Prevalence, incidence, risk factors, and outcomes of pressure injuries in patients receiving extracorporeal membrane oxygenation (ECMO): A systematic review and meta-analysis.

INTRODUCTION: Pressure injury is a common, preventable harm in critically ill patients, yet its burden among those receiving extracorporeal membrane oxygenation (ECMO) remains poorly quantified. We aimed to synthesize the prevalence and incidence of PI in ECMO-supported patients, analyzed separately by epidemiological measure, and to summarize associated factors and clinical outcomes. METHODS: We systematically searched five bibliographic databases (PubMed, Embase, Scopus, Web of Science, and CINAHL), supplemented by ProQuest and Google Scholar, for observational studies reporting pressure injury in patients receiving ECMO. Every reported estimate was classified into one of five epidemiological categories: prevalence at ECMO initiation or transfer, point prevalence, period prevalence during the index admission, cumulative incidence of ECMO-acquired PI, or incidence density, according to its baseline PI status, time origin, ascertainment window, numerator, denominator, and stage threshold. Quantitative pooling was restricted to studies reporting the same measure with a comparable structure; all other categories were synthesized narratively. Methodological quality was appraised with the Joanna Briggs Institute checklist for prevalence studies. RESULTS: Nine observational studies (3106 ECMO-supported patients) were included. Four adult cohorts reported the proportion of patients with at least one PI of any stage at any time during the index admission, with prevalent cases included (period prevalence); the pooled estimate across these four studies was 56.0% (95% confidence interval [CI]: 44.4-67.6; I2 = 78.47%). No included study reported point prevalence on a predefined assessment date. A PI was already present at admission or transfer to the ECMO center in 17.0-30.8% of patients. The cumulative incidence of ECMO- or ICU-acquired PI ranged from 4.2% to 44.4% across five studies that differed in stage thresholds, ascertainment windows, and populations, and was therefore not pooled; one study reported an incidence density of 1.7 new PIs per 100 ECMO patient-days. Reported correlates of PI included prone positioning, longer ECMO and ventilation duration, older age, male sex, and renal replacement therapy; device-related injury was predominantly endotracheal-tube-related. CONCLUSION: The burden of PI among ECMO-supported patients appears high in every epidemiological category examined, but the individual measures answer different questions and cannot be reduced to a single summary figure. Estimates should be interpreted within their category and confirmed in standardized prospective studies that report clearly defined measures of prevalence and incidence.