Characterizing Antimicrobial Resistance Post-Hospitalization: A Retrospective Analysis Using Administrative Claims Data
Author(s)
Arackal J1, Beyene K2, Venker B3, Moenster R2, Mcdougall V2, Micek S2
1University of Health Sciences and Pharmacy in St. Louis, St. Louis, MO, USA, 2University of Health Sciences and Pharmacy in St. Louis, St Louis, MO, USA, 3Roivant Science, New York, NY, USA
Presentation Documents
OBJECTIVES: To evaluate the incidence and characteristics of antimicrobial resistance (AMR) in patients following hospitalization for serious infections using a comprehensive administrative claims database.
METHODS: This retrospective analysis used Inovalon's closed claims data to identify patients hospitalized for serious infections from January 2017 to December 2021 using a previously validated algorithm. The study followed patients from their initial hospitalization until either the termination of insurance coverage or the onset of AMR. The primary outcome was the incidence of new AMR, identified as a documented ICD-10 claim for AMR post-hospitalization with no prior AMR history in the previous 12 months. Descriptive statistics were used to summarize the data.
RESULTS: A total of 1,225,279 patients were hospitalized for serious infections with an average follow-up duration of 894 days. Of these, 343,757 (28.1%) had received antibiotics within 3 months prior to their hospitalization, and 895,801 (73.1%) received antibiotics during the follow-up. A total of 42,190 patients (3.4%) developed AMR, with the average time to AMR development being 304 days. The most common AMR types were beta-lactam resistance (19,556 cases), vancomycin resistance (3,733 cases), and multidrug resistance (7,261 cases). When comparing geographic regions, the South region had the highest rate of AMR (3.7%) followed by the West (3.5%), Northeast (3.3%), and Midwest (3.0%). AMR was higher in females compared to males, there was also an increase in AMR with age.
CONCLUSIONS: This study is among the first utilizing real-world administrative claims data and ICD-10 codes to estimate AMR incidence post-hospitalization in the United States. Our large sample size and extended follow-up provide a more realistic estimate of the AMR burden and associated characteristics.
Conference/Value in Health Info
Value in Health, Volume 27, Issue 6, S1 (June 2024)
Code
EPH162
Topic
Clinical Outcomes, Epidemiology & Public Health, Study Approaches
Topic Subcategory
Clinical Outcomes Assessment, Public Health
Disease
Drugs, Infectious Disease (non-vaccine)