Abstract 246: Cost-analysis and Clinical Outcomes in Medicare Patients Who Have Received Holter Monitoring

Introduction: The diagnostic odyssey for arrhythmia detection utilizing ambulatory ECG monitoring is not well documented. Our aim was to characterize the diagnosis, additional monitoring, clinical events and costs that occur after an index Holter in Medicare patients. Methods: We performed a retrospective, longitudinal analysis of the Medicare Standard Analytical Files (SAF), a 5% random sample of Medicare fee-for-service beneficiaries that contains every fully adjudicated claim. Beneficiary IDs were encrypted and consistent between all claims in a given year and across years. We studied years 2008 through 2011 in Medicare Carrier (Part B), Inpatient, Outpatient, and Denominator files. Inclusion criteria: 1) beneficiary with an index Holter in 2009, 2) enrolled in Medicare Parts A and B from one year prior to the date of the first Holter to death date or two years after the date of the index Holter event in 2009. Exclusion criteria: 1) use of a Holter or other cardiac monitoring device one year prior to the date of the first Holter use in 2009, 2) diagnosis of atrial fibrillation or ablation one year prior to the date of the first Holter. The population with an index Holter performed in 2009 was 46,840. After applying exclusion criteria, the study cohort reduced to 17,887 patients. To meet the aims of the study, four areas were identified: diagnosis, additional monitoring, clinical events and costs. Diagnosis was identified by ICD-9 codes and additional ambulatory ECG monitoring by CPT codes and date of service. Clinical events were identified by site of service, ICD-9 codes for Emergency Department, Inpatient, Observation stays and ICD-9 codes for stroke, TIA, and cardiac arrests. Principal diagnosis was used for clinical events. Costs were derived from claims data. Results: Eleven percent of the population attained a diagnosis with an index Holter without additional monitoring or clinical event (diagnosis attained cohort) and incurred total allowed charges in the two year period (charges) of $48 mil. Over a quarter (27%) of the population did not reach a diagnosis (no diagnosis cohort) and continued in the diagnostic and clinical event odyssey. Half of the patients in the no diagnosis cohort (13%) - experienced clinical events despite repeat monitoring and incurred charges of $63 million dollars. A significant number of those patients had their first clinical event within three months of their index Holter including stroke, TIA and cardiac arrest. The other half of the no diagnosis cohort (14%) received additional cardiac monitoring and total charges of $61 mil. Holter was the most commonly repeated test. Conclusions: Repeat monitoring, often of the same type as done initially, frequently did not yield a diagnosis and patients continued to experience clinical events at a substantial healthcare cost. Additional diagnostic paradigms should be explored to improve these patient and system outcomes.