An electronic tool for systematic reporting of fractures on skeletal surveys in suspected child abuse: prototype development and physician feedback.

OBJECTIVE: To describe a skeletal survey data entry and compilation tool and assess physician attitudes toward this reporting approach. BACKGROUND: Narrative skeletal survey reports are highly variable and prone to inconsistencies with potential adverse impact on patients. MATERIALS AND METHODS: The prototype skeletal survey data entry and compilation tool was developed and introduced into clinical practice at a large urban children's hospital. Pediatric radiologists and child protection team (CPT) pediatricians completed a survey of reporting preferences. Skeletal survey reports between March 1, 2013, and March 1, 2014, were reviewed to assess use of the tool. RESULTS: The survey response rate was 70% (14/20) for radiologists and 100% (4/4) for CPT pediatricians. Among responding radiologists, 54.5% (6/11) indicated that a skeletal survey data entry and compilation tool was helpful for skeletal surveys with >3 fractures; 80% (8/10) of responding radiologists indicated that tabulated data from prior skeletal survey was helpful when interpreting a follow-up skeletal survey with >3 fractures; 90.9% (10/11) of radiologists thought the tool improved report organization; 72.7% (8/11) thought it improved accuracy. Most radiologists (11/12, 91.7%) and 100% (4/4) of CPT clinicians preferred reports with both free text and a tabulated fracture list for testifying in court when >3 fractures were present. The tool was used in the reporting of 14/23 (61%) skeletal surveys with >3 fractures during a 1-year period. A case example using the application is presented. CONCLUSION: Most radiologists and CPT physicians at our center prefer skeletal survey reports with tabulated data and narrative description; 91.7% (11/12) of radiologists and all CPT clinicians prefer this approach for testifying in court when >3 fractures are present.

Yield of radiographic skeletal surveys for detection of hand, foot, and spine fractures in suspected child abuse.

OBJECTIVE: Previous studies have found that fractures involving the spine, hands, and feet are rare on skeletal surveys in cases of suspected child abuse, leading some authors to suggest eliminating these regions from the initial skeletal survey protocol. We assessed this recommendation by performing a historical review of these injuries in a pediatric population undergoing film screen-based radiographic skeletal surveys for suspected child abuse. MATERIALS AND METHODS: This cross-sectional retrospective study reviewed reports of initial skeletal surveys of all children younger than 2 years with suspected abuse imaged between April 1988 and December 2001. Radiographic skeletal survey imaging was performed according to American College of Radiology standards. Sixty-two percent (225/365) of all skeletal surveys had positive findings, and 44% (98/225) showed more than one fracture. Surveys with fractures involving the spine, hands, or feet were identified, and the data were tabulated and analyzed. RESULTS: Twenty of 365 studies (5.5%) yielded fractures involving the spine, hands, or feet. Of all positive skeletal surveys, 8.9% (20/225) had fractures involving the spine, hands, or feet. Of all patients with more than one fracture on skeletal survey, 20.4% (20/98) had fractures involving these regions. CONCLUSION: These data, acquired during the film-screen era, suggest that fractures of the spine, hands, and feet may not be rare in infants and toddlers in cases of suspected child abuse. The benefits of eliminating views of these regions from the initial skeletal survey should be carefully weighed against the cost of missing these potentially important injuries in at-risk pediatric populations.