Prognostic ability of diabetic foot classification systems for lower limb amputation in Latin America and the Caribbean: a systematic review with meta-analysis

Objective: To estimate the prognostic accuracy for lower limb amputation using diabetic foot scale studies conducted in Latin America and the Caribbean. Materials and methods: Longitudinal studies were included without language restriction, sourced from Scielo, Embase, PubMed, Web of Science (WOS), and Scopus until October 2024. The studies recorded the incidence of major or any-level amputation according to the application of a diabetic foot scale at the onset of medical care. A meta-analysis was conducted to estimate sensitivity, specificity, positive likelihood ratio (LR+), negative likelihood ratio (LR-), diagnostic odds ratio, and the area under the ROC curve. The risk of bias was assessed using the QUADAS-2 tool, and the certainty of evidence was analyzed according to GRADE. Results: Fifteen articles were included, identifying five scales: IDSA, San Elián, WIfI, Meggitt-Wagner (with thresholds 3 and 4), and University of Texas from nine countries. For any-level amputation, the classification that showed the highest sensitivity for both outcomes were the Wagner scale at threshold grade 3, with a value of 0.91. The San Elián scale exhibited the highest specificity, with a value of 0.92. The most significant LR+ was associated with San Elián, with a value of 4,1, while the most impactful LR- was that of Wagner in category 3, with a value of 0.20. The 15 GRADE evaluations revealed a very low certainty of evidence. Conclusion: The Wagner scale at threshold grade 3 and the San Elián scale demonstrated the highest sensitivity and specificity, respectively. However, our study has significant limitations, such as deficits in follow-up reporting, high heterogeneity, and a limited number of studies, explaining the very low certainty of evidence. There is a need for studies with greater methodological rigor in our region.