Rehospitalization after hip fracture: predictors and prognosis from a national veterans study

Abstract

Objectives: To estimate the risk and long-term prognostic significance of 30-day readmission postdischarge of a 4-year cohort of elderly veterans first admitted to Medicare hospitals for treatment of hip fractures (HFx), controlling for comorbidities.

Design: Retrospective, national secondary data analysis.

Setting: National Medicare and Veterans Health Administration (VHA) facilities.

Participants: The study cohort was 41,331 veterans with a HFx first admitted to a Medicare eligible facility during 1999 to 2002.

Measurements: HFxs were linked with all other Medicare and VHA inpatient discharge files to capture dual inpatient use. Logistic regression was used to examine the relationship between 30-day readmission and age, sex, inpatient length of stay, and selected Elixhauser comorbidities.

Results: Approximately 18.3% (7,579/41,331) of HFx patients were readmitted within 30 days. Of those with 30-day readmissions, 48.5% (3,675/7,579) died within 1 year, compared with 24.9% (8,388/33,752) of those without 30-day readmissions. Readmission risk was significantly greater in the presence of specific comorbidities, ranging from 11% greater risk for patients with fluid and electrolyte disorders (95% confidence interval (CI)=1.04-1.20) to 43% for renal failure (95% CI=1.29-1.60). For this cohort, cardiac arrhythmias (24%), chronic pulmonary disease (28%), and congestive heart failure (16%) were common comorbidities, and all affected the risk of 30-day readmission.

Conclusion: Patients with HFx with 30-day readmissions were nearly twice as likely to die within 1 year. Identification of several predictive comorbidities at discharge and examination of reasons for subsequent readmission suggests that readmission was largely due to active comorbid clinical problems. These comorbidity findings have implications for the current Centers for Medicare and Medicaid Services (CMS) pay-for-performance initiatives, especially those related to better coordination of care for patients with chronic illnesses. These comorbidity findings for elderly patients with HFx may also provide data to enable CMS and healthcare providers to more accurately differentiate between comorbidities and hospital-acquired complications under the current CMS initiative related to nonpayment for certain types of medical conditions and hospital acquired infections.