A recent study found that mobility prediction models overestimate the likelihood of mobility in socioeconomically disadvantaged people following a major amputation. The researchers performed a retrospective review of patients with a major amputation secondary to peripheral arterial disease from 2016-2022, excluding patients who were nonambulatory before amputation or with a contralateral amputation.
They investigated the AmpPredict and Vascular Quality Initiative (VQI) model, which predict one-year mobility, and the AMPSIMM, which predicts the degree of mobility with a prosthesis at one year, and compared the predicted mobility rates versus actual mobility rates.
The cohort consisted of 126 patients (71 percent male, 60 percent nonwhite race, with a mean state Area Deprivation Index of 9/10), whose actual mobility at one year was 43 percent. Baseline characteristics were significantly different between the study and derivation cohorts.
- Of the 38 patients with an AmpPredict one-year mobility of at least 70 percent, 45 percent actually achieved mobility.
- Of 101 patients with a “high” predicted probability from the VQI score (at least 71 percent), 48 percent achieved mobility.
- The mean difference between AmpPredict and VQI for a given patient was 36 percent.
- AMPSIMM predicted 87 percent of patients would be community ambulators (versus home ambulators), and 32 percent of patients actually achieved community ambulation.
The researchers suggested that the overestimation may be related to demographics/comorbidities of the cohort being significantly different from the derivation cohorts, and recommend caution when applying prediction models to a population with significantly different characteristics from the population used to derive the model.
The open-access study, “Applying mobility prediction models to real world patients with major amputations,” was published in the Journal of Vascular Surgery.
