1471. Medical Claims Analysis as a Tool to Evaluate Empiric and Targeted Antibiotic Therapy in UTIs

Abstract

BackgroundEmpiric therapy is a mainstay of the inpatient management of urinary tract infections (UTIs), and the choice of empiric antibiotic is shaped by local epidemiology and patient risk factors and comorbidities. Pathogen identification (ID) and antibiotic susceptibility testing (AST) provide information that can guide adjustments in therapy, allowing de-escalation to more targeted, narrow-spectrum antibiotics.MethodsReal-world data (hospital billing claims) from 2017 was used to extract relevant information from general hospitals on inpatients with bacterial UTIs including patient demographics and antibiotics line of therapy progression. Patients were projected to the USA national event totals and validated with other projection-related data sources (HCUP) and secondary market research. Data obtained in the claims analysis was validated by primary market research (PMR).ResultsAnalysis of 33M claims identified 169K patients with a code for UTI; in at least one-third of patients, there were no codes associated with ID/AST assays. Among those with codes for ID/AST assays, the vast majority were performed in the first 3 days following hospital admission. Approximately two-thirds of patients with associated ID/AST codes were already receiving an antibiotic when the assays were performed, which was assumed to be the empiric treatment. Analysis of the line of antibiotic therapy progression in patients where ID/AST was performed identified subsequent changes in antibiotic prescribing in approximately one-third of patients within 3 days, compatible with changes due to delivery of conclusive results and were interpreted as a transition from empiric to targeted treatment.ConclusionTo the best of our knowledge, this is the first attempt at understanding the impact of ID/AST assays on prescribing practices with basis on analysis of claims data. Our results align with PMR conducted by DRG internally, supporting the validity of this methodology. Although this claims analysis delivers reliable data when claims are associated with ID/AST assays, it is limited by incompletely filled claims, which may underestimate the use of these assays.Disclosures All authors: No reported disclosures.