2023 HSR&D/QUERI National Conference


Lead/Presenter: Sundar Natarajan,  VA New York Harbor Healthcare System, New York, NY
All Authors: Zahrieh I (New York University and VA New York Harbor Healthcare System), Nicholson, A (New York University Grossman School of Medicine and VA New York Harbor Healthcare System) Lipsitz, S (Brigham and Women's Hospital and Harvard Medical School) Natarajan, S (VA New York Harbor Healthcare System and New York University Grossman School of Medicine)

Diabetes afflicts approximately a quarter of US veterans and foot complications are a distressing consequence. These veterans require specialized foot care for neuropathy, calluses, ulcers, infections and amputations. Since these patients are quite fragile, and there is debate regarding risk for death in these veterans and the treatment intensity they require, we formally compared these individuals with established high-risk conditions

We analyzed national veteran data from the National health Interview Survey. Variables concerning diabetes and podiatrist visits were used to create a diabetes and foot disease (DFD) variable. The relationship between DFD and mortality was compared to other serious conditions, specifically previous cancer or prevalent cardiovascular disease (CVD). Data on mortality, cause of death, and follow-up time were obtained from the National Death Index-linked mortality files, with maximum 15 years of follow-up. After initially computing mortality rates and calculating rate ratios, we evaluated the independent effect of DFD, CVD and cancer using multivariate Cox proportional hazards models that adjusted for age, sex, race, marital status, education, US geography, physical activity, dietary habits, smoking, alcohol consumption, and obesity. Finally, to compare DFD to CVD and cancer, we created an 8-level variable with DFD, CVD and cancer combinations, and used DFD alone as referent in Cox models using the same covariates. Analyses were with SAS 9.4 and incorporated the complex sampling to yield population estimates.

There were 39,310 veterans (2001-2014). At baseline, there were 1,649 with DFD, 5,231 with CVD and 6,470 with cancer. Over the 15-year follow-up, participant deaths were: DFD 438; CVD 1,222 and cancer 1,485. Of the DFD deaths, 104 (23.7%) was due to CVD and 94 (21.5%) was due to cancer. The overall mortality rate was 54/1000 person-years for DFD, 44/1000 person-years for CVD and 43/1000 person-years for cancer. The rate ratio for all-cause mortality (with DFD as referent) was 0.817 (95%CI [confidence intervals] 0.732, 0.913) for CVD and 0.809 (95% CI 0.727, 0.902). The multivariate hazard ratio (HR) for all-cause mortality was 1.48 (95% CI 1.31, 1.66) for DFD, 1.36 (95% CI 1.26, 1.48) for CVD and 1.17 (95% CI 1.08, 1.25) for cancer. There were no significant interactions between DFD and CVD or cancer. When CVD and cancer were directly compared to DFD, the HR was 0.99 (95% CI 0.82, 1.21) for CVD and 0.82 (95% CI 0.68, 0.99) for cancer.

Veterans with diabetes who were being seen by podiatry were at equivalent risk for mortality compared to prevalent CVD at baseline and at higher risk for mortality than those with previous cancer at baseline.

This population-based analysis emphases the magnitude of the mortality-risk in veterans with diabetes and foot disease, and quantifies the dramatically increased risk by comparing DFD to CVD and cancer. Even though the VA has developed special clinics for veterans with diabetes and there is easy access to podiatry, more intensive and personalized approaches to prevent mortality are needed. Current recommendations for care and the structure of care may need to be further refined to match intensity of treatment to mortality risk.