Interobserver Reliability of Three Validated Scoring Systems in the Assessment of Diabetic Foot Ulcers.

Abstract

Scoring systems for diabetic foot ulcers may be used for clinical purposes, research or audit, to help assess disease severity, plan management, and even predict outcomes. While many have been validated in study populations, little is known about their interobserver reliability. This prospective study aimed to evaluate interobserver reliability of 3 scoring systems for diabetic foot ulceration. After sharp debridement, diabetic foot ulcers were classified by a multidisciplinary pool of trained observers, using the PEDIS (Perfusion, Extent, Depth, Infection, Sensation), SINBAD (Site, Ischemia, Neuropathy, Bacterial infection, Depth), and University of Texas (UT) wound classification systems. Interobserver reliability was assessed using intraclass correlations (0 = no agreement; 1 = complete agreement). Thirty-seven patients (78.4% male) were assessed by a pool of 12 observers. Single observer reliability was slight to moderate for all scoring systems (UT 0.53; SINBAD 0.44; PEDIS 0.23-0.42), but multiple observer reliability was almost perfect (UT 0.94; SINBAD 0.91; PEDIS 0.80-0.90). The worst agreement for single observers was when scoring infection (SINBAD 0.28; PEDIS 0.28), ischemia (SINBAD 0.26; PEDIS 0.23), or both (UT 0.25); however, this improved to almost perfect agreement for multiple observers (infection: 0.83; ischemia: 0.80-0.82; both: 0.81). These classification systems may be reliably used by multiple observers, for example, when conducting research and audit. However, they demonstrate only slight to moderate reliability when used by a single observer on an individual subject and may therefore be less helpful in the clinical setting, when documenting ulcer characteristics or communicating between colleagues.