjusted life-years.

The Markov decision analysis model has been previously described.13 For the choice of primary amputation, four transitional states would be possible initially: healing of the wound and the ability to ambulate with a prosthesis, healing of the wound without the ability to ambulate, nonhealing of the wound, and perioperative death (Fig 1). Within a specified cycle interval, patients would be predicted to have a finite probability to either die or remain alive in the same state or, alternatively, pass into another transitional state. For the choice of bypass, initial transitional states included hemodynamically satisfactory bypass with healing, unsatisfactory bypass without improvement and/or healing, early thrombosis of the bypass, and perioperative death. Transitional states of being ambulatory or nonambulatory with a healed amputation, for example, were not available as an initial result of the bypass. However, they were included in this branch of the decision tree, because they represent potential states after amputation should the bypass not prove to be satisfactory and not be revised. These states subsequently acted as clones of the Primary Amputation subtree, as is indicated by the letter C, thus allowing the extended decision tree to be illustrated more easily. For the choice of angioplasty, initial transitional states included hemodynamically satisfactory angioplasty, unsatisfactory angioplasty, and periprocedural death (Fig 2). The initial transitional state associated with nonoperative medical management assumed that the patient would remain the same. This led to the possible outcomes of dying during the course of the first cycle; requiring primary amputation, bypass operation, or angioplasty; or remaining the same. The transitional states associated with these other management options are again included in this subtree. The cumulative time in each transitional state earned quality-of-life points based on the value placed on that state as determined by the utility assessment. Utilities were discounted at a rate of 5% per annum according to the authors’ best estimate of overall deterioration in quality of life for patients of this age group related to further aging but independent of vascular disease. A disutility (negative utility) of −0.10 was assigned for the performance of each additional operation and −0.08 for each additional angioplasty according to the authors’ best estimate of the negative quality-of-life effect of repeated interventions. The Markov rollback used monthly cycles and was terminated when 99.9% of the model was absorbed within the transitional state for death or at a maximum of 600 cycles, corresponding to 50 years after initial intervention.

  • View full-size image.
  • Fig 1. 

    Markov subtrees for primary amputation and bypass operation. Transitional states for amputation and amputation reflect the potential for these outcomes. Circled C indicates a clone of named subtrees for amputation or angioplasty. #Remaining outcome probability calculated as a residual from the probability of all other events.

  • View full-size image.
  • Fig 2. 

    Markov subtrees for angioplasty and nonoperative medical management. Transitional states for amputation, bypass, and angioplasty reflect the potential for these outcomes. Circled C indicates a clone of named subtrees for amputation, bypass, or angioplasty. #Remaining outcome probability calculated as a residual from the probability of all other events.

Quality of life and intervention 

Immediately after enrollment, generalized quality of life (Short Form-36; SF-36) and disease-specific quality of life (Walking Impairment Questionnaire [WIQ], value range, 0-70) were entered by patients, supervised by a trained registered nurse, on a touch-screen computer. Individual health concept items of the SF-36, which range from 0 to 100, are specifically weighted and combined to yield overall Physical and Mental Component Scores, with mean (range) values of 43.5 (8-59) and 52.6 (21-74) for the US general population aged 65 to 74 years.14 The initial plan for treatment was recorded during the initial visit, and the final treatment plan was recorded 1 week later, after the surgeon was informed or not informed (according to randomization) of the predictions of the Markov model. Quality-of-life assessments were repeated at 4 months and at 1 year after enrollment by using the same tools as described previously.

Statistical analysis 

Descriptive analysis included mean and standard deviation for demographic and patient history data. Correlation of data was performed by using Pearson two-tailed correlation analysis. Comparisons of individual interval data were assessed by using t tests or analysis of variance. Proportional data were compared by using χ2 tests. Agreement between models, decisions, and interventions was assessed by using the κ statistic.

Back to Article Outline

Results 

From July 1999 through November 2001, 214 patients were screened, 208 gave consent, and 205 were enrolled with data available. The average patient age at enrollment was 63 ± 15 years (mean ± SD). Most patients were white (58%) and male (70%). The most severe symptom of peripheral vascular disease was claudication in 58%, rest pain in 38%, and tissue loss in 14%. A minority of patients had experienced a prior lower extremity vascular bypass operation (28%) or prior major amputation (7%), whereas 32% had a family history of amputation for vascular disease. Twenty-four patients died during the first year of follow-up.

The types of operation recommended by the Markov model and those actually performed have previously been reported.13 To summarize, the optimal treatments predicted by decision analysis differed significantly from the surgeon’s initial plan and consisted of bypass for 30% and 29% for the model and surgeon plan, respectively; angioplasty for 28% and 11%; amputation for 8% and 6%; and medical management for 34% and 54% (agreement, 50%; κ = 0.28). Patients for whom the model agreed with the surgeon’s initial plan were less likely to undergo bypass (13% vs 30%; P < .01). The study was originally designed to test whether surgeon awareness of the Markov model results might alter the treatment plan, but this hypothesis could not be proven. Surgeon awareness of the decision model results did not alter the verbalized final plan, but it did trend toward less frequent use of bypass. The primary intervention actually performed consisted of amputation for 9, bypass for 42, angioplasty for 8, and medical treatment for 146 patients.

Follow-up quality-of-life data were not included in the short-term report, but they can now be reported. Higher SF-36 values indicate better quality of life. In contrast, because higher WIQ values indicate greater impairment in walking, correlations would be expected to be negative to be predictive. In fact, the ability of the model to predict the WIQ was weak (r = −0.08). For the entire cohort of patients, no improvements in mean overall patient quality of life measured by the generalized SF-36 or disease-specific WIQ were noted 12 months after counseling and treatment by the vascular surgeons (Table I). Bodily Pain was the only individual SF-36 category to improve (40 ± 23 vs 49 ± 25; P = .03), whereas observed changes in Physical Functioning (27 ± 19 vs 22 ± 16; P = .36) were not significant. Improvements in Bodily Pain were more likely to be observed among patients with limb-threatening symptoms (68 ± 25 vs 37 ± 23; P = .02) compared with patients presenting with claudication (41 ± 24 vs 44 ± 23; P = .4). This was also observed for patients undergoing bypass (60 ± 29 vs 31 ± 22; P = .02). Despite experiencing a significant improvement in Bodily Pain (54 ± 26 vs 37 ± 22; P < .01) not observed among men (40 ± 24 vs 43 ± 23; P = .5), at 1 year WIQ was worse for women compared with men (49 ± 21 vs 37 ± 22; P < .01).

Table I. For the entire cohort, change in generalized (Short Form-36; SF-36) and disease-specific (Walking Impairment Questionnaire; WIQ) variables after patient enrollment
VariableBaseline4 mo12 mo
Bodily Pain40±2347±2549±25⁎
Energy/Fatigue40±2145±2343±19
Social Function62±3060±2863±27
Mental Health69±2269±2069±19
Role-Physical21±3024±3630±38
Physical Functioning27±1928±2322±16
General Health47±2146±1943±19
Role-Emotional64±4466±4658±45
Physical Component Score27±828±928±8
Mental Component Score51±1251±1151±11
Walking Impairment Questionnaire39±2235±2239±23

Higher SF-36 values indicate better quality of life, whereas higher WIQ values indicate greater impairment in walking.

Data are mean±SD.

⁎P < .05 compared with baseline.

One year after enrollment in the study, a significant correlation was observed between many of the SF-36 quality-of-life parameters and the quality of life predicted at baseline by the Markov model for the intervention that the patient subsequently received (Table II). Clearly, predictions of outcome by the model rely heavily on the method of utility analysis chosen, as confirmed in our previous study.13 The visual analog scale seems to be the simplest and has the highest level of patient comprehension. Furthermore, because use of the visual analog scale to provide the utility assessment for the model seemed to correlate better with actual quality of life, the visual analog scale was used to determine how the model compared with the surgeon’s treatment. It is noteworthy that when the surgeon’s treatment was incongruent with the Markov model, patients were more likely to report a poorer Physical Component Score quality of life both at baseline (28 ± 9 vs 25 ± 7; P < .01) and at 1 year (25 ± 8 vs 29 ± 8; P < .01; Table III). As previously reported, most disagreement was on whether to offer bypass operation, which was the initial intervention performed 30% of the time when the model and surgeons were in agreement but only 13% of the time when the two disagreed (P < .01). Aggressive surgical management did not always seem to benefit patient quality of life at 1 year as measured by these assessment tools.

Table II. Quality of life predicted by the Markov model by using patient-derived individualized visual analog scale (VAS) or time trade-off (TTO) utility values correlated with actual measured (Short Form-36) quality of life 12 months later
VariableVAS (r)TTO (r)
Bodily Pain0.33⁎0.11
Energy/Fatigue0.44⁎0.26
Social Function0.30⁎0.14
Mental Health0.42⁎0.02
Role-Physical0.240.19
Physical Functioning0.29⁎0.09
General Health0.31⁎0.11
Role-Emotional0.24−0.12
Physical Component Score0.230.15
Mental Component Score0.32⁎0.00
Walking Impairment Questionnaire−0.080.05

Data are mean±SD.

⁎P < .05 by two-tailed test of significance.

Table III. Change in generalized (Short Form-36) and disease-specific (Walking Impairment Questionnaire) variables according to agreement of the surgeon’s plan with the model recommendation
VariableBaseline4 mo12 mo
Bodily Pain
Agree with model43±2447±2347±21
Disagree with model35±22⁎47±2748±32
Role-Physical
Agree with model25±3328±3935±39
Disagree with model17±2718±3119±34
Physical Functioning
Agree with model32±2132±2324±16
Disagree with model22±1623±2218±16
Physical Component Score
Agree with model28±929±929±8
Disagree with model25±7⁎27±925±8⁎
Walking Impairment Questionnaire
Agree with model32±2132±2036±23
Disagree with model48±2139±2648±24

Data are mean±SD.

⁎P < .05, “agree” vs “disagree.”

Back to Article Outline

Discussion 

Over the last decade, there has been increasing interest in demonstrating that interventions performed in patients with peripheral arterial disease do improve the quality of life for those patients.4, 15, 16, 17, 18 Operative revascularization has been found to be superior to primary amputation in many studies.19, 20, 21, 22, 23, 24, 25 Such observations have not always been consistent, because others have minimized the quality-of-life benefit from revascularization for leg ischemia.5, 15, 17 Benefits are often delayed during the recovery period from operation.26 Indeed, few patients report feeling back to their normal baseline level of health for the first 6 months after infrainguinal bypass.20 Concurrent medical illness and complex personal, financial, and social factors may distract them from any perceived improvement in quality of life. By contrast, their expectations for intervention seem to be high. Two of three patients believe that their ability to perform daily activities, leg pain, and walking ability will improve, whereas most have a poor understanding of their own long-term survival.27 These patients seem willing to undertake significant operative risk with, perhaps, unrealistic expectations for the achievable results.

Several authors have reported the development of standardized algorithms for treatment of peripheral arterial disease. Some may be limited to specific clinical scenarios, such as defining the conditions under which to perform angioplasty with or without a covered stent vs bypass in elderly patients according to the location and length of the stenosis.7 By this algorithm, longer occlusions would be bypassed and shorter stenoses in the femoral and popliteal arteries would be managed by a covered stent. Other algorithms are much more broad in scope and, thereby, more clinically useful. The Lower Extremity Grading System (LEGS) score favors primary amputation in patients with lower pre-existing functional status, more medical comorbidities, compromised venous conduit, and (paradoxically) less severe ischemia and a shorter extent of disease.28 Open surgical revascularization is offered for good-risk patients with fewer medical comorbidities who are ambulatory at the time of presentation and exhibit more extensive arterial involvement. Endovascular techniques are generally considered best suited for patients residing between these extremes.6 Application of this relatively simple system of interventional standardization has demonstrated excellent results, including significant improvement in SF-36 patient quality of life. Although these results are quite encouraging, we have chosen a different approach—namely, the burgeoning science of decision analysis geared toward individual patients.

Decision analysis has frequently been used to justify the benefits of intervention in large populations of patients with peripheral arterial occlusive disease of the lower extremities, but it has rarely been used clinically as a tool to aid in the decision-making process.29, 30, 31 Formal decision analysis combines the probabilities of various potential outcomes of an intervention with the value of these outcomes to patients.8 This type of analysis goes beyond physician assessment of the patient’s level of function as a proxy for their quality of life, and it actually incorporates patient values and preferences for outcome. The greatest challenge for decision analysis is accurate and fair assessment of patient utility value scores.32 Utility values are heavily influenced during the process of assessment, so that utility values elicited by using different assessment strategies may not be directly comparable.33, 34 It has been shown that utility values are best derived from the patients themselves, because values derived from either treating surgeons or patient surrogates inaccurately predict the patient’s actual views.35, 36, 37 The challenge of accurate derivation of patient utility values remains a controversial yet critical component to the success of the model. Because it correlates most directly with patient quality of life and because it seems to be comprehensible to most patients, we have chosen to use utility values derived by the visual analog scale. The Markov model is also quite dependent on the accuracy of the estimates of probability for various outcomes of intervention. Algorithms derived from large databases such as the National Surgical Quality Improvement Project may improve the accuracy of this estimate, at least regarding mortality. The evolution of endovascular means to manage increasingly complex peripheral vascular problems will also affect the estimated rate of success for these procedures, such that the “angioplasty” option might better be considered “endovascular.” Fundamentally, to demonstrate the validity of the individualized Markov decision analysis model, we must first show it to accurately predict the actual quality of life experienced by the patient as measured by generalized and disease-specific instruments. In fact, this study has demonstrated that a Markov decision model can be predictive of the quality of life after intervention for these patients, although that ability is limited, with even the highest correlation (r = 0.44), with the Energy and Fatigue component only expressing a 20% predictive value. It is interesting to note that when the treatment provided followed the Markov model, patients reported a better quality of life at 1 year. This observation suggests that the lack of congruency between the model and the surgeon’s initial plan identifies a group of patients who have the most to gain by reconsidering a different course of treatment.

However, it is especially noteworthy that these observations suggest that regardless of intervention, the effects on overall patient quality-of-life improvement are relatively minute. This is somewhat at odds with the LEGS report, although their study included only patients who were candidates for intervention and not conservative medical therapy alone.28 In fact, compared with our observations, their patients scored much lower than those in this study at baseline in the SF-36 categories of Social Function, Role Emotional, Mental Health, and General Health, among others. In this study, improvement in quality of life was limited to those patients with the most severe symptoms and those undergoing bypass operations. The fact that most of our patients did not undergo direct intervention may well affect the lack of greater improvement in the group overall. Although only 29% had direct intervention, the remainder were followed up medically, which is certainly an important option for patients referred with vascular disease. In fact, much of the disagreement between the surgeon’s plan and the model’s recommendation was that the surgeon recommended medical therapy far more often, whereas the model suggested either amputation or angioplasty. It is across these broad options for treatment that the model may be most useful.

The WIQ may be more sensitive to patients with claudication and less so for limb-threatening ischemia, and it has recently been confirmed that the SF-36 is less sensitive for detecting patient changes after intervention for chronic lower limb ischemia compared with other measures such as the Nottingham Health Profile or the VascuQol.38, 39 These quality-of-life indicators are disease specific and, therefore, more sensitive to vascular disease. Had we used these tools in this study, it is assumed that a greater difference would have been detected as a result of direct intervention or even medical therapy. However, because these tools are much less generalizable for purposes of comparison to other disease states in terms of overall value gained, it can be argued that the need to use alternative measures to detect improvement simply reinforces the original observation that the effects on overall patient quality of life are relatively small. This study identified Bodily Pain to be the only parameter that consistently improved, and even then it did so only in patients with the most severe symptoms or after bypass. Although it may be assumed that most surgeons believe that the effect of our interventions is great, these observations should help put the value of our interventions into perspective.

Back to Article Outline

Author contributions 


Conception and design: TEB, JGR, BME

Analysis and interpretation: TEB, JGR, BME

Data collection: TEB, JGR, BME

Writing the article: TEB

Critical revision of the article: TEB, JGR, BME

Final approval of the article: TEB, JGR, BME

Statistical analysis: TEB

Obtained funding: TEB

Overall responsibility: TEB

Back to Article Outline

 

The assistance of Montgomery H. Cox, MD, in the conception, design, and data collection for this work is gratefully acknowledged.

Back to Article Outline

Appendix 

Additional material for this article may be found online at www.jvascsurg.org.

Appendix Table I (online only). Pooled values of utility scores derived from individual patients
VariableVASTTOP value⁎
Bypass successful0.67(0.25;0.25-0.86)0.85(0.26;0.4-1.0)<.001
Stayed the same0.53(0.26;0.19-0.72)0.77(0.31;0.3-0.9)<.001
Bypass failed0.42(0.26;0.06-075)0.71(0.35;0.1-0.9)<.001
Amputate, ambulate0.39(0.28;0.10-0.65)0.71(0.35;0.2-0.9)<.001
Amputate, nonambulate0.26(0.25;0.05-0.65)0.60(0.39;0.2-0.8)<.001
Amputate, nonhealed0.20(0.24;0.02-0.52)0.52(0.41;0.1-0.8)<.001

Data are mean (SD;range).

⁎Paired t-test.

Appendix Table II (online only). Pooled values of estimates of outcome probabilities for input into the model
Potential outcomeMeanSD (range)
Patient mortality (annual)⁎0.130.04(0.05-0.50)
Primary amputation
Amputation healed, ambulatory0.460.18(0.00-0.90)
Amputation not healed0.110.04(0.01-0.40)
Death from amputation/reamputation0.040.03(0.01-0.30)
Stay healed ambulatory⁎0.150.19(0.02-1.00)
Wound breakdown⁎0.100.02(0.05-0.30)
Reamputation healed, ambulatory0.400.16(0.00-0.75)
Reamputation not healed0.110.04(0.05-0.40)
Bypass operation
Bypass satisfactory0.870.13(0.02-0.98)
Early thrombosis after bypass⁎0.140.14(0.02-0.98)
Death from bypass/rebypass0.030.05(0.00-0.50)
Late thrombosis†0.510.17(0.02-0.95)
Amputation for unsatisfactory bypass0.290.16(0.01-0.90)
Rebypass satisfactory0.770.14(0.00-0.95)
Early thrombosis after rebypass0.250.17(0.00-0.99)
Angioplasty after bypass failure0.100.06(0.00-0.50)
Angioplasty
Angioplasty satisfactory0.890.18(0.00-0.99)
Early thrombosis after angioplasty⁎0.440.31(0.00-1.00)
Death from angioplasty/reangioplasty0.010.01(0.00-0.05)
Late thrombosis†0.820.14(0.20-1.00)
Reangioplasty satisfactory0.160.12(0.00-0.60)
Bypass for unsatisfactory angioplasty0.850.21(0.00-0.98)
Early thrombosis after reangioplasty0.570.27(0.00-1.00)
Nonoperative management
Symptoms do not worsen†0.440.32(0.00-0.95)
Delayed angioplasty⁎0.160.19(0.00-0.80)
Delayed bypass⁎0.600.32(0.00-1.00)

Data are mean (SD;range).

⁎One-year estimate.

†Five-year estimate. These estimates were used to derive monthly probability estimates.

Back to Article Outline

References 

  1. Brothers TE, Robison JG, Elliott BM, Arens C. Is infrapopliteal bypass compromised by distal origination of the proximal anastomosis?. Ann Vasc Surg. 1995;9:172–178
  2. Elliott BM, Robison JG, Brothers TE, Cross MA. Limitations of peroneal artery bypass grafting for limb salvage. J Vasc Surg. 1993;18:881–888
  3. Robison JG, Ross JP, Brothers TE, Elliott BM. Distal wound complications following pedal bypass: analysis of risk factors. Ann Vasc Surg. 1995;9:53–59
  4. Abou-Zamzam AM, Lee R, Moneta GL, et al. Functional outcome after infrainguinal bypass for limb salvage. J Vasc Surg. 1997;25:287–297
  5. Duggan MM, Woodson J, Scott TE, et al. Functional outcomes in limb salvage vascular surgery. Am J Surg. 1994;168:188–191
  6. Taylor SM, Kalbaugh CA, Gray BH, et al. The LEGS score: a proposed scoring system to grade and treat chronic lower extremity ischemia. Ann Surg. 2003;237:812–818
  7. Killewich LA. Improving functional status and quality of life in elderly patients with peripheral arterial disease. J Am Coll Surg. 2006;202:345–355
  8. Eiseman B. Surgical decision making and elderly patients. Bull Am Coll Surg. 1996;81:8–11
  9. Clarke JR. The application of decision analysis to clinical medicine. Interfaces. 1987;17:27–32
  10. Weinstein MC. Utility analysis (clinical decisions involving many possible outcomes). In:  Weinstein MC,  Fineberg HV editor. Clinical decision analysis. Philadelphia: WB Saunders; 1980;p. 184–227
  11. Birkmeyer JD, Welch HG. A reader’s guide to surgical decision analysis. J Am Coll Surg. 1997;184:589–595
  12. Brothers TE, Rios GA, Robison JG, Elliott BM. Predicting outcome for arterial bypass using individualized decision analysis. Evidence-Based Surg. 2003;1:39–47
  13. Brothers TE, Cox MH, Robison JG, Elliott BM, Nietert P. Prospective decision analysis modeling indicates that clinical decisions in vascular surgery often fail to maximize patient expected utility. J Surg Res. 2004;120:278–287
  14. Ware JE, Kosinski M, Keller SD. SF-12: how to score the SF-12 Physical and Mental Health Summary scales. 2nd ed.. Boston (MA): The Health Institute, New England Medical Center; 1995;
  15. Albers M, Fratezi AC, DeLuccia N. Assessment of quality of life of patients with severe ischemia as a result of infrainguinal arterial occlusive disease. J Vasc Surg. 1992;16:54–59
  16. Crosby FE, Ventura MR, Frainier MA, et al. Well-being and concerns of patients with peripheral arterial occlusive disease. J Vasc Nurs. 1993;11:5–11
  17. Chetter IC, Dolan P, Spark JI, et al. Correlating clinical indicators of lower limb ischaemia with quality of life. Cardiovasc Surg. 1997;5:361–366
  18. Guralnik JM, Ferrucci L, Simonsick EM, et al. Lower-extremity function in persons over the age of 70 years as a predictor of subsequent disability. N Engl J Med. 1995;332:556–561
  19. Paaske WP, Lausten J. Femorodistal bypass grafting: quality of life and socioeconomic aspects. Eur J Vasc Endovasc Surg. 1995;10:226–230
  20. Gibbons GW, Burgess AM, Guadagnoli E, et al. Return to well-being and function after infrainguinal revascularization. J Vasc Surg. 1995;21:35–45
  21. Schneider JR, McHorney CA, Malenka DJ, et al. Functional health and well-being in patients with severe atherosclerotic peripheral vascular occlusive disease. Ann Vasc Surg. 1993;7:419–428
  22. Pell JP. Impact of intermittent claudication on quality of life. Eur J Vasc Surg. 1995;9:469–472
  23. Kaira HS, Hanger R, Shearman CP. Quality of life in patients with intermittent claudication. Eur J Vasc Endovasc Surg. 1996;11:65–69
  24. Thompson MM, Sayers RD, Reid A, et al. Quality of life following infragenicular bypass and lower limb amputation. Eur J Vasc Endovasc Surg. 1995;9:310–313
  25. Hosie KB, Kockelberg R, Newbury-Ecob RA, et al. A retrospective review of the outcome of patients over 70 years of age considered for vascular reconstruction in a district general hospital. Eur J Vasc Surg. 1990;4:313–315
  26. Mangione CM, Goldman L, Orav EJ, et al. Health-related quality of life after elective surgery: measurement of longitudinal changes. J Gen Intern Med. 1977;12:686–697
  27. Schwarze ML, Sayla MA, Alexander GC. Exploring patient preferences for infrainguinal bypass operation. J Am Coll Surg. 2006;202:445–452
  28. Taylor SM, Kalbaugh CA, Blackhurst DW, et al. Determinants of functional outcome after revascularization for critical limb ischemia: an analysis of 1000 consecutive vascular interventions. J Vasc Surg. 2006;44:747–756
  29. Brothers TE, Rios GA, Robison JG, et al. Justification of intervention for limb-threatening ischemia: a surgical decision analysis. Cardiovasc Surg. 1999;7:62–69
  30. Brothers TE, Robison JG. Assessing outcomes, function, and quality of life following vascular surgery: introducing new tools and trends. Perspect Vasc Surg. 1997;9:20–43
  31. Brothers TE, Frank CE, Frank B, et al. Is duplex venous surveillance worthwhile after arthroplasty?. J Surg Res. 1996;67:72–78
  32. Lenert LA, Cher DJ, Goldstein MK, et al. The effect of search procedures on utility elicitations. Med Decis Making. 1998;18:76–83
  33. Llewellyn-Thomas H, Sutherland HJ, Tibshirani R, et al. The measurement of patient’s values in medicine. Med Decis Making. 1982;2:449–462
  34. Froberg DG, Kane RL. Methodology for measuring health-state preferences: II—scaling methods. J Clin Epidemiol. 1989;42:459–471
  35. Rios GA, Brothers TE, Robison JG, Elliott BM. The surgeon’s decision for lower extremity bypass operation frequently disagrees with formal surgical decision analysis. Surg Forum. 1998;49:556–557
  36. Tsevat J, Dawson NV, Wu AW, et al. Health values of hospitalized patients 80 years or older. JAMA. 1998;279:371–375
  37. Enemark U, Lyttkens CH, Troeng T, et al. Implicit discount rates of vascular surgeons in the management of abdominal aortic aneurysms. Med Decis Making. 1998;18:168–177
  38. Klevsgard R, Froberg BL, Risberg B, Hallberg IR. Nottingham Health Profile and Short-Form 36 Health Survey questionnaires in patients with chronic lower limb ischemia: before and after revascularization. J Vasc Surg. 2002;36:310–317
  39. Nguyen LL, Moneta GL, Conte MS, Bandyk DF, Clowes AW, Seely BL. Prospective multicenter study of quality of life before and after lower extremity vein bypass in 1404 patients with critical limb ischemia. J Vasc Surg. 2006;44:977–984

 Supported by an Established Investigator Award of the American Heart Association and by the Office of Research and Development, Medical Research Service, Department of Veterans Affairs.

 Competition of interest: none.

 Additional material for this article may be found online at www.jvascsurg.org.

PII: S0741-5214(07)00975-5

doi:10.1016/j.jvs.2007.05.045

Journal of Vascular Surgery
Volume 46, Issue 4 , Pages 701-708.e2, October 2007

Article Tools

* Image Usage & Resolution