Presentation (CTI)

OBJECTIVES: The National Agency of Supplementary Health from Brazil (ANS) collects standardized hospital procedure data from the private health insurance operators (PHIO). This data is openly shared but underutilized for cost inputs in ANS submissions. This study explores ANS data to generate condition-specific costs from the ANS perspective, capturing geographic and PHIO modality variations, using multiple myeloma (MM) hospitalization as a case study.
METHODS: ANS Hospital Procedures open data (2020-2024) was analyzed. MM hospitalization events (ICD C90.0) were identified and linked to detailed cost data. Events were filtered for length of stay >24 hours, non-zero total value, and no pre-established service packages. Descriptive analysis examined variations by state and PIHO modality.
RESULTS: Of 33,641,733 hospital events, 37% had missing diagnosis data and 12,467 involved MM. Of these, 8.625 (69%) had the minimum LOS, non-zero total value and not-predefined service package. 74.6% of the events from the Southeast region, contrasting with the proportion of the Brazilian population living in that area, 41.6%. Overall, the mean (SD) claimed value by event was BRL $60,557 ($92,116), varying considerably between different Brazilian states ($5,147-$142,609). Also, mean costs varied by PIHO modality, with Specialized Health Insurer presenting twice ($78,692) the value reported by Philanthropy modality ($37,942).
CONCLUSIONS: Despite underreporting of diagnoses (37%) and Southeast data concentration, the ANS Hospital Procedures dataset excels in capturing cost variability across Brazil’s private health system, far better than the use of a single PIHO source or reference tables mapped to expert opinions, which is often the case in ANS submissions. This publicly available resource provides reliable, granular insights into geographic and operator-specific differences, making it an essential tool for accurate health economic analyses.