Impact of clinical history on choice of abdominal/pelvic CT protocol in the Emergency Department

• Published in: PloS one 2018-08, Vol.13 (8), p.e0201694-e0201694
• Main Authors: Dang, Wilfred, Stefanski, Pawel D, Kielar, Ania Z, El-Khodary, Mohamed, van der Pol, Christian, Thornhill, Rebecca, Jaberi, Arash, Fu, Angel Y N, McInnes, Matthew D
• Format: Article
• Language: English
• Subjects: Abdomen; Accuracy; Analysis; Biology and Life Sciences; CAT scans; Communication; Computed tomography; Computerized physician order entry; Consultants; Correlation; Differential diagnosis; Emergency medical care; Emergency medical services; Evaluation; Medical diagnosis; Medical imaging; Medical personnel; Medical protocols; Medical referrals; Medicine; Medicine and Health Sciences; Order entry; Patients; People and Places; Physicians; Practice; Quality; Quality assessment; Radiologists; Reliability analysis; Research and Analysis Methods; Thyroid gland
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0201694

Radiologists and other specialty consultants play a role in diagnosing patients with acute abdominal conditions. Numerous Computed Tomography (CT) protocols are available and radiologists' choices are influenced by the clinical history provided. We hypothesize that the quality of the initial communication between referring physicians and radiologists greatly affects the utilization of health resources and subsequent patient care. The purpose of this pilot study was to employ a grading system to quantitatively evaluate a provided history. We also sought to evaluate inter-rater reliability by having radiologists evaluate sample histories and finally, to assess whether the quality of history has an impact on the number of CT protocols radiologists choose as potentially appropriate, with less potential protocols being seen as a positive outcome. Four reviewers, (2 attendings and 2 residents) evaluated 350 consecutive clinical histories provided for patients presenting to a tertiary care Emergency Department (ED) between September-October, 2012. Reviewers graded histories on a 5-point scale using 4 categories of criteria. This includes a) presenting complaint, b) relevant past medical history or symptom evolution, c) objective laboratory or prior examination results and d) differential diagnosis. There was substantial agreement among all four reviewers when evaluating the quality of history, ICC 0.61, (95% CI 0.48-0.71). In particular, agreement amongst attending radiologists was substantial, with ICC 0.69 (0.48-0.80). Significant negative correlation was observed between history grade and number of potentially appropriate protocols in 3 of 4 reviewers (Spearman's rho: -0.394, -0.639, -0.864, p