Related Experiment Video
Updated: Aug 17, 2026
Inverse Probability of Treatment Weighting (Propensity Score) using the Military Health System Data Repository and National Death Index
Published on: January 8, 2020
Concordance between childhood injury diagnoses from two sources: an injury surveillance system and a physician
A Kostylova1, B Swaine, D Feldman
1Centre de recherche interdisciplinaire en réadaptation du Montréal métropolitain, Institut de réadaptation de Montréal; Montreal, Quebec, Canada.
Insights
Physician billing claims data show substantial concordance with injury surveillance systems for identifying pediatric head injuries and orthopedic injuries. This suggests billing data can reliably estimate injury occurrence in children.
Area of Science:
- Public Health
- Epidemiology
- Health Informatics
Background:
- Accurate injury surveillance is crucial for public health.
- Comparing data sources like injury surveillance systems and billing claims is essential for validating data.
- Understanding the concordance between different data sources can improve injury research.
Purpose of the Study:
- To determine the concordance between injury diagnoses (head injury, probable head injury, orthopedic injury) captured by an injury surveillance system and a physician billing claims database for children presenting to emergency departments.
- To assess the sensitivity and specificity of diagnostic and procedure codes in physician billing claims for identifying head injuries and orthopedic injuries in children.
Main Methods:
- A cross-sectional cohort study linked data from 3049 children (2000-01) using their health insurance numbers.
- Children were categorized into head injury, probable head injury, and orthopedic injury groups using physician-recorded codes from the Canadian Hospitals Injury Research and Prevention Program (CHIRPP).
- An algorithm using ICD-9-CM diagnostic and procedure codes from the Regie de l'assurance maladie de Quebec (RAMQ) database classified children into the same three injury groups.
Main Results:
- Substantial concordance (weighted Kappa 0.66) was found between the CHIRPP surveillance system and the RAMQ billing claims database.
- The sensitivity of RAMQ codes for identifying head injuries was 0.61, and for orthopedic injuries was 0.97.
- The specificity of RAMQ codes for identifying head injuries was 0.97, and for orthopedic injuries was 0.58.
Conclusions:
- Combining diagnostic and procedure codes in physician billing claims databases, such as RAMQ, appears to be a valid method for estimating pediatric injury occurrence.
- Billing data offers a potentially valuable resource for injury surveillance and research when used appropriately.
- Further research may explore optimizing algorithms for improved accuracy in injury coding from billing claims.
Objectives:
(1) To determine the concordance between injury diagnoses (head injury (HI), probable HI, or orthopedic injury) for children visiting an emergency department for an injury using two
Data Sources:
an injury surveillance system (Canadian Hospitals Injury Research and Prevention Program, CHIRPP) and a physician billing claims database (Regie de l'assurance maladie de Quebec, RAMQ), and (2) to determine the sensitivity and specificity of diagnostic and procedure codes in billing claims for identifying HI and orthopedic injury among children.
Design:
In this cross sectional cohort, data for 3049 children who sought care for an injury (2000-01) were obtained from both sources and linked using the child's personal health insurance number.
Methods:
The physician recorded diagnostic codes from CHIRPP were used to categorize the children into three groups (HI, probable HI, and orthopedic), while an algorithm, using ICD-9-CM diagnostic and procedures codes from the RAMQ, was used to classify children into the same three groups.
Results:
Concordance between the data sources was "substantial" (weighted Kappa 0.66; 95% CI 0.63 to 0.69). The sensitivity of diagnostic and procedure codes in the RAMQ database for identifying HI and for orthopedic injury were 0.61 (95% CI 0.57 to 0.64) and 0.97 (95% CI 0.96 to 0.98), respectively. The specificity for identifying HI and for orthopedic injury were 0.97 (95% CI 0.96 to 0.98) and 0.58 (95% CI 0.56 to 0.63), respectively.
Conclusion:
Combining diagnostic and procedures codes in a physician billing claims database (the RAMQ database) may be a valid method of estimating injury occurrence among children.
