The accuracy and cost-effectiveness of strategies used to identify peripheral artery disease among patients with diabetic foot ulcers

Background: Patients with diabetic foot ulcers (DFUs) should be evaluated for peripheral artery disease (PAD). We sought to estimate the overall diagnostic accuracy for various strategies that are used to identify PAD in this population. Methods: A Markov model with probabilistic and deterministic sensitivity analyses was used to simulate the clinical events in a population of 10,000 patients with diabetes. One of 14 different diagnostic strategies was applied to those who developed DFUs. Baseline data on diagnostic accuracy of individual noninvasive tests were based on a meta-analysis of previously reported studies. The overall sensitivity and cost-effectiveness of the 14 strategies were then compared. Results: The overall sensitivity of various combinations of diagnostic testing strategies ranged from 32.6% to 92.6%. Cost-effective strategies included ankle-brachial indices for all patients; skin perfusion pressures (SPPs) or toebrachial indices (TBIs) for all patients; and SPPs or TBIs to corroborate normal pulse examination findings, a strategy that lowered leg amputation rates by 36%. Strategies that used noninvasive vascular testing to investigate only abnormal pulse examination results had low overall diagnostic sensitivity and were weakly dominated in cost-effectiveness evaluations. Population prevalence of PAD did not alter strategy ordering by diagnostic accuracy or cost-effectiveness. Conclusions: TBIs or SPPs used uniformly or to corroborate a normal pulse examination finding are among the most sensitive and cost-effective strategies to improve the identification of PAD among patients presenting with DFUs. These strategies may significantly reduce leg amputation rates with only modest increases in cost.