Variation in CT pediatric head examination radiation dose: results from a national survey.

Abstract

OBJECTIVE. The purpose of this article is to examine the variation in radiation dose, CT dose index volume (CTDIvol), and dose-length product (DLP) for pediatric head CT examinations as a function of hospital characteristics across the United States. MATERIALS AND METHODS. A survey inquiring about hospital information, CT scanners, pediatric head examination protocol, CTDIvol, and DLP was mailed to a representative sample of U.S. hospitals. Follow-up mailings were sent to nonrespondents. Descriptive characteristics of respondents and nonrespondents were compared using design-based Pearson chi-square tests. Dose estimates were compared across hospital characteristics using Bonferroni-adjusted Wald test. Hospital-level factors associated with dose estimates were evaluated using multiple linear regressions and modified Poisson regression models. RESULTS. Surveys were sent out to 751 hospitals; 292 responded to the survey, of which 253 were eligible (35.5% response rate, calculated as number of hospitals who completed surveys [n = 253] divided by sum of number who were eligible and initially consented [n = 712] plus estimated number who were eligible among those who refused [n = 1]). Most respondents reported using MDCT scanners (99.2%) and having a dedicated pediatric head CT protocol (93%). Estimated mean reported CTDIvol values were 27.3 mGy (95% CI, 24.4-30.1 mGy), and DLP values were 390.9 mGy × cm (95% CI, 346.6-435.1 mGy × cm). These values did not vary significantly by region, trauma level, teaching status, CT accreditation, number of CT scanners, or report of a dedicated pediatric CT protocol. However, estimated CTDIvol reported by children's hospitals was 19% lower than that reported by general hospitals (p < 0.01). CONCLUSION. Most hospitals (82%) report doses that meet American College of Radiology accreditation levels. However, [corrected] the mean CTDI(vol) at children's hospitals was approximately 7 mGy (21%, adjusted for covariates), lower than that at nonchildren's hospitals.