Lead/Presenter: Alyson Littman, Seattle Epidemiologic Research and Information Center
All Authors: Littman AJ (Seattle-Denver COIN), Timmons AK (Seattle Epidemiologic Research and Information Center) Korpak A (Seattle Epidemiologic Research and Information Center) Chan KC (Seattle Epidemiologic Research and Information Center) Jones KT (VA Office of Health Equity) Shirley S (VHA Innovation Ecosystem) Nordrum K (VHA Innovation Ecosystem) Masadeh S (Cincinnati Veteran Affairs Medical Center) Moy E (VA Office of Health Equity)
Objectives:
Monitoring foot skin temperatures is one of the few interventions recommended for the prevention of ulcer recurrence in patients with diabetes but has not been evaluated in a large multi-centered study. We aimed to evaluate the effectiveness of remote temperature monitoring (RTM) in the Department of Veterans Affairs healthcare system in terms of lower extremity amputation (LEA), all-cause hospitalization, and death.
Methods:
We conducted a matched retrospective cohort study. Patients were considered eligible if they were at high risk for ulceration and amputation (e.g., primarily [>90%] due to a recent diabetic foot ulcer) and did not have any of the exclusion criteria. Eligible patients enrolled in RTM between 2019 and 2021 were matched (on age, rurality, Veterans Integrated Service Network, history of a foot ulcer, LEA, and diabetes) up to 3:1 to eligible patients who were not enrolled. Follow-up was through August 31, 2021. We used conditional Cox regression (with death as a competing event) to account for matched data and estimate unadjusted and adjusted cause-specific hazard ratios (HR) and corresponding 95% confidence intervals (CI) for LEA (primary outcome) and all-cause hospitalization and death (secondary outcomes). Sensitivity analyses included restriction to the most experienced sites; time varying covariates to account for a hypothesized lag in benefit; exclusion of those with a prior LEA, or chronic kidney disease or end stage renal disease (two separate analyses), and “as-treated†analyses using data on average weekly mat use.
Results:
The study included 924 patients enrolled in RTM and a comparison group of 2757 patients. Less than 2% of patients were under 50 years of age, nearly half were between 70 and 79 years, about 16% Black, and were 76% white. After adjusting for covariates (race, hemoglobin A1c, osteomyelitis, Charcot foot, kidney disease, Gagne comorbidity index, drive time to specialty care, and ER/urgent care visits), RTM was not associated with LEA incidence (4.7 per 100 person-years in both groups, HR = 0.89, 95% CI 0.61-1.31) or all-cause hospitalization (adjusted HR = 0.97, 95% CI 0.83-1.13), but was inversely associated with death (adjusted HR = 0.67, 95% CI 0.52-0.84). Sensitivity analyses that excluded those with LEA at baseline suggested a meaningful reduction in incident LEA risk (HR = 0.72, 95% CI 0.43-1.20), though confidence intervals were wide.
Implications:
This study does not provide support that RTM reduces the risk of LEA or all-cause hospitalization, though RTM may be effective in those without a history of LEA. Randomized controlled trials can overcome important limitations and identify implementation gaps that can be filled to improve effectiveness. Future research may also help define subsets of patients that are more or less likely to benefit from RTM.
Impacts:
Rigorous evaluation of technologies and systems is critical to ensure high quality care. As VHA cares for over 2 million patients with diabetes, reducing the physical and psychological burden of diabetic foot ulcer recurrence on individuals and the high costs on the healthcare system is paramount.