Related Experiment Video
Updated: Aug 20, 2026
Assessment of Intestinal Transcytosis of Neonatal Escherichia coli Bacteremia Isolates
Published on: February 17, 2023
Child maltreatment in Missouri: combining data for public health surveillance
Patricia G Schnitzer1, Paula Slusher, Mark Van Tuinen
1Department of Family and Community Medicine, University of Missouri, Columbia, Missouri 65212, USA. schnitzerp@health.missouri.edu
Insights
Public health surveillance for child maltreatment in Missouri linked child protective services (CPS) data with medical records. Linking these data sources identified child abuse and neglect cases but medical data added few unique cases.
Area of Science:
- Public Health
- Epidemiology
- Child Welfare
Background:
- Established public health surveillance for nonfatal child maltreatment in Missouri using dual data sources.
- Described child maltreatment epidemiology via this surveillance system.
- Assessed the utility of integrating medical data with child protective services (CPS) reports for surveillance.
Purpose of the Study:
- To describe the epidemiology of nonfatal child maltreatment in Missouri.
- To evaluate the effectiveness of combining medical data with CPS reports for surveillance.
Main Methods:
- Linked child abuse and neglect reports (2000) from Missouri's Division of Family Services (DFS) with hospital discharge and emergency department (ED) data.
- Classified children as maltreated based on ICD codes or substantiated DFS reports.
- Validated ICD maltreatment codes using a 10% sample of medical records.
Main Results:
- Identified 5657 nonfatal child maltreatment cases (rate: 7.4/1000 children).
- Higher rates observed in infants, females, African Americans, and rural populations.
- Medical data uniquely identified only 10% of cases, though it captured more African American, urban, physically abused, and infant cases.
Conclusions:
- Child protective services (CPS) and medical data are linkable for surveillance purposes.
- Medical data contribute minimally to identifying unique maltreatment cases.
- Medical data capture only a subset of all maltreated children.
Background:
Public health surveillance was established for nonfatal child maltreatment in Missouri using two data sources. This paper describes the epidemiology of nonfatal child maltreatment using this surveillance system, and assesses the usefulness of combining medical data with child protective services (CPS) reports for child maltreatment surveillance.
Methods:
Child abuse and neglect reports for children aged <10 years for 2000 were obtained from the Missouri Division of Family Services (DFS) and linked to hospital discharge and emergency department (ED) data for 2000. Children were classified as maltreated if they had an International Classification of Disease (ICD) maltreatment code for a hospital or ED visit, or a substantiated report to DFS. Validity of the ICD maltreatment codes was assessed in a 10% random sample of the hospital/ED visits. Medical records were reviewed to determine the accuracy of the maltreatment code assigned. The data analyses reported here were conducted in 2002 and 2003.
Results:
In the linked data, 5657 children met the case definition of maltreatment, providing a nonfatal maltreatment rate of 7.4/1000 children. Rates were higher among children aged <1 year (8.4/1000), females (7.7/1000), African Americans (11.8/1000), and children residing in rural counties (8.7/1000). The hospital/ED data identified proportionately more children who were African American, urban, physically abused, or infants, and uniquely identified only 10% of the total cases. In the validation sample, maltreatment was documented in 110 (87%) of the 127 records reviewed.
Conclusions:
CPS and medical data can be linked for surveillance. However, the medical data add few unique cases and identify only a subset of maltreated children.
Related Concept Videos
01:23Investigation of Disease Outbreaks
01:26Principles of Disease Surveillance