A Practical Approach to Identifying Early Neonatal Conditions in a De-Identified Database: National Health Insurance Research Database (NHIRD)
Author(s)
Yu-Hsuan Kuo, MSc, PharmD1, Hung-Wei Lin, MS, RPh1, Kai-Pei Chou, MSc1, Yuan-Liang Wen, PhD Candidate2, Sheng-Yin To, PhD2, Hui-Wen Yang, PhD2, Li-Ting Kao, PhD2.
1Real World Solutions, IQVIA Solutions Taiwan, Taipei, Taiwan, 2Graduate Institute of Life Sciences, National Defense Medical Center, Taipei, Taiwan.
1Real World Solutions, IQVIA Solutions Taiwan, Taipei, Taiwan, 2Graduate Institute of Life Sciences, National Defense Medical Center, Taipei, Taiwan.
OBJECTIVES: Mother-baby linkage (MBL) is fundamental for longitudinal cohort studies in pregnancy to connect prenatal exposures with newborn outcomes. In Taiwan, although the Birth Reporting Database (BRD) provides comprehensive newborn outcomes at birth, there is no reliable way to collect neonatal (i.e., 28 days after birth) information due to the delayed time for granting identification number (ID) for newborns. Therefore, the study aimed to propose an alternative approach by using maternal IDs to identify neonatal conditions.
METHODS: The National Health Insurance Research Database (NHIRD) was utilized to identify the neonatal conditions through International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM: P00-P96). The IDs of existing pregnant women in BRD served as reference maternal IDs. Moreover, neonatal deaths were also captured from the National Death Registry (NDR) to confirm whether their IDs could be linked to maternal IDs. The data periods of all data sources were between 2022 and 2023.
RESULTS: A total of 71,920 patients with diagnosis of neonatal conditions were identified during inpatient visit, of whom 86.5% of IDs could be matched to the maternal IDs in the BRD. The proportions remained similar (50.6% and 86.1%) for neonatal infections (P35-P39). In the NDR, 760 neonatal deaths were identified. However, none of their IDs could be matched to maternal IDs in the BRD.
CONCLUSIONS: The study findings suggested that in NHIRD, neonatal outcomes requiring hospitalization could be identified by maternal IDs before the neonate’s ID is granted. This alternative approach enables longitudinal follow-up from pregnancy to neonatal period when MBL is not available. Moreover, the unmatchable IDs of neonatal deaths indicated that death registry were documented with neonates’ IDs instead of maternal ones. Further research may focus on establishing valid family ID linkages to complement the methodology of identifying neonatal conditions through father’s ID.
METHODS: The National Health Insurance Research Database (NHIRD) was utilized to identify the neonatal conditions through International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM: P00-P96). The IDs of existing pregnant women in BRD served as reference maternal IDs. Moreover, neonatal deaths were also captured from the National Death Registry (NDR) to confirm whether their IDs could be linked to maternal IDs. The data periods of all data sources were between 2022 and 2023.
RESULTS: A total of 71,920 patients with diagnosis of neonatal conditions were identified during inpatient visit, of whom 86.5% of IDs could be matched to the maternal IDs in the BRD. The proportions remained similar (50.6% and 86.1%) for neonatal infections (P35-P39). In the NDR, 760 neonatal deaths were identified. However, none of their IDs could be matched to maternal IDs in the BRD.
CONCLUSIONS: The study findings suggested that in NHIRD, neonatal outcomes requiring hospitalization could be identified by maternal IDs before the neonate’s ID is granted. This alternative approach enables longitudinal follow-up from pregnancy to neonatal period when MBL is not available. Moreover, the unmatchable IDs of neonatal deaths indicated that death registry were documented with neonates’ IDs instead of maternal ones. Further research may focus on establishing valid family ID linkages to complement the methodology of identifying neonatal conditions through father’s ID.
Conference/Value in Health Info
2025-09, ISPOR Real-World Evidence Summit 2025, Tokyo, Japan
Value in Health Regional, Volume 49S (September 2025)
Code
RWD138
Topic Subcategory
Health & Insurance Records Systems
Disease
SDC: Pediatrics