Testing the study protocol and interrater reliability of a new end-of-life wound assessment tool: a feasibility study

Some dying individuals can develop skin injuries at the end-of-life (EOL) due to factors associated with the disease processes, aging or both. These EOL wounds, which include Kennedy terminal ulcers, Trombley-Brennan terminal tissue injuries, Skin Changes at Life's End and end-stage skin failure, have distinguishing features. Yet, they can appear similar to pressure injuries (PIs), making assessment difficult. Compounding this was the lack of clinical assessment tool for EOL wounds. In 2022, we conducted a modified Delphi panel to develop a new EOL wound assessment tool for use in dying adults and established the face and content validity of the items. The new tool does not differentiate between a PI and EOL wound; rather, it aids clinicians' assessment of EOL wound characteristics and suggests the development of a multidisciplinary management plan. The next step in the tool development is to determine its reliability. The aim of this study was to test the study protocol and interrater reliability of a new EOL wound assessment tool.

This feasibility study was conducted in dying hospitalised adult patients admitted to medical and palliative care units at three hospitals across southeast Queensland, Australia. We gathered quantitative data according to the study protocol including participant screening, recruitment, consent, data collection and interrater reliability. Our four research assistants (RAs) and an independent blinded outcome assessor were trained in the study protocol and use of the new EOL wound assessment tool. Using a pragmatic approach, patients with a new reported PI were screened for study eligibility. For recruited participants, clinical data, skin blanching, and a deidentified wound photograph were first collected. Next, the RAs used the new tool to assess the patient and the skin to determine the presence of an EOL wound (Yes/No). An off-site independent blinded outcome assessor accessed the participant research data and, using the new tool, undertook the same assessment as the RA. Frequencies and percentages were computed for the feasibility outcomes. Cohen's kappa statistic was calculated to determine the interrater reliability agreement.

Over 20 months, 140 patients were screened, with 23 (16.4%) eligible for recruitment, exceeding our ≥ 10% target. Ten (43.5%) participants were recruited, which fell short of our ≥ 50% target, with study refusal and imminent death the reasons for non-recruitment. Among the 10 recruited study participants, 13 wounds were observed on the sacrum, coccyx, and lower extremities. The interrater reliability between the two assessors was moderate (n = 8/13; 61.5%), with disagreement on five wounds, all located on the heels and toes.

Assessing for EOL wounds in dying patients is a clinical imperative. With minor study protocol adjustments, such as having two clinicians concurrently undertake independent wound assessment and only recruiting from palliative care units, conducting a larger multisite study testing the inter- and intrarater reliability of the new EOL wound assessment tool is feasible.