Andrew Kent Moriarity MD, Presenter: Nothing to Disclose

Safwan Halabi MD, Abstract Co-Author: Nothing to Disclose

ACR Appropriateness Criteria (AC) exist for 183 clinical scenarios in 22 specialty areas. These consensus opinions of recognized experts and stakeholders are based on currently available evidence. Computerized radiology order entry (CROE) systems allow greater integration of decision support (DS) software, which is often based on AC. Radiation and cost savings are predicted through increased use of DS. There is no current data on how closely ordering habits without the use of DS already follow the AC. We examine how closely the provided clinical information from our hospitalist service matches current AC for the requested examination.

All CROE requests by the hospitalist service were collected for 6 months. Providers selected a ‘relevant clinical scenario’ and ‘signs and symptoms’ from a pull-down menu during CROE. No DS was presented; an AC score (1-9) was generated but not displayed. Protocolling did not alter the initial score. Scores were analyzed according to available AC.

Using CROE, 3737 requests were submitted for 1859 patients by 21 of the 38 employed hospitalists. 2788 requests (74%) were for ‘non-advanced imaging’ and excluded from scoring. No AC was available for the provided clinical scenario in 685 requests (18%). 264 requests (7%) had an applicable AC and a score was automatically generated at the time of the request; the average score was 7 (range 1 to 9).

Examinations with an applicable AC had an average score of 7, “Usually appropriate” per ACR definitions. The range of scores encompassed all possible values from 1, “Usually not appropriate,” up to the maximum score of 9. We discuss variances from AC with regard to modality, clinical history/scenario and hospitalist demographics including type of graduate medical education and years in practice; we address how these may influence ordering habits. We also identify multiple common clinical scenarios for which there is no current AC. 

Significant baseline variations from ACR exist, there are potential radiation and cost savings among inpatient examinations if DS can improve AC adherence.

Moriarity, A, Halabi, S, Examining the Balance between Clinical Information, Physician Ordering Habits and Decision Support Criteria: A Comparison of Typical Hospitalist Imaging Requests with the Current ACR Appropriateness Criteria.  Radiological Society of North America 2012 Scientific Assembly and Annual Meeting, November 25 - November 30, 2012 ,Chicago IL. http://archive.rsna.org/2012/12035336.html