Palliative care consultation and hospital length of stay.

It has frequently been claimed that palliative care (PC) consultation services reduce hospital length of stay (LOS). We review 12 published studies comparing patients receiving PC or similar intervention and patients receiving usual care with regard to average total hospital LOS. None of the six observational studies showed LOS impact. Three of the four quasi-experiments and one of the two randomized controlled trials reported LOS reduction for at least one subsample. Reduced LOS was demonstrated only for decedents in intensive care unit-based interventions using experimental or quasi-experimental research designs. PC program leaders are cautioned against promising that their inpatient consultations will reduce the length of those admissions because this may be nearly impossible for a typical hospital-based PC program to demonstrate using observational data. Research to date has been handicapped by designs and methods not suitable for detecting an impact on LOS. Only three studies included survivors and decedents and disaggregated them in analysis and interpretation, despite profound differences in the meaning and implications of reduced LOS for survivors and decedents. Recommendations for future studies include conceptualizing, analyzing, and reporting outcomes separately for survivors and decedents; strengthening study design to reduce the likelihood of failing to detect actual LOS impact; using methods that allow for creation of a reasonable comparison group; and addressing the fundamental problem that LOS is both a predictor and criterion variable in observational studies of palliative care consultation services.