Reporting instruction for radiology residents

Abstract

Rationale and objectives: To determine the amount of formal instruction and evaluation about reporting given to radiology residents in the U.S.A., to document report generation methods and to quantify the performance of physician coding.

Materials and methods: E-mail requests with links to a web-based, anonymous survey were sent to program directors of all accredited radiology residencies in the USA. Demographic questions included university or private affiliation, number of residents, geographic location, and number of hospitals covered. Subject-specific items covered the amount of didactic instruction, formal evaluation of reports, and use of structured reports. A didactic activity index (DAI) was calculated as the sum of answers to domain-specific questions and tested for relation to demographic variables. We also asked about dictation methods and International Classification of Diseases (ICD) or Common Procedural Terminology (CPT) coding of examinations by radiologists.

Results: Of the 191 active radiology residencies, 151 (79%) completed the survey. Responses for hours of didactic instruction in reporting given more than a 4-year residency were distributed as follows: 0-1 = 40%, 2-4 = 46%, >4 = 14%. The percentage of resident reports formally graded was distributed as follows: 0-1 = 82%, 2-4 = 8%, >4 = 10%. The extent to which faculty-designed, structured reports were used by residents was distributed as follows: none = 16%, minimal = 25%, few = 17%, some = 33%, most = 9%. The DAI was normally distributed with a mean of 14.8 and a standard deviation of 2.4. Military programs had higher DAIs than university residencies (P = .03). There was no significant relation between any other program demographic variables and the DAI (P > .05). A substantial number of programs reported that physicians performed coding for some or most studies: ICD-9 = 30%, CPT = 26%. The dominant method for report generation was human transcription in 79% followed by speech recognition at 19%. Speech recognition penetration (departments reporting use of the technology for at least some dictation) was estimated to be 38%.

Conclusion: In 86% of sampled radiology residencies, trainees receive no more than one hour of didactic instruction in radiology reporting per year. An aggregate measure of didactic activity about interpretative reporting was identical across all program demographic variables except that military residencies seemed to do slightly more than those located at universities.