Clinical question: What are the risk factors associated with potentially avoidable readmissions (PARs) for end-of-life care issues?
Background: The 6% of Medicare beneficiaries who die each year account for 30% of yearly Medicare expenditures on medical treatments, with repeated hospitalizations a frequent occurrence at the end of life. There are many opportunities to improve the care of patients at the end of life.
Study design: Nested case-control.
Setting: Academic, tertiary-care medical center.
Synopsis: There were 10,275 eligible admissions to Brigham and Women’s Hospital in Boston from July 1, 2009 to June 30, 2010, with a length of stay less than one day. There were 2,301 readmissions within 30 days of the index hospitalization, of which 826 were considered potentially avoidable. From a random sample of 594 of these patients, 80 patients had PAR related to end-of-life care issues. There were 7,974 patients who were not admitted within 30 days of index admission (controls). The primary study outcome was any 30-day PAR due to end-of-life care issues. A readmission was considered a PAR if it related to previously known conditions from the index hospitalization or was due to a complication of treatment.
The four factors that were significantly associated with 30-day PAR for end-of-life care issues were: neoplasm (OR 5.6, 95% CI: 2.85-11.0), opiate medication at discharge (OR 2.29, 95% CI: 1.29-4.07), Elixhauser comorbidity index, per five-unit increase (OR 1.16, 95% CI: 1.10-1.22), and number of admissions in previous 12 months (OR 1.10, 95% CI: 1.02-1.20). The model that included all four variables had excellent discrimination power, with a C-statistic of 0.85.
Bottom line: The factors from this prediction model can be used, formally or informally, to identify those patients at higher risk for readmission for end-of-life care issues and prioritize resources to help minimize this risk.
Citation: Donzé J, Lipsitz S, Schnipper JL. Risk factors for potentially avoidable readmissions due to end-of-life care issues. J Hosp Med. 2014;9(5):310-314.