Risk assessment scales for pressure ulcers in intensive care units : A systematic review with meta - analysis

Aims: To identify risk assessment scales for pressure ulcers that have been used in critical care units. Determine which of them have been validated according to validity, predictive capacity and reliability criteria, developing, where possible, aggregate indicators. Methods: Systematic review of the clinical literature with meta-analysis. The main international health science databases were searched for prospective studies on the validity and/or predictive capacity of pressure ulcers risk assessment scales published between 1962 and 2009 in any language and with a loss to follow-up of less than 25%. We excluded grey literature, reviews, and retrospective and crosssectional studies. The methodological quality of the studies was assessed according to the guidelines of the Critical Appraisal Skills Programme. The indicators analysed were validity, effect size (RR – relative risk) and reliability. When two or more valid original studies were found, a meta-analysis was conducted using the random effect model followed by a sensitivity analysis. Results: A total of 255 articles were found. These articles identified 16 risk assessment scales designed specifically for ICUs. Twenty-six (26) studies measure the validity of these scales. Only three scales have been validated by more than one study (NM Bienstein, Cubbin-Jackson, JacksonCubbin). Four general scales have also been validated in ICUs (Braden, Norton, BM Choi Song-and Waterlow). Waterlow and NM Bienstein are not valid due to their low sensitivity. The CubbinJackson, Jackson-Cubbin and Norton scales present very similar predictive validity data, but have only been tested with very small samples. The Braden scale presents the best test results in ICUs, with adequate parameters and predictive validity. Conclusions: We recommend the use of the Braden Scale for the pressure ulcer risk assessment in critical care units. Other scales, such as the CubbinJackson, Jackson-Cubbin, Norton or BM Choi-Song scales, may be useful but need to be tested in larger numbers of patients.