Effects of hospital palliative care on health, length of stay, and in-hospital mortality across intensive and non-intensive-care units: A systematic review and metaanalysis

Abstract

Objective: Hospital palliative care has been shown to improve quality of life and optimize hospital utilization for seriously ill patients who need intensive care. The present review examined whether hospital palliative care in intensive care (ICU) and non-ICU settings will influence hospital length of stay and in-hospital mortality. Method: A systematic search of CINAHL/EBSCO, the Cochrane Library, Google Scholar, MEDLINE/Ovid, PubMed, and the Web of Science through 12 October 2016 identified 16 studies that examined the effects of hospital palliative care and reported on hospital length of stay and in-hospital death. Random-effects pooled odds ratios and mean differences with corresponding 95% confidence intervals were estimated. Heterogeneity was measured by the I-2 test. The Grading of Recommendations Assessment, Development, and Evaluation (GRADE) system was utilized to assess the overall quality of the evidence. Results: Of the reviewed 932 articles found in our search, we reviewed the full text of 76 eligible articles and excluded 60 of those, which resulted in a final total of 16 studies for analysis. Five studies were duplicated with regard to outcomes. A total of 18,330 and 9,452 patients were analyzed for hospital length of stay and in-hospital mortality from 11 and 10 studies, respectively. Hospital palliative care increased mean hospital length of stay by 0.19 days (pooled mean difference = 0.19; 95% confidence interval [CI95%] = -2.22-2.61 days; p = 0.87; I-2 = 95.88%) and reduced in-hospital mortality by 34% (pooled odds ratio = 0.66; CI95% = 0.52-0.84; p < 0.01; I-2 = 48.82%). The overall quality of evidence for both hospital length of stay and in-hospital mortality was rated as very low and low, respectively. Significance of results: Hospital palliative care was associated with a 34% reduction of in-hospital mortality but had no correlation with hospital length of stay.