The impact of guidelines on the proportion of OA consultations associated with an X-ray request: a time trend analysis

Problem

Osteoarthritis (OA) is a frequent cause of primary care consultations. X-rays have often been used to confirm a diagnosis of OA. The current NICE and Royal College of Radiologists (RCR) guidelines discourage the routine use of X-rays as they do not improve diagnostic confidence, have cost implications, and may inappropriately alter management. It is unclear if guidelines have affected X-ray requests. This study aimed to describe trends in X-ray requests in primary care OA consulters, and the potential impact of guideline publication.

Approach

Consultation data from seven general practices in North Staffordshire were extracted from a primary healthcare database (CiPCA). The practice and overall annual proportion of X-ray requests in OA consulters from 2000-2012 (defined as at least one X-ray Read code within 30 days of an OA consultation) were calculated. More recent data was unavailable for analysis. Joinpoint regression analysis was used to construct trend models; a likelihood ratio test statistic was calculated to identify the model with the best fit to the observed data. Joinpoint dates were compared with the publication dates of four relevant UK guidelines.

Findings

Between 2000-2012, 18,114 patients had at least one recorded OA consultation. The proportion of X-ray requests in patients consulting for OA ranged between 14.3% in 2000 to 19.8% in 2003 (mean 17.3%). Marked variation existed between GP practices, with quarterly X-ray requests ranging from 10.5% to 25.3%. From 2000 Q1-2003 Q2, X-ray requests increased by 2.6% per quarter (95% CI:1.0%, 4.3%). From 2003 Q2-2012 Q4 the model identified a fall in the proportion of X-ray requests by 0.5% per quarter (95% CI: -0.8%, -0.2%). This amounted to a 16.4% decrease from 2003 Q2-2012 Q4. The joinpoint at 2003 Q2 coincided with the publication date of the RCR 2003 guideline (June 2003).

Consequences

The high proportions of X-ray requests in OA consulters found in this study are consistent with other studies. The publication date of one of the four national guidelines coincided with a change in trend. The simultaneous nature of the publication of the RCR guideline and the change in trend is not necessarily supportive of a causal association, as there would have been no time for dissemination. Therefore, the cause of the reduction in X-ray requests is unknown, though guidelines appear to have a limited impact on X-ray requests. The need to look beyond guidelines for an effective solution is highlighted in the latest NHS report on 'Transforming radiology elective care services' (2019). Further exploration of the variation seen between practices may help identify factors that lead to an X-ray request for OA.