Variation in the length and structure of reports written by reporting radiographers: A retrospective study

The literature suggests that there is variation in various features of the written radiology report for a range of body areas and imaging modalities. The retrospective study presented here aims to determine if similar variation is demonstrated in a group of 5 reporting radiographers in a UK NHS Trust.

Full reports for 1530 knee radiographic examinations performed from accident and emergency referrals were extracted for a 12-month period from a Radiology Information System (RIS) into Excel. Copied into Word, the word count function was used for each report and the number of words and characters (without spaces) was returned into Excel. Average word count and word length per report, by radiographer, were calculated for the following sections of the report: report title, main body and signature. SPSS was used to perform inferential statistical analysis.

A wide range in the maximum and minimum average report lengths (60.88 v 17.83 words) was demonstrated. Statistically significant differences (p < 0.05) were seen between all but one pair-wise comparison (Rad 2 v Rad 4; p = 0.98) for the overall report length; for the length of the findings section, four pair-wise comparisons did not reach significance. Average word length demonstrated less variation. 4 out of 5 radiographers always included a report title; 3 out of 5 never included a report signature. There was a strong negative correlation between experience and report length.

Variation in report structure and length, as well as word length, was seen, comparable to studies of radiologist reports. Further research is required to investigate the drivers of this variation, and determine if there is any clinical significance.