From: Risk prediction models for diabetic foot ulcer development or amputation: a review of reviews
Author (year); Risk of Bias (ROBIS); Search Dates; Sources; Study type | Population | Outcome | #Studies/ #Models | Limitations | Authors’ Conclusions | Our Conclusions |
---|---|---|---|---|---|---|
Beulens et al. (2021) [9]; Low ROBa; Inception-10/21/2020; PubMed and EMBASE; Systematic review and external validation study; Funding: Dutch Diabetes Research Foundation | Patients with type 2 diabetes | Foot ulcer development, amputation, or neuropathy, or a combination of these over a minimal 1 year follow-up | 21/34 | (i) Low [5-year] incidence of amputation in the external validation cohort (70/7624; 0.9%), (ii) inability to differentiate between major and minor imputations (missing data), (iii) limited generalizability to populations/settings different from validation cohort (iv) inability to validate models including variables not available in validation cohort | The models by Boyko et al., [14] PODUS 2015, [15] and Martins-Mendes et al., [16] performed well to predict outcomes of either amputation or foot ulcer | PODUS 2015 was developed as a risk classification system with no time horizon for risk prediction. Hence, it was excluded from further consideration. The models by Boyko et al. and Martins-Mendes et al. are prognostic models |
Fernandez-Torres et al. (2020) [10]; Moderate ROB; Inception- 12/30/2019; PubMed, Scopus, SciELO, CINAHL, Cochrane, PEDro, and EMBASE; Systematic review; Funding: None | Patients with diabetic foot disease including neuropathy, regardless of the type of diabetes | Neuropathy risk, ulceration risk, and diabetic foot ulcer outcome (amputation risk, healing, infection assessment, and measurement) | 29/39 | Exclusion of tools not published in English, French, Spanish, Portuguese, and Italian, or reporting psychometric characteristics not captured by the review’s inclusion criteria | The Queensland High Risk Foot Form (QHRFF) was valid and reliable for the assessment of ulceration risk | QHRFF was developed as a risk classification system with no time horizon for prediction. Hence, it was excluded from further consideration |
Monteiro-Soares et al. (2011) [11]; Low ROB; Inception until 4/15/2010; MEDLINE; Systematic review; Funding: None | Patients with diabetes, type unspecified | Foot ulcer development | 13/5 | Quality assessment, data analysis, and extraction were performed by one reviewer who was not blinded to authors or institutions | The best method for assessment of risk stratification is not immediately apparent | Identical to the authors conclusions |