|
|
|
Back to 2014 Annual Meeting Abstracts Predictors and Outcomes of Readmission for Clostridium difficile in a National Sample of Medicare Beneficiaries Courtney Collins*1, M. Didem Ayturk1,2, Frederick a. Anderson1,2, Heena P. Santry1 1Surgery, University of Massachusetts, Worcester, MA; 2Center for Outcomes Research, Worcester, MA
Introduction: The incidence and severity of Clostridium difficile (CD) infections is rapidly increasing in the US, particularly in the elderly. Older patients may also be at increased risk of recurrence of CD, placing them at further risk for adverse outcomes. We examined readmission rates and patient characteristics in a national sample of Medicare beneficiaries with CD and determined predictors of CD readmission. Methods: We queried a 5% random sample of Medicare claims data(2009-2011) and identified hospitalized patients with CD by primary or secondary ICD-9 diagnosis codes. Patients were considered readmitted for CD after these index hospitalizations if they had a subsequent admission with CD as the primary diagnosis. Demographics (race, age, sex), clinical information (Elixhauser index, gastrointestinal comorbidities), and original CD hospitalization characteristics (length of stay (LOS), need for ICU care, CD as primary vs. secondary diagnosis) were compared for patients who were and were not readmitted for CD. A multivariable analysis determined independent predictors of CD readmission. Results:Of 8998 beneficiaries surviving a hospitalization with CD, 1267(14%) were readmitted for CD during the study period. Median time to CD readmission was 23 days(IQR 9-47 days). Readmitted patients were more likely to be female, 75-84 years old, have more comorbidities, and have a history of inflammatory bowel disease(IBD) than patients not readmitted (Table 1). Readmitted patients were significantly more likely to have had CD as the primary diagnosis, shorter LOS, and lower rates of ICU care during index admission. Of readmitted patients, 382(30%)received outpatient CD treatment (vancomycin and/or metronidazole), 176(14%) were exposed to other oral antibiotics, and 331(26%) had ≥1 admission for reasons other than a primary diagnosis of CD prior to readmission for CD. Readmitted patients stayed a median of 7 days(IQR 4-17) with 159(12.5%) requiring ICU care; 85(6.7%) died during their readmission. Colectomy rate was low at 0.9%(12 patients). On multivariable analysis female gender, Elixhauser index, age 75-84 years, IBD, original LOS <7 days, and ICD-9 as primary diagnosis were independent predictors of CD readmission(Table 2). About one fourth of readmitted patients (22%) were readmitted more than once. Conclusion: Nearly 1 in 6 patients hospitalized for CD experience recurrence requiring readmission, often in the first month after their index hospitalization. It appears that older patients with more comorbidities, in particular IBD, admitted primarily for CD who had shorter treatment courses during index admission are at highest risk of readmission. These findings suggest that older patients at risk for CD readmission may benefit from documented eradication of CD prior to hospital discharge or more aggressive post-discharge CD treatment and monitoring. Back to 2014 Annual Meeting Abstracts |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© 2024 Society for Surgery of the Alimentary Tract. All Rights Reserved. Read the Privacy Policy.