Nonblanchable erythema for predicting pressure ulcer development: a systematic review with an individual participant data meta-analysis

Shi, C, Bonnett, LJ ORCID: 0000-0002-6981-9212, Dumville, JC and Cullum, N (2020) Nonblanchable erythema for predicting pressure ulcer development: a systematic review with an individual participant data meta-analysis. BRITISH JOURNAL OF DERMATOLOGY, 182 (2). pp. 278-286.

Text Manuscript_BJD.docx - Author Accepted Manuscript Download (89kB)

Abstract

<h4>Background</h4>Empirical evidence is uncertain regarding the value of nonblanchable erythema in predicting the incidence of stage 2 (or more severe) pressure ulcers.<h4>Objectives</h4>To investigate whether nonblanchable erythema is an independent prognostic factor for pressure ulcer incidence using individual patient data.<h4>Methods</h4>We performed an electronic database search in February 2017 to identify longitudinal studies that considered nonblanchable erythema for predicting pressure ulcer risk in any population. We collected individual participant data for the included studies, and assessed the risk of bias of these studies using the Quality In Prognosis Studies tool. We analysed individual participant data in Stata using mixed-effects logistic regression to investigate the association of interest. The certainty of evidence from individual participant data analysis was assessed using the Grades of Recommendation Assessment, Development and Evaluation. The study was registered with PROSPERO (CRD42017081151).<h4>Results</h4>From the 13 included studies (total 68 077 participants) we had access to individual participant data from four (n = 3223), and 11·9% of participants (383 of 3223) developed new pressure ulcers of stage 2 or above within 28 days. Mixed-effects logistic regression showed that participants with nonblanchable erythema had higher odds of developing new pressure ulcers of stage 2 or above within 28 days of follow-up than those without nonblanchable erythema (multivariable association: n = 2684; odds ratio 2·72, 95% confidence interval 2·02-3·69; τ² = 0; moderate-certainty evidence).<h4>Conclusions</h4>This first prognostic factor review with individual-level data analysis in patients with pressure ulcers suggests that people with nonblanchable erythema are more likely to develop new pressure ulcers of stage 2 or above within 28 days than people without nonblanchable erythema. It is important to identify nonblanchable erythema in practice and to intervene appropriately to prevent pressure ulceration. What's already known about this topic? Pressure ulcer reduction is a high priority for healthcare systems. Regularly inspecting skin to identify skin abnormalities is one key practice for preventing ulceration. Nonblanchable erythema - discoloration of the skin that does not turn white when pressed - is one clinically important skin abnormality. Empirical evidence synthesized using conventional meta-analysis is uncertain regarding the value of nonblanchable erythema for predicting open pressure ulcer incidence; this is partly because the conventional technique has weakness in terms of pooling prognostic effects of different multivariable analyses across studies. What does this study add? This prognostic factor review used individual-level data analysis to overcome the limitations of the conventional meta-analysis technique. For the first time there is confirmatory and moderate-certainty evidence on the association of nonblanchable erythema with pressure ulcer incidence. People with nonblanchable erythema are more likely to develop new pressure ulcers of stage 2 or more severe within 28 days than people without nonblanchable erythema, regardless of their age, baseline pressure ulcer risk or received support surfaces.

Item Type:	Article
Uncontrolled Keywords:	Skin, Humans, Erythema, Incidence, Pressure Ulcer, Data Analysis
Depositing User:	Symplectic Admin
Date Deposited:	12 Jun 2019 10:42
Last Modified:	19 Jan 2023 00:40
DOI:	10.1111/bjd.18154
Related URLs:	Author Publisher
URI:	https://livrepository.liverpool.ac.uk/id/eprint/3045540