2219. Evaluation of Medicare Claims to Assess Burden of Pertussis Disease in Persons Aged ≥65 Years

BACKGROUND: Pertussis in adults may be underdiagnosed and underreported; there is limited information on the incidence and severity of pertussis in older adults. We compared pertussis diagnoses identified using medical claims data with national surveillance data to examine the use of claims data as a source for disease burden estimates. METHODS: We examined claims data in persons aged ≥65 years in the United States enrolled in Medicare A and B from January 1, 2008 to December 31, 2017. We identified provider-diagnosed pertussis through pertussis-related ICD9/ICD10 diagnostic codes (033.XX, 484.3, A37.XX). We examined whether any were categorized as inpatient claims and if there were claims for laboratory tests within 30 days of the initial pertussis claim. We estimated claims-based pertussis incidence using person-time for all Medicare A/B enrollees and compared incidence estimates to those reported by the Nationally Notifiable Diseases Surveillance System (NNDSS) for the same period and age group. RESULTS: Among 27,269,361 Medicare beneficiaries, 24,355 (0.09%) had claims with pertussis diagnostic codes. Of these, 1,875 (7.7%) had claims associated with inpatient hospitalizations; 7,964 (33%) had laboratory testing performed. The mean annual incidence of claims-based pertussis was 11.5/100,000 person-years (range: 9.37 to 14.3/100,000 person-years) (figure). In contrast, 6,722 pertussis cases in persons aged ≥65 years were reported to NNDSS. Among the 5,101 cases whose hospitalization status was known, 783 (15%) were hospitalized. Mean annual reported pertussis incidence was 1.5/100,000 person-years (0.67 cases to 2.63 cases/100,000 person-years) in this age group. CONCLUSION: Many more Medicare beneficiaries with pertussis-related claims were identified than pertussis cases in persons ≥65 years reported to public health authorities, suggesting pertussis is likely diagnosed more frequently in older adults than national incidence estimates indicate. A smaller proportion of Medicare beneficiaries with pertussis-related claims were hospitalized compared with reported cases and a majority did not have laboratory testing performed. It is unknown what proportion of pertussis-associated claims represent true pertussis disease. [Image: see text] DISCLOSURES: All authors: No reported disclosures.