Interobserver reliability of scoring systems for diabetic foot ulcer assessment

Each person with a diabetic foot ulcer should be systematically evaluated, as recommended in Australian and international guidelines. Systematic evaluation can be used to determine who is at risk of a poor outcome, who needs urgent treatment, and what disciplines should be involved. Another reason for such evaluation is to capture, analyse and benchmark the outcomes of your own patients against (inter)national standards.

DFA strongly supports systematic evaluation of people with foot ulcers, and developed together with various stakeholders the “Australian Diabetic Foot Ulcer Minimum Dataset”. Using this dataset, diabetic foot services can collect the minimally relevant information to evaluate their processes and outcomes. In the minimum dataset, diabetic foot ulcer related items are chosen in line with the PEDIS classification.

However, as with any scoring system, clinicians may observe the same ulcers differently. New research from the UK has investigated the reliability of the three most commonly used systems, PEDIS (Perfusion, Extent, Depth, Infection, Sensation), SINBAD (Site, Ischemia, Neuropathy, Bacterial infection, Depth) and UT (University of Texas). They found slight to moderate single observer agreement for all systems, but multiple observer reliability was almost perfect.

Missing from the article, unfortunately, are baseline characteristics of the observations made. Agreement will never be perfect, but the magnitude of disagreement could have been better interpreted with that information. For example, differences in observations on infection may not lead to major clinical changes when this primarily concerns disagreement whether an infection is mild or moderate; however, when disagreements are found between whether or not an ulcer is infected, this may directly impact clinical practice.

The authors conclude that caution is needed when these systems are used in daily clinical practice. Moderate single-observer agreement suggests that referral to other specialties or development of treatment plans for individual patients should not be solely based on these systems. However, most clinicians will always use more information than just the four to six items covered with PEDIS, SINBAD and UT. The good news is the conclusion that all three systems may be reliably used by multiple observers, for example in research or audits. This supports the use of the PEDIS system in the Australian Diabetic Foot Ulcer Minimum Dataset, and we look forward seeing the first results of Australian clinical sites using that.