Incidence and Predictors of Hospital Readmission Among Patients With Chronic Obstructive Pulmonary Disease in the Department of Veterans Affairs

SESSION TITLE: COPD Poster Discussion SESSION TYPE: Original Investigation Poster Discussion PRESENTED ON: Sunday, October 25, 2015 at 01:30 PM - 03:00 PM PURPOSE: This study describes 30- and 60-day all-cause and 30-day chronic obstructive pulmonary disease (COPD)-related readmission and factors associated with all-cause readmission for adults with COPD in the Department of Veterans Affairs (VA). METHODS: A retrospective cohort analysis was conducted using data from the national VA VINCI database of COPD patients hospitalized for COPD (indicated by primary discharge diagnosis) between 1/1/2004 and 7/1/2014 with ≥ 1 year baseline data. Readmission incidence was calculated as number of index hospitalizations with a subsequent hospitalization within 30 or 60 days after discharge divided by total number of index hospitalizations. Multivariate logistic regression was used to determine predictors of the first readmission including demographics, care site measures, history of hospitalizations, smoking, comorbidities, index hospital length of stay, ICU stay, and discharge disposition. RESULTS: Overall, 89,502 COPD patients had an index hospitalization during the study period. All-cause 30-day readmission was 17.3% and 60-day readmission was 26.7%. COPD-related 30-day readmission was 7.8%. Of the 11,977 COPD patients with readmission within 30 days of first COPD hospitalization, mean age was 70.6 years and 97% were men. Top primary discharge diagnoses for the first readmission were COPD-related (34.3%) and pneumonia (5.9%), with the remaining distributed among multiple conditions (the most common being heart failure, 5.8%). An initial regression model showed that previous hospitalizations (OR: 1.39 for 1 hospitalization to 3.24 for 5 or more hospitalizations, p<0.001), moderate/severe liver disease (OR: 1.44, p=0.003), paraplegia (OR: 1.41, p<0.001), pulmonary hypertension (OR: 1.32, p=0.016), substance abuse (OR: 1.28, p<0.001) and heart failure (OR:1.25, p<0.001) were the strongest significant independent predictors of all-cause 30-day readmission among COPD patients in the VA. CONCLUSIONS: COPD readmission incidence in the VA is consistent with that identified by Medicare and resulted from multiple causes. Initial investigation showed that the risk of 30-day all-cause readmission among COPD patients was associated with multiple chronic difficult-to-manage comorbid conditions. CLINICAL IMPLICATIONS: Since hospital readmissions have been targeted under the Affordable Care Act for cost control, characterizing readmission risk factors will help in understanding the potential for lowering the rate among COPD patients. DISCLOSURE: Scott DuVall: Grant monies (from industry related sources): research investigator Kevin Brown: Grant monies (from industry related sources): research investigator Linda Nici: Consultant fee, speaker bureau, advisory committee, etc.: research consultant Carolyn Rochester: Consultant fee, speaker bureau, advisory committee, etc.: research consultant Patricia Nechodom: Grant monies (from industry related sources): research investigator Kathleen Fox: Other: Contractor for AstraZeneca Danielle Chun: Grant monies (from industry related sources): research investigator Benjamin Viernes: Grant monies (from industry related sources): research investigator Michael DePietro: Employee: Employee of AstraZeneca No Product/Research Disclosure Information