Peripheral Vascular Disease
Barthel Index and Outcome of Open Bypass for Critical Limb Ischemia
2018 Volume 82 Issue 1 Pages 251-257
Details
2018 Volume 82 Issue 1 Pages 251-257
Background: Daily activity should be an important factor to consider when deciding on a treatment strategy for critical limb ischemia (CLI), and we hypothesized that there was a close relationship between activities of daily living (ADL) and prognosis. The aim of this study was to investigate the association between ADL and outcomes of open bypass for CLI.
Methods and Results: A total of 226 patients undergoing infrainguinal bypass for CLI between January 2005 and December 2015 were reviewed. They were divided into 2 groups based on Barthel index (BI) at admission (pre-BI; ≥60 and <60) and compared with respect to the incidence of early adverse events and 3-year overall survival and amputation-free survival. Adverse events were more frequently observed in patients with pre-BI <60. Patients with pre-BI <60 had worse long-term outcomes. On multivariable analysis, hypoalbuminemia, end-stage renal disease, and BI at discharge <60 were significant predictors of 3-year outcomes, whereas pre-BI <60 was not a significant predictor of either outcome.
Conclusions: It is not BI at admission, but BI at discharge that determines long-term outcome.
Critical limb ischemia (CLI) is the final stage of peripheral artery disease (PAD) and revascularization is essential to treat CLI completely. When deciding on the revascularization method, it is essential to obtain information about a given patient’s risk factors. Since the BASIL trial demonstrated the superiority of the bypass-first strategy in patients with a life expectancy >2 years,1 many studies have identified a variety of factors as predictors of long-term outcome of revascularization for CLI,2–6 and several scoring systems with risk stratification and confirmed validity regarding the prediction of outcomes have been introduced.5–8
Activities of daily living (ADL) or physical function should be considered when determining the treatment strategy because patients with poor ADL often have significant comorbidities, and there is little expectation of improvement in their ADL or physical function even after revascularization.9 Some studies have reported an association between the physical activity or functional status and mortality in patients with PAD.10–12 Except for the ambulation ability,9 however, a simple but very important index of ADL, no studies have described ADL as a significant factor for long-term outcome in CLI. The Barthel index (BI), a scale measuring disability or dependence in ADL, was originally developed to measure ADL in stroke patients and subsequently its use has been extended to geriatric patients.13 Although not all patients with low ADL are frail, low BI is generally associated with high levels of frailty.14 This suggests that the BI measured at admission (pre-BI) can be used to predict early and long-term outcomes of CLI, which can help guide decision-making concerning the revascularization strategy.
The aim of this study was therefore to examine whether or not pre-BI was associated with the incidence of adverse events during hospitalization and the long-term outcomes of infrainguinal bypass for CLI, and to determine which perioperative factors including pre-BI and the BI measured at discharge (post-BI) affect long-term outcome.
This study was approved by the institutional review board of Saiseikai Yahata General Hospital and Steel Memorial Yawata Hospital. Informed consent was waived for this retrospective review.
Database and Patient SelectionThe data of 303 consecutive patients who underwent infrainguinal bypass for CLI (Rutherford 4–6) due to arteriosclerosis obliterans at Steel Memorial Yawata Hospital between January 2005 and March 2015 and at Saiseikai Yahata General Hospital between April 2015 and December 2015 were reviewed and the clinical records retrieved from electronic medical records. Seventy-four patients with a history of surgical bypass in either lower limb, and 3 patients with a history of major amputation in the contralateral limb, were excluded. Ultimately, 226 patients were enrolled for analysis. With regard to post-BI, 215 patients were enrolled because the ADL of 11 patients who died during hospitalization were not evaluated (Figure 1).
Patient flow. BI, Barthel index; CLI, critical limb ischemia.
The following data were retrieved as variables for analysis: age, gender, body mass index (BMI), Rutherford classification, comorbidities (ambulation status, diabetes mellitus [DM], hypertension [HT], coronary artery disease [CAD], chronic obstructive pulmonary disease [COPD], end-stage renal disease [ESRD], stroke, malignancy, dyslipidemia, and smoking history), preoperative laboratory data (hemoglobin [Hb], serum albumin [Alb]), and BI measured at admission (pre-BI) and at discharge (post-BI).
BI measures 10 basic aspects of activity related to self-care and mobility: feeding; bathing; grooming; dressing; bowels; bladder; toilet use; transfers (bed to chair and back); mobility on level surface; and stairs. The full score is 100, and lower scores indicate greater dependency.
EndpointsOverall survival (OS) and amputation-free survival (AFS) were set as the primary endpoints, and operative death, in-hospital death, postoperative systemic complications during hospitalization, operative wound complications, and major adverse leg event (MALE) within 30 days were set as the secondary endpoints.
BI Cut-offThe BI cut-off was set at 60 because patients with BI <60 were indicated to have functional dependency.15
CLI Strategy and Surveillance MethodPrimary amputation was performed for patients with severe dementia, in a bed-rest state, with irremediable joint contracture, and with large wounds accompanied by uncontrollable infection and sepsis. Otherwise, revascularization was attempted for limb salvage.
The revascularization strategy changed over time. Until March 2010, open surgery was mainly performed; from April 2010 to March 2015, however, an endovascular therapy (EVT)-first strategy was chosen in patients with simple, short, and stenotic lesion in the superficial femoral artery (SFA). EVT for the popliteal or distal artery was not performed in principle. For patients with long occluded lesion in the SFA, open surgery using a vein or prosthetic graft was performed. From April 2015, EVT was actively performed in patients with long occluded lesion in the SFA if an available vein for an arterial conduit could not be found. EVT for the popliteal or distal artery was performed when the lesion was simple and short.
Rehabilitation was started on the next day after surgery and was continued to the day before discharge unless the patient refused it. The criteria for discharge were as follows: (1) wound healing completed or could be predicted without any further invasive treatments (i.e., revascularization or debridement) in patients with tissue loss; (2) all stitches removed; and (3) postoperative complications resolved.
After discharge, the patients were followed up at regular intervals of 2–4 weeks if wound healing had not been completed. After the completion of wound healing, graft surveillance was performed at regular intervals of 1–3 months. Interview by phone was performed for patients who could not visit the hospital. Revision procedures were recommended when obvious stenosis with a peak systolic velocity >300–350 cm/s or velocity ratio >3.0–3.5 was detected on duplex ultrasonography. Percutaneous transluminal angioplasty with a cutting balloon was selected as the first strategy for revision and graft interposition, or elongation of the bypass was performed if the restenosis occurred regardless of repeated revision procedures. If CLI recurred after graft occlusion, redo surgery was performed.
A final review was conducted in February 2017, and up to 3 years’ worth of data were analyzed.
DefinitionsHT was defined as either or both systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg, or ongoing therapy for HT. DM was defined as fasting blood glucose ≥105 mg/dL or ongoing therapy for DM with an oral hypoglycemic agent or insulin injection. CAD was defined as a history of angina pectoris, myocardial infarction, percutaneous coronary intervention, or coronary artery bypass grafting. Stroke was defined as a history of transient ischemic attack, cerebral infarction, or cerebral bleeding. COPD was defined as percent forced expiratory volume in 1 s <70%. ESRD was defined as requirement for hemodialysis or peritoneal dialysis. Dyslipidemia was defined as either or both fasting total cholesterol >220 mg/dL, or triglyceride >207 mg/dL in men, or >137 mg/dL in women, or ongoing therapy for dyslipidemia. Malignancy was defined as a history of any treatment for cancer neoplasm. Major amputation was defined as the loss of the leg above the ankle. Operative death was defined as death within 30 days after surgery and in-hospital death was defined as death before discharge after surgery. Postoperative systemic complication was defined as complication observed in the heart (ischemic heart attack, acute heart failure), lung (pneumonia, atelectasis), kidney (acute renal failure excluding postoperative renal failure in patients with chronic renal failure), gastrointestinal tract (bleeding), brain (stroke), and other areas (sepsis/DIC and other organs) that occurred during hospitalization. Operative wound complication included infection, dehiscence, necrosis, or lymphocele of the operative wound, graft infection, anastomotic aneurysm, and hemorrhaging or hematoma formation requiring interventions or blood transfusion after the first operation. Operative wound complication after redo or revision surgery was not included. MALE was defined as major amputation or major reintervention including a new bypass graft, jump/interposition graft revision, or thrombectomy/thrombolysis related to the target lesion.
Statistical AnalysisThe categorical and continuous variables were compared using Fisher’s exact test and unpaired Student’s t-test, respectively. The survival curves were calculated with the Kaplan-Meier method and evaluated on log-rank test. To determine the hazard ratio for each variable, univariate and multivariate analysis were performed using a Cox proportional hazard regression analysis. P<0.05 was considered to be statistically significant. Statistical analysis was performed using JMP version 11.2 for Mac (SAS Institute, Cary, NC, USA).
The data of 226 patients (134 patients with pre-BI ≥60 and 92 patients with pre-BI <60) were analyzed. The average age and the prevalence of ≥80 years of age, BMI <18.0 kg/m2, non-ambulatory status, tissue loss, HT, COPD, Hb <10 g/dL, and Alb <3.5 g/dL were significantly higher in the pre-BI <60 group than in the pre-BI ≥60 group, while the prevalence of smoking history was higher in the pre-BI ≥60 group than in the pre-BI <60 group (Table 1).
| Variables | Pre-BI ≥60 (n=134) |
Pre-BI <60 (n=92) |
P-value |
|---|---|---|---|
| Age (years) | 73.1±9.2 (44–93) | 76.1±8.7 (56–93) | 0.013 |
| ≥70 | 93/134 (69) | 72/92 (78) | 0.170 |
| ≥80 | 31/134 (23) | 42/92 (46) | <0.001 |
| Female | 43/134 (32) | 38/92 (41) | 0.161 |
| BMI <18.0 kg/m2 | 10/134 (7) | 20/92 (22) | 0.003 |
| Non-ambulatory | 23/134 (17) | 49/92 (53) | <0.001 |
| Tissue loss | 106/134 (79) | 86/92 (93) | 0.004 |
| DM | 87/134 (65) | 58/92 (63) | 0.780 |
| HT | 103/134 (77) | 81/92 (88) | 0.038 |
| CAD | 56/134 (42) | 38/92 (41) | 1.000 |
| COPD | 14/130 (11) | 25/90 (28) | 0.002 |
| ESRD | 48/134 (36) | 43/92 (48) | 0.129 |
| Stroke | 53/134 (40) | 48/92 (52) | 0.077 |
| Cancer | 17/134 (13) | 14/91 (15) | 0.562 |
| Dyslipidemia | 68/132 (52) | 46/92 (50) | 0.892 |
| Smoking history | 88/133 (66) | 48/92 (52) | 0.038 |
| Hb <10 g/dL | 230/134 (22) | 33/91 (36) | 0.034 |
| Alb <3.5 g/dL | 50/134 (37) | 52/91 (56) | 0.004 |
Data given as mean±SD (range) or n (%). Alb, serum albumin; BMI, body mass index; CAD, coronary artery disease; COPD, chronic obstructive pulmonary disease; DM, diabetic mellitus; ESRD, end-stage renal disease; Hb, hemoglobin; HT, hypertension; pre-BI, Barthel index at admission.
The prevalence of in-hospital death, postoperative systemic complications, post-BI <60 or in-hospital death, and post-BI <60 excluding 11 patients who died during hospitalization, was significantly higher in the pre-BI <60 group than in the pre-BI ≥60 group (Table 2). The causes of death in the pre-BI ≥60 group were cerebral infarction in 1 (within 30 days) and chronic renal failure in 1 (after 31 days), and in the pre-BI <60 group they were cerebral infarction, mesenteric infarction, and hemorrhagic shock in 1 each (within 30 days) and acute heart failure in 3, and sepsis, chronic renal failure, and gastrointestinal bleeding in 1 each (after 31 days). Postoperative systemic complications in the pre-BI ≥60 group were observed in the heart in 5, lung in 3, and as sepsis/DIC in 3 patients, while in the pre-BI <60 group they were observed in the heart in 8, as sepsis/DIC in 6, in the brain in 4, lung in 3, kidney in 2, and gastrointestinal tract in 2. Operative wound complications in the pre-BI ≥60 group were wound infections in 7, lymphocele in 6, dehiscence in 5, hemorrhaging/hematoma in 4, and graft infection in 2, and in the pre-BI <60 group they were wound infections in 9, dehiscence in 4, lymphocele in 3, hemorrhaging/hematoma in 3, and graft infection in 2.
| Pre-BI ≥60 (n=134) |
Pre-BI <60 (n=92) |
P-value | |
|---|---|---|---|
| Operative death | 1 (1) | 3 (3) | 0.307 |
| In-hospital death | 2 (2) | 9 (10) | 0.008 |
| Systemic complications | 9 (7) | 18 (20) | 0.006 |
| Wound complications | 21 (16) | 16 (17) | 0.855 |
| MALE within 30 days | 7 (5) | 12 (13) | 0.050 |
| Post-BI <60 or in-hospital death | 20 (15) | 64 (70) | <0.001 |
| Post-BI <60† | 18/132 (14) | 55/83 (66) | <0.001 |
Data given as n (%). †Excluding deaths during hospitalization. BI, Barthel index; MALE, major adverse limb event; pre-BI, BI at admission; post-BI, BI at discharge.
The prevalence of post-BI<60 or in-hospital death, which implies failure of functional independence, was 70% in the pre-BI <60 group. In other words, 30% of patients with pre-BI <60 recovered to functional independence after surgical revascularization. Conversely, 15% in the pre-BI ≥60 group lost functional independence or died during hospitalization after surgery.
Unadjusted Survival Curve of 3-Year OS and AFSPre-BI ≥60 vs. Pre-BI <60 The OS for pre-BI ≥60 vs. pre-BI <60 was 91±3% vs. 82±4% at 1 year, 82±3% vs. 65±5% at 2 years, and 75±4% vs. 56±6% at 3 years, respectively. Pre-BI ≥60 was superior to pre-BI <60 in OS (P=0.003; Figure 2A). The AFS for pre-BI ≥60 vs. pre-BI <60 was 87±3% vs. 74±5% at 1 year, 77±4% vs. 58±5% at 2 years, and 70±4% vs. 50±6% at 3 years, respectively. Pre-BI ≥60 was superior to pre-BI <60 in AFS (P=0.002; Figure 2B).
(A) Overall survival and (B) amputation-free survival in 226 patients with a record of Barthel index measured at admission (pre-BI).
Post-BI ≥60 vs. Post-BI <60 The OS for post-BI ≥60 vs. post-BI <60 was 96±2% vs. 84±4% at 1 year, 89±3% vs. 60±6% at 2 years, and 79±4% vs. 54±7% at 3 years, respectively. Post-BI ≥60 was superior to post-BI <60 in OS (P<0.001; Figure 3A). The AFS for post-BI ≥60 vs. post-BI <60 was 92±2% vs. 73±5% at 1 year, 84±3% vs. 51±6% at 2 years, and 75±4% vs. 46±6% at 3 years, respectively. Post-BI ≥60 was superior to post-BI <60 in AFS (P<0.001; Figure 3B).
(A) Overall survival and (B) amputation-free survival in 215 patients with a record of Barthel index measured at discharge (post-BI).
Age, gender, BMI, tissue loss, CAD, stroke, ESRD, serum Alb, as well as pre-BI and post-BI were selected for multivariate analysis. Factors other than BI and gender were indicated to be significant in previous reports.2–6 Post-BI <60, ESRD, and hypoalbuminemia were identified as significant predictors of both OS and AFS on multivariate analysis (Table 3).
| Unadjusted | Adjusted | |||||
|---|---|---|---|---|---|---|
| HR | 95% CI | P-value | HR | 95% CI | P-value | |
| OS | ||||||
| Pre-BI <60 | 2.08 | 1.27–3.42 | 0.004 | 0.59 | 0.28–1.22 | 0.159 |
| Post-BI <60 | 3.12 | 1.81–5.41 | <0.001 | 3.17 | 1.59–6.32 | 0.001 |
| Age (years) | ||||||
| ≤69 | 1 | 1 | ||||
| 70–79 | 1.93 | 1.01–3.78 | 0.046 | 2.03 | 0.91–4.68 | 0.084 |
| ≥80 | 2.84 | 1.52–5.52 | 0.001 | 2.37 | 0.99–5.98 | 0.054 |
| Female | 1.29 | 0.78–2.11 | 0.321 | 1.12 | 0.61–2.02 | 0.720 |
| BMI <18.0 kg/m2 | 1.63 | 0.83–2.95 | 0.148 | 0.81 | 0.36–1.65 | 0.580 |
| Tissue loss | 3.08 | 1.27–10.2 | 0.010 | 1.53 | 0.60–5.18 | 0.406 |
| CAD | 2.18 | 1.33–3.61 | 0.002 | 1.18 | 0.66–2.12 | 0.583 |
| Stroke | 1.16 | 0.71–1.90 | 0.544 | 0.70 | 0.40–1.23 | 0.214 |
| ESRD | 5.62 | 3.29–10.1 | <0.001 | 5.43 | 2.95–10.4 | <0.001 |
| Alb <3.5 g/dL | 3.77 | 2.23–6.61 | <0.001 | 3.10 | 1.60–6.27 | <0.001 |
| AFS | ||||||
| Pre-BI <60 | 2.05 | 1.30–3.24 | 0.002 | 0.71 | 0.37–1.36 | 0.305 |
| Post-BI <60 | 3.10 | 1.90–5.10 | <0.001 | 3.31 | 1.76–6.26 | <0.001 |
| Age (years) | ||||||
| ≤69 | 1 | 1 | ||||
| 70–79 | 1.37 | 0.77–2.45 | 0.279 | 1.15 | 0.59–2.27 | 0.673 |
| ≥80 | 1.98 | 1.13–3.51 | 0.016 | 1.20 | 0.59–2.50 | 0.618 |
| Female | 1.46 | 0.91–2.29 | 0.112 | 1.56 | 0.91–2.64 | 0.105 |
| BMI <18.0 kg/m2 | 1.15 | 0.57–2.08 | 0.679 | 0.54 | 0.23–1.11 | 0.099 |
| Tissue loss | 3.72 | 1.54–12.2 | 0.002 | 1.96 | 0.78–6.56 | 0.166 |
| CAD | 1.75 | 1.11–2.76 | 0.016 | 1.14 | 0.68–1.92 | 0.616 |
| Stroke | 1.45 | 0.92–2.29 | 0.110 | 1.00 | 0.61–1.66 | 0.994 |
| ESRD | 3.44 | 2.16–5.58 | <0.001 | 3.09 | 1.85–5.24 | <0.001 |
| Alb <3.5 g/dL | 3.15 | 1.96–5.19 | <0.001 | 2.53 | 1.44–4.58 | 0.001 |
AFS, amputation-free survival; OS, overall survival. Other abbreviations as in Tables 1,2.
This study on the association of pre-BI with early and long-term outcomes of CLI found that the prevalences of in-hospital death and postoperative systemic complications were significantly higher and the 3-year OS and AFS significantly lower in patients with pre-BI <60 than in those with pre-BI ≥60. Adverse events during hospitalization lead to a long hospitalization and increase the cost of treatment, and poor long-term outcomes reduce the merits of surgical revascularization, which is invasive but durable. Accordingly, these findings suggest that surgical bypass may not be suitable as a first strategy for patients with poor ADL.
Given that the biased background of the pre-BI <60 group, which included higher percentages of patients ≥80 years of age or with lower BMI, non-ambulatory status, tissue loss, HT, COPD, anemia, and hypoalbuminemia is compatible with the conditions of frailty with regard to impaired mobility and physical activity, malnutrition, and multiple comorbidities,12,14,16 the present results showing poor outcomes in patients with poor ADL seem predictable. Multivariable analysis, however, failed to indicate that pre-BI <60 was a significant predictor of long-term outcome. Hypoalbuminemia and ESRD were found to be significant preoperative predictors for 3-year OS and AFS. Both of these factors have been frequently demonstrated to be significant predictors so far,2–6 and should be included when determining treatment strategy. The disappearance of pre-BI from the list of significant predictors of long-term outcome on multivariate analysis, however, suggests that poor ADL at admission is not always an absolute determinant of revascularization strategy in patients with CLI. This is also suggested by the fact that 30% of patients with pre-BI <60 recovered to functional independence by treatment with surgical revascularization.
Additionally, we investigated the association of post-BI with 3-year outcomes and found that OS and AFS in patients with post-BI ≥60 were significantly better than those in patients with post-BI<60. Post-BI was found to be a significant predictor on multivariate analysis, along with ESRD and hypoalbuminemia. The difference between pre-BI and post-BI is reflected by severe ischemic symptoms. Taking a strong analgesic might affect ADL. In addition, measuring the correct BI at admission might be difficult in older or cognitive patients with CLI.13 In contrast, post-BI can be measured after successful revascularization followed by adequate wound management and rehabilitation. With these ideal treatments, restrictions on daily life can be removed. Accordingly, post-BI following successful treatment can reflect the true ADL, which is close to that before the onset of CLI.
The present results are similar to the findings in CLI patients with or without preceding intermittent claudication (IC). Patients with CLI can be divided into 2 groups based on the presence of IC before onset of CLI. CLI without IC accounts for 37–63% of CLI patients.17 Some cases of CLI without IC are classified as chronic subclinical limb ischemia (CSLI). Takahara et al showed that CSLI includes more patients with non-ambulation, DM, severe tissue loss, or ESRD,18 and Shirasu et al also identified non-ambulation, DM, severe tissue loss, or hypoalbuminemia as features of CLI without preceding IC.4 The ischemic stage in such patients was already relatively severe, and the AFS after revascularization was worse than in patients with IC.5 Patients with “asymptomatic” PAD had worse physical function and less physical activity and tended to live a sedentary lifestyle linked to higher incidences of mortality and morbidity than those with IC.19 This suggests that patients with a high pre-BI or low pre-BI in association with high post-BI have CLI with IC, and patients with low pre-BI and low post-BI even after successful revascularization have CLI without IC.
The present subjects were limited to those who underwent open surgery. Open bypass is invasive, and activity is reduced immediately after surgery. After the resolution of ischemic symptoms and relief of pain from surgical wounds, BI recovers and increases with continued rehabilitation throughout hospitalization. Given that open bypass is superior to EVT in durability,20–22 ADL will continue to improve after discharge as long as surgical bypass holds. That is an absolute merit of open bypass compared with EVT. We must determine which factors most affect the recovery of ADL after surgical bypass and how long high ADL can be maintained after surgical bypass, given that the recovery of ADL to the level before onset of CLI (i.e., improvement in the quality of life) is a major goal of treatment for CLI. Subsequent follow-up of patients and further analyses are required.
Several limitations associated with the present study warrant mention. First, this was a retrospective study conducted by a single vascular surgery team continuously at 2 hospitals. Second, the patient group was small. Third, the period for enrollment was too long and the treatment strategy changed with time. Fourth, the rules of discharge were not completely fixed. The BI at discharge might have been affected by the timing of discharge and where patients were staying after discharge. Fifth, postoperative medication that might have affected long-term outcome was not included in the variables.
The BI measured at admission (pre-BI) was an indicator of early adverse events in patients who underwent infrainguinal bypass for CLI. It was not, however, the BI measured at admission, but the BI measured at discharge (post-BI) that determined long-term outcome. Pre-BI should not be used to determine treatment strategy.
The authors declare no conflicts of interest.
