[Assessment of birth registration coverage in the Hospital Information System of the Brazilian Unified National Health System, by hospital of admission, Brazil, 2012-2020].

This study aimed to evaluate the national coverage of birth registration within the Brazilian Hospital Information System of the Brazilian Unified National Health System (SIH/SUS), according to the hospital of birth, and identify associated institutional characteristics. A descriptive ecological study was conducted using available data from the SIH/SUS, the Brazilian Information System on Live Births (SINASC), and the Brazilian National Registry of Health Establishments (CNES) from 2012 to 2020. Hospital admissions of women aged 10 to 49 for vaginal childbirth or cesarean section in the SIH/SUS were compared to live birth records from hospital deliveries in SUS-contracted establishments with more than 100 live births per year in SINASC. Coverage was measured by the proportion of admissions for childbirth (SIH/SUS) in relation to the total number of live births in SINASC. Supervised classification models, decision trees, and random forests were used to identify hospital characteristics predictive of coverage. Coverage in the SIH/SUS was estimated at 86.9%, and 80.6% after excluding hospitals with > 100% coverage. Higher coverage rates were observed in the North and Northeast regions, whereas lower rates were seen in the South and Central-West regions. National coverage increased from 77.9% to 82.3% during the period. The main predictive factors were the proportion of cesarean sections, the number of obstetric beds of SUS, the administrative sphere, and facility size, with lower coverage in services with a higher proportion of cesarean sections. Inconsistencies in CNES registration were identified in SINASC. Birth registration coverage in the SIH/SUS is high across the country, though lower in hospitals with a high proportion of cesarean sections. Strategies for continuous improvement of quality for information system records are necessary.