As the prevalence of diabetes and peripheral artery disease rise, chronic lower limb (LL) wounds are an increasingly important public health issue. These wounds can impair function and reduce quality of life. Given the intensity and complexity of outpatient medical management required to heal chronic wounds and prevent subsequent wounds, the issue of access to wound care is of paramount importance, especially in rural areas where barriers to specialty care exist.
The goals of this study were to: (1) to identify and characterize an urban and a rural cohort of VISN20 Veterans with a chronic LL wound in FY2007 with follow-up for one year, (2) to determine whether rural and urban Veterans with chronic LL wounds were equally likely to receive evidence-based good wound care, and (3) to compare wound outcomes for rural and urban Veterans.
Using a set of high probability ICD-9 codes, we identified a cohort of 160 rural and 160 urban Veterans with incident chronic LL wounds in VISN20 first treated at a VA facility between October 1, 2006 and September 30, 2007. We used VA's classification system to define urban and rural residence. We defined a study wound as an open wound on a Veteran's LL that did not heal within 30 days from the first treatment visit. Using VA medical record notes, we assessed whether several components of evidence-based care were provided at each visit and classified Veterans as receiving or not receiving evidence-based care during at least 80% of visits. We also recorded the number of visits and the provider type for all wound care visits within VA. We searched Medicare claims to identify dual wound care use, defined as at least one outpatient or inpatient encounter with one of the ICD-9 codes used to identify the sample, during the study period. We recorded outpatient visit and inpatient stay counts and outpatient provider type as we did within VA. We calculated the mean number of outpatient visits and inpatient (hospital and skilled nursing) stays in VA and Medicare and used Poisson regression models with robust standard errors to compare utilization for rural and urban Veterans and for dual users and VA-exclusive users, adjusting for potential confounders. We used proportional hazards models to estimate the hazard ratio (HR) of wound healing, accounting for the competing risks of amputation and death, and adjusting for confounding by various factors based on the literature.
During one year of follow-up, 234 Veterans' wounds healed (73%; 123 rural and 111 urban), 27 Veterans underwent amputation (8%;16 rural and 11 urban), 20 Veterans died with the wound present (6%; 7 rural and 13 urban), 5 Veterans were lost to follow-up (2%; 2 rural and 3 urban), and 34 had ongoing wounds at the end of follow-up (11%; 12 rural and 22 urban). Evidence-based wound care was similar for rural and urban Veterans within VA, but most Veterans did not receive evidence-based care during at least 80% of visits. For example, only 20% of Veterans with venous wounds had edema assessments documented in at least 80% of visits and only 19% of Veterans with diabetic wounds had debridement performed during at least 80% of visits. Rural Veterans had lower outpatient wound care utilization than urban Veterans (mean 6.8 versus 9.9 visits), including fewer visits with wound care specialists (mean 4.2 versus 6.9 VA wound care specialty visits). Inpatient utilization was low and similar for rural and urban Veterans (0.9 among rural Veterans and 0.8 among urban Veterans). After adjusting for age, age squared, number of comorbid conditions, dual use, SCD rating 50%, and time of follow-up, rural Veterans had significantly lower outpatient utilization compared to urban Veterans (p<0.001) although the model did not fit the data well. In the adjusted analysis, there was no statistically significant difference in the number of inpatient stays for rural and urban Veterans (p=0.96). Rural Veterans had a similar hazard of wound healing (HR=1.11, 95% CI: 0.84-1.47, p=0.45) compared to urban Veterans. Amputation was more common among rural Veterans (HR=2.65, 95% CI: 1.02-6.87, p=0.045) and 50% of rural Veterans versus 9% of urban Veterans had a transtibial or transfemoral amputation. This result was based on a small number of events and is accompanied by a wide confidence interval. Of the cohort, 71% were enrolled in Medicare but only 13% of cohort members were dual system wound care users. Medicare-VA dual users had significantly higher observed utilization (mean 11.6 outpatient visits and 1.7 inpatient stays compared to 7.5 and 0.7, respectively). In a model adjusted for age, age squared, rural residence, the original reason for Medicare eligibility, the number of comorbid conditions, follow-up time, SCD rating 50%, and baseline wound severity, we found dual users' outpatient wound care utilization was not statistically significantly different than VA-exclusive users' (p=0.093), but inpatient utilization did differ significantly after adjustment (p=0.001). Neither utilization model fit the observed data well. The proportion of Veterans who received evidence-based care within VA was similar for dual users and VA-exclusive users. For example, 26% of dual users and 23% of VHA-exclusive users had sharp debridement performed during at least 80% of visits. Dual users had a significantly lower hazard rate of wound healing compared to VA-exclusive users (HR=0.38, 95%CI: 0.25-0.56, p<0.001), while the hazards for amputation (adjusted HR=3.73, 95% CI: 1.10-12.63, p=0.034) and death (adjusted HR=3.89, 95%CI: 1.23-12.28, p=0.020) both were significantly higher for dual users compared to VA-exclusive users.
VA wound care quality was similar for rural and urban Veterans but evidence-based guideline concordant care was not consistently provided; for components assessed, fewer than half of veterans received the care in at least 80% of visits. In spite of lower utilization among rural Veterans, we found no difference in wound healing between rural and urban Veterans. Our findings suggest that Medicare-VA dual use is associated with higher utilization and detrimental to wound outcomes. Additional research is needed to measure wound care quality within Medicare and to determine whether coordination across VA and Medicare might improve wound outcomes among Veterans.
External Links for this Project
Grant Number: I01HX000207-01A1
None at this time.