Program Schedule

Predictors of 30-day hospital readmission in patients with community-onset pneumonia

Session: Poster Abstract Session: Clinical Respiratory Infections
Friday, October 10, 2014
Room: The Pennsylvania Convention Center: IDExpo Hall BC
Background: Predictors of hospital readmission after community-onset pneumonia are not well characterized. Aim: compare characteristics of readmitted vs non-readmitted patients to identify predictors of 30-day readmission in a cohort of patients with an index hospitalization for community-onset pneumonia. 

Methods: Retrospective evaluation of adults admitted to a 624-bed U.S. academic medical center with an ICD-9 diagnosis code of pneumonia from 1/2010 - 12/2011. All patients had a new or progressive infiltrate on chest imaging and at least two of: hypo/hyperthermia, WBC <4k or >12k, or respiratory symptoms (e.g. cough, shortness of breath). Chi-squared, Student’s t-test, and Mann-Whitney U were used for univariate analysis. Variables with a p<0.1 in univariate analysis were candidates for the stepwise multivariate logistic regression to identify factors independently associated with readmission.

Results: 508 patients included with 19.3% (98/508) experiencing 30-day readmission. Baseline characteristics at index hospitalization for the readmitted/non-readmitted groups were the following: median age: 66.5y/65y (p=0.77); CURB-65 score >2: 33.7%/12.2% (p<0.001); ICU admission: 7.1%/6.6% (p=0.85); WBC: 12.9K/12.3K (p=0.48); immunocompromised status: 19.4%/9% (p=0.003); MRSA colonization: 12.2%/3.7% (p=0.001); HCAP: 63.3%/44.9% (p=0.001); aspiration risk: 26.5%/18.8% (p=0.086) COPD: 33.7%/26.3% (p=0.145); heart failure: 21.4%/15.4% (p=0.15); diabetes 30.6%/26.8% (p=0.45); hemodialysis 12.2%/3.7% (p=0.001); antibiotic use in the last 90 days: 41.8%/28.9% (p=0.013); median number of hospitalized days in the last 180 days: 4/0 (p<0.001). In the final multivariate model, primary viral pneumonia was protective for 30d hospital readmission OR=0.115 [95%CI 0.01-0.96] and four variables were independently associated with readmission: MRSA colonization, OR=3.13 [95%CI 1.33-7.38]; hemodialysis, OR=2.56 [95%CI 1.07-6.13]; CURB-65 score >2 OR=3.58 [95%CI 2.05-6.24]; immunocompromised status, OR=2.71 [95%CI 1.39-5.32].

Conclusion: Independent predictors of 30-day readmission in multivariate analysis are typically not modifiable; preventive interventions should be focused on these four patient populations to decrease hospital readmission.

Alan E. Gross, PharmD, BCPS, Univ. of Illinois Hospital and Health Sciences System, Chicago, IL; Pharmacy Practice, Univ. of Illinois at Chicago, Chicago, IL, Trevor Van Schooneveld, MD, University of Nebraska Medical Center, Omaha, NE, Keith M. Olsen, PharmD, FCCP, FCCM, College of Pharmacy, Univ. of Nebraska Med. Ctr., Omaha, NE, Mark E. Rupp, MD, Division of Infectious Diseases, University of Nebraska Medical Center, Omaha, NE and Andre C. Kalil, MD, MPH, Internal Med., Univ. of Nebraska Med. Ctr., Omaha, NE


A. E. Gross, None

T. Van Schooneveld, None

K. M. Olsen, None

M. E. Rupp, 3M: Consultant and Grant Investigator, Consulting fee and Research grant

A. C. Kalil, None

Findings in the abstracts are embargoed until 12:01 a.m. EDT, Oct. 8th with the exception of research findings presented at the IDWeek press conferences.

Sponsoring Societies:

© 2014, All Rights Reserved.

Follow IDWeek