Establishment of national diagnostic reference levels for computed tomography procedures in Sri Lanka: first nationwide dose survey.

The main purpose of this study was to establish for the first time national diagnostic reference levels (NDRLs) for common computed tomography (CT) procedures in Sri Lanka. Patient morphometric data, exposure parameters and dose data such as volume CT dose index (CTDIvol) and dose-length product (DLP) were collected from 5666 patients who underwent 22 types of procedure. The extreme dose values were filtered before analysis to ensure that the data come from standard size patients. The median of the dose distribution was calculated for each institution, and the third quartile value of the median distribution was considered as the NDRL. Based on the inclusion and exclusion criteria, data from 4592 patients and 17 procedure types were considered for establishment of a NDRL, covering 41% of the country's CT machines. The proposed NDRLs based on CTDIvoland DLP were: non-contrast-enhanced (NC) head, 82.2 mGy/1556 mGy cm; contrast-enhanced (CE) head, 82.2 mGy/1546 mGy cm; chest NC, 7.4 mGy/350 mGy cm; chest CE, 8.3 mGy/464 mGy cm; abdomen NC, 10.5 mGy/721 mGy cm; abdomen arterial (A) phase, 13.4 mGy/398 mGy cm; abdomen venous (V) phase, 10.8 mGy/460 mGy cm; abdomen delay (D) phase, 12.6 mGy/487 mGy cm; sinus NC, 30.2 mGy/452 mGy cm; lumbar spine NC, 24.1 mGy/1123 mGy cm; neck NC, 27.5 mGy/670 mGy cm; high-resolution CT of chest, 10.3 mGy/341 mGy cm; kidneys ureters and bladder NC, 19.4 mGy/929 mGy cm; chest to pelvis (CAP) NC, 10.8 mGy/801 mGy cm; CAP A, 10.4 mGy/384 mGy cm; CAP V, 10.5 mGy/534 mGy cm; CAP D, 16.8 mGy/652 mGy cm. Although the proposed NDRLs are comparable with those of other countries, the observed broad dose distributions between the CT machines within Sri Lanka indicate that dose optimisation strategies for the country should be implemented for most of the CT facilities.