• B Krug, M Boettge, T Reineke, S Coburger, M Zähringer, U Harnischmacher, M Lüngen, K W Lauterbach, W Lehmacher, and K Lackner.
• Institut und Poliklinik für Radiologische Diagnostik der Universität zu Köln, Cologne. Barbara.Krug@medizin.uni-koeln.de
• Rofo. 2003 Mar 1; 175 (3): 346-60.

PurposeIn the state of North Rhine-Westphalia (NRW), Germany, a survey was conducted on radiologic examinations ordered by general practitioners (GPs). Part II of this study aims to determine the quality of the process and outcome. The reference standard is the assessment of both radiologists and physicians without board certification in radiology working at a university hospital and in outpatient facilities.Materials And MethodsAllGPs in NRW were asked to cooperate. Participating GPs filled out a questionnaire for each patient. The patients recorded the symptoms prompting the imaging examinations. The radiologists or other physicians performing the examinations were asked to provide the images and written reports and to complete a questionnaire. A file was created for each of the 394 patients with image documentation of at least one examination. Each file, which included medical history, physical findings, imaging documentation and written report, was sequentially forwarded to a board-certified radiologist and to a physician without board certification in radiology working in a university hospital and in an outpatient facility. All physicians were requested to complete a structured questionnaire for each file.ResultsThe referral diagnoses were rated as medically plausible in 81%, the indications for imaging found correct in 76%, the examination techniques considered appropriate in 69%, the clinical question answered in 63%, the interpretation judged medically correct in 50% and all incidental findings documented in 49%. In retrospect, 32 % of the examinations were judged superfluous. The sequence of multiple examinations performed on a particular patient was rated as appropriate in 51%. The interpretation revealed specialty-related differences. The plausibility of the referral diagnoses had a significant impact on the appropriateness of subsequent diagnostic investigations. Marked deficits showed sonography, performance by non-radiologists, self-referrals by GPs, gastroenterologic radiology and the ICD-10 coding (suspicion of cardiovascular disease).ConclusionIn the "best-case" scenario, the process quality proved to have moderate deficiencies and the outcome quality severe deficiencies. In consequence, GPs and radiologists should be more communicative by sharing information and exchanging opinions. GP self-referrals should be restricted. Sonography and examinations performed by physicians without board certification in radiology should undergo stricter quality controls. A more intensive interdisciplinary collaboration is needed to determine the optimum implementation of diagnostic imaging of gastroenterologic and cardiovascular diseases.

28 keywords...

hide…