A multilevel analysis of three randomised controlled trials of the Australian Medical Sheepskin in the prevention of sacral pressure ulcers

Abstract

Objective: To assess the effectiveness of the Australian Medical Sheepskin in preventing sacral pressure ulcers (PUs), based on combined data from existing published trials.

Design and setting: Data from two randomised controlled trials (RCTs) among Australian hospital patients and one RCT among Dutch nursing home patients were pooled, comprising a total population of 1281 patients from 45 nursing wards in 11 institutions. These data were analysed in two ways: with conventional meta-analysis based on the published effect sizes; and with multilevel binary logistic regression based on the combined individual patient data. In the multilevel analysis, patient, nursing ward and institution were used as levels and we controlled for sex, age, PU risk and number of days of observation.

Main outcome measure: Incidence of sacral PUs.

Results: Overall, the incidence of sacral PUs was 12.2% in the control group versus 5.4% in the intervention group with an Australian Medical Sheepskin. Conventional meta-analysis showed significantly reduced odds of developing a PU while using the sheepskin (odds ratio [OR], 0.37 [95% CI, 0.17-0.77]). Multilevel analysis gave an OR of 0.35 and narrowed the confidence interval by almost 50% (95% CI, 0.23-0.55).

Conclusions: These analyses of pooled data confirm that the Australian Medical Sheepskin is effective in preventing sacral PUs. Multilevel analysis of individual patient data gives a more precise effect estimate than conventional meta-analysis.