Maki Wakana, Kitaoka Hiroki, Aso Shotaro, Ono Sachiko, Kato Motohiro, Yasunaga Hideo
Department of Clinical Epidemiology and Health Economics, School of Public Health, The University of Tokyo, Tokyo, Japan.
Department of Pediatrics, The University of Tokyo, Tokyo, Japan.
Pediatr Int. 2025 Jan-Dec;67(1):e70178. doi: 10.1111/ped.70178.
Administrative claims data are used in clinical studies; however, recorded diagnoses and procedures have not been fully validated for pediatric patients. We aimed to examine the validity of recorded information on pediatric patients in the Japanese Diagnosis Procedure Combination (DPC) database, a national inpatient database that includes administrative claims data.
We validated the DPC data using medical charts as the reference standard. We included patients aged <16 years admitted to a single academic hospital in Tokyo between 2018 and 2022. Positive predictive values were assessed for six diagnoses (cardiac arrest, blood cancer, acute myocarditis, nontraumatic intracranial hemorrhage, hypoplastic left heart syndrome, and trisomy 21) and in-hospital death. We evaluated intubation, mechanical ventilation, and high-flow nasal cannulas in 100 randomly selected patients from the neonatal intensive care unit (NICU) and pediatric intensive care unit, and the gestational age and birth weight in NICU patients.
The positive predictive values of the diagnoses ranged from 70% (non-traumatic intracranial hemorrhage) to 100% (trisomy 21 and in-hospital death). The sensitivities of the procedures ranged from 36% (high-flow nasal cannula in the NICU) to 89% (mechanical ventilation in the NICU), and the specificities were all >95%. The mean gestational ages were 34.2 weeks in the chart and 33.8 weeks in the claims, and the mean birth weights were 2101 g in the chart and 2104 g in the claims, demonstrating high concordance.
Japanese claims data can be useful for pediatric studies focusing on selected diagnoses, procedures, and birth records with confirmed validity.
行政索赔数据用于临床研究;然而,儿科患者的记录诊断和手术信息尚未得到充分验证。我们旨在检验日本诊断程序组合(DPC)数据库中儿科患者记录信息的有效性,该数据库是一个包含行政索赔数据的全国性住院患者数据库。
我们以病历作为参考标准来验证DPC数据。纳入2018年至2022年间入住东京一家学术医院的16岁以下患者。对六种诊断(心脏骤停、血癌、急性心肌炎、非创伤性颅内出血、左心发育不全综合征和21三体综合征)以及院内死亡的阳性预测值进行评估。我们对新生儿重症监护病房(NICU)和儿科重症监护病房中随机抽取的100例患者的插管、机械通气和高流量鼻导管使用情况进行了评估,并对NICU患者的胎龄和出生体重进行了评估。
诊断的阳性预测值范围为70%(非创伤性颅内出血)至100%(21三体综合征和院内死亡)。手术的敏感性范围为36%(NICU中的高流量鼻导管)至89%(NICU中的机械通气),特异性均>95%。病历中的平均胎龄为34.2周,索赔数据中的为33.8周,病历中的平均出生体重为2101克,索赔数据中的为2104克,显示出高度一致性。
日本的索赔数据对于专注于选定诊断、手术和具有确认有效性的出生记录的儿科研究可能有用。