Identification of physical abuse cases in hospitalized children: accuracy of International Classification of Diseases codes.

OBJECTIVE: Hospital discharge databases are being increasingly used to track the incidence of child physical abuse in the United States. These databases use International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) codes to categorize illnesses and injuries in hospitalized patients. We assessed the accuracy of the assignment of these codes for cases of child physical abuse. STUDY
DESIGN: Participants were all children (N = 133) evaluated by a child abuse pediatrician (CAP) for suspicion of abuse at Yale-New Haven Children's Hospital from January 1, 2007-December 31, 2010. These children included both those judged to have injuries from abuse and those judged to have injuries accidental/medical in nature. We compared the ICD-9-CM codes entered in the hospital discharge database for each child with the decisions made by the CAPs, as documented in their child abuse registry. The CAPs' decisions were considered to be the gold standard. Sensitivity and specificity were calculated. Medical records were reviewed for cases in which the ICD-9-CM codes disagreed with the CAP's decision.
RESULTS: In 133 cases of suspected child physical abuse, the sensitivity and specificity of ICD-9-CM codes for abuse were 76.7 % (CI 61.4%, 88.2%) and 100% (CI 96.0%, 100%), respectively. Analysis of the 10 cases of abuse not receiving ICD-9-CM codes for abuse revealed that errors in physician documentation (n = 5) and in coding (n = 5) contributed to the reduction in sensitivity.
CONCLUSIONS: Despite high specificity in identifying child physical abuse, the sensitivity of ICD-9-CM codes is 77%, indicating that these codes underestimate the occurrence of abuse.