Caregiver questionnaire responses collected during routine 18-month pediatric health checkups in Japan can stratify families by their likelihood of administrative registration with a municipal child protection coordination system, according to a cross-sectional secondary analysis of administrative records. Child maltreatment represents a major global public health concern, and researchers sought to determine whether routine health screening tools could assist frontline personnel in identifying households in need of coordinated municipal assistance.
In Japan, municipal governments mandate health checkups for children at 18 months and 3 years of age, achieving average community attendance rates exceeding 96 percent. Questionnaires completed by caregivers during these visits routinely capture information regarding the home environment and primary caregiving circumstances. The country also operates Yōtaikyō, a municipal child protection coordination network that administratively registers children identified as experiencing or facing elevated risks of maltreatment to provide structured support.
To investigate whether checkup responses could reliably associate with subsequent child protection tracking, investigators analyzed administrative records from a single Japanese municipality. The study cohort comprised 11,072 children born between fiscal years 2016 and 2018 whose caregivers attended both statutory examinations. Researchers screened candidate questionnaire items from each checkup using a significance threshold of P < .01, subsequently entering eligible variables into stepwise logistic regression models to compute weighted risk scores for each household.
The primary outcome evaluated in the statistical analysis was a documented history of registration with the Yōtaikyō child protection coordination system through April 2023. Model performance and discriminatory ability were assessed using the area under the receiver operating characteristic curve alongside bootstrap internal validation and calibration analysis. The investigative team developed separate predictive models for the 18-month and 3-year examinations to evaluate how predictive capacity changed as children aged.
The predictive model derived from 18-month questionnaire records analyzed 10,959 children and retained 14 distinct predictor variables. This model demonstrated an apparent area under the receiver operating characteristic curve of 0.769, an optimism-corrected value of 0.763, and a calibration slope of 0.976. Within this evaluation, the strongest individual predictors were a caregiver reporting a public health nurse as a source of support and an absence of spousal support.
Applying the optimal cut-point derived from the 18-month scoring system categorized 24.3 percent of the analyzed cohort into a designated high-risk group. Children assigned to this high-risk classification exhibited a 5.4-fold higher rate of registration with the municipal coordination system than peers assigned to the low-risk classification, indicating significant separation between stratified tiers in administrative tracking.
By contrast, the model constructed using 3-year questionnaire records incorporated eight predictor variables across a sample of 10,643 children but exhibited lower overall discrimination. That model showed an apparent area under the curve of 0.663 and an optimism-corrected metric of 0.659. In the 3-year analysis, poor physical or mental health reported by the caregiver emerged as the single strongest predictive variable for system registration.
The study authors concluded that questionnaire information routinely gathered during universal 18-month examinations stratifies families by their probability of Yōtaikyō registration with sound internal validity. Implementing systematic risk-stratification models within municipal health checkups and broader pediatric healthcare environments could provide an objective method to complement the clinical judgment of frontline child health practitioners seeking to deliver early family support.
Subject of Research: Child protection risk stratification using universal pediatric health checkup questionnaires in Japan
Article Title: Risk stratification for registration with a municipal child protection coordination system using caregiver questionnaire data from universal child health checkups in Japan: a cross-sectional secondary analysis of administrative data
Article References: Kanezaki, A., Inoue, M., & Ihara, K. (2026). Risk stratification for registration with a municipal child protection coordination system using caregiver questionnaire data from universal child health checkups in Japan: a cross-sectional secondary analysis of administrative data. BMC Pediatrics. https://doi.org/10.1186/s12887-026-08081-x
Image Credits: AI Generated
DOI: 10.1186/s12887-026-08081-x
Keywords: child maltreatment, pediatric checkups, Japan public health, Yotaikyo registration, child protection, risk stratification, caregiver questionnaires, administrative data, logistic regression, AUC discrimination, internal validation, spousal support, public health nurse, early intervention, pediatrics
News Source: Harold Sullivan. (November 20, 2026). Japan Checkup Data Stratify Child Protection System Registration Risk. Scienmag.



