Reduction of routine radiographs in the follow‐up of distal radius and ankle fractures: Barriers and facilitators perceived by orthopaedic trauma surgeons

Abstract Rationale, aims, and objectives Studies suggest that routine radiographs during follow‐up of distal radius and ankle fractures result in increased radiation exposure and health care costs, without influencing treatment strategies. Encouraging clinicians to omit these routine radiographs is challenging, and little is known about barriers and facilitators that influence this omission. Therefore, this study aims to identify barriers and facilitators among orthopaedic trauma surgeons that might prove valuable towards the design of a deimplementation strategy. Methods A mixed‐method approach was used. First, interviews were conducted with orthopaedic trauma surgeons and patients (n = 16). Subsequently, a questionnaire was developed. This questionnaire was presented to 228 orthopaedic trauma surgeons in the Netherlands. Regression analyses were performed in order to identify which variables were independently associated to the decision to stop performing routine radiographs 6 and 12 weeks after trauma if proven not effective in a large randomized controlled trial. Results In total, 130 (57%) respondents completed the questionnaire. Of these, 71% indicated they would stop ordering routine radiographs if they were proven not effective. Three facilitators were independent predictors for the intention to omit routine radiographs: This will “lead to lower health care costs” (Odds Ratio [OR]: 5.38 and 4.38), the need for “incorporation in the regional protocol” (OR: 3.66 and 2.66), and this will “result in time savings for the patient” (OR: 4.84). Conclusions We identified three facilitators that could provide backing for a deimplementation strategy aimed at a reduction of routine radiographs for patients with distal radius and ankle fractures.


| INTRODUCTION
Over the past decade, the reduction of low-value care has become progressively more important to increase the overall quality of health care.
One of the driving forces behind this change is the "Choosing Wisely" campaign, which started in 2012 in the United States. Choosing Wisely is committed to reducing the use of diagnostic tests, treatments, and procedures if there is evidence of overuse, potential harm, or significant and unjustifiable costs. 1 Routine radiography in the postacute follow-up of distal radius and ankle fractures (ie, after an initial follow-up period of 4 wk) is an example of diagnostic imaging with questionable value. 2,3 Distal radius and ankle fractures are common for all ages. The incidence rate is approximately 70 to 160 per 100 000 persons for distal radius fractures and 187 per 100 000 persons for ankle fractures. [4][5][6][7][8][9] Due to the ageing of the population, incidence rates are expected to increase over the coming decades. 10 Patients with these fractures present a significant burden to the health care system. In order to allow for optimal functional recovery, conservative and operative fracture treatment options aim to optimize and maintain anatomical reduction until fracture healing occurs. [11][12][13] Radiographs are used to monitor the position of the fracture fragments or the osteosynthesis material, the alignment of the joint, and the bone-healing process during the initial phase of follow-up (ie, the first 3 mo). Additional reasons for the use of radiographs include reassurance of the physician and/or patient and medico-legal motives. 14 The frequency and timing of routine radiographs are empirically based. National and international protocols recommend two to four radiographs during the initial phase of follow-up. Typical moments for radiographs in both ankle and distal radius fracture treatment are 1, 2, 6, and 12 weeks after trauma or operative fixation. [15][16][17][18] Studies that evaluated the value of postsplinting radiographs and radiographs taken at the first post-operative outpatient clinic visit after a distal radius fracture suggest that these radiographs do not lead to changes in treatment strategies if they were ordered without a clear clinical indication. 14,19,20 A prospective randomized controlled trial (RCT)the WARRIOR-trial-is currently conducted to confirm the safety and cost-effectiveness of omitting routine follow-up radiographs at 6 and 12 weeks among patients with distal radius or ankle fractures. 21 If the WARRIOR-trial confirms that omitting ("de-implementing") follow-up radiographs without a clear clinical indication is safe and cost-effective, this may lay a foundation for a change in the radiographic follow-up of wrist and ankle fractures. Radiographs taken without a clear clinical indication can then be added to the list of low-value diagnostic tests that can be consulted at the Choosing Wisely website (www. choosingwisely.org).
When performing a clinical trial, research on how to implement possible findings is an important step. Previous studies have shown that solely publishing trial results that demonstrate the redundancy of a certain treatment or test, or simply publishing Choosing Wisely recommendations, did not typically lead to an abandonment of low-value care. [22][23][24] To actually change practice, a strategy is needed to address barriers and facilitators to the change. 25,26 Currently, detailed insight in barriers and facilitators influencing orthopaedic trauma surgeons to adopt a suggested change in follow-up protocol of distal radius and ankle fractures is lacking. Therefore, this study aims to identify the specific barriers and facilitators among orthopaedic trauma surgeons for reducing the use of routine radiographs in the follow-up of distal radius and ankle fractures. We achieved to identify several independently associated facilitators influencing the reduction of routine radiography in the follow-up of distal radius and ankle fractures.

| Study design
In this cross-sectional survey, orthopaedic trauma surgeons in the Netherlands were invited to complete an Internet-based questionnaire. The Medical Ethics Committee of the Leiden University Medical Center approved the study (protocol number P14.214).

| Questionnaire development
To explore potential barriers and facilitators for the de-implementation of routine radiographs during follow-up of distal radius and ankle fractures, semistructured interviews were performed with 10 health care professionals (orthopaedic trauma surgeons) and with six patients (three with a distal radius fracture and three with an ankle fracture). Purposive sampling of health care professionals was applied to obtain contrasting views and identify all potentially relevant barriers and facilitators. To increase generalizability orthopaedic trauma surgeons from different regions in the Netherlands, working in university and nonuniversity hospitals was selected. For practical reasons, only patients treated at a single university hospital were asked to participate. They were contacted during their first outpatient clinic visit, or by phone in the week following their first visit.
The frameworks of Grol and Wensing 27 and Cabana 28 were used to compose the questions of the semistructured interviews. In both frameworks, barriers and facilitators for behaviour change are grouped in several domains (ie, the innovation itself, the individual professional, the patient, the social context, the organizational context, and the economic and political context). In addition to the barriers and facilitators, the professionals were asked about their current follow-up protocol for distal radius and ankle fractures. This was done because current usage of radiography might influence the willingness to adopt a protocol using less routine radiography. Both the professionals and the patients were asked for their opinion about a protocol prescribing radiographs at 6 and 12 weeks only on clinical indication.
The interviews were audiotaped, transcribed, and saved anonymously.  In a pilot, two local orthopaedic trauma surgeons filled out the questionnaire to test the comprehensibility of the questions and the response categories. No changes to the initial questionnaire were deemed necessary after this assessment.

| Population
The developed Internet-based questionnaire was sent to all surgeons registered with the Dutch Trauma Association (n = 236). Nonresponders received four reminders at 1, 3, 4, and 5 weeks after the first invitation.

| Statistical analysis
Data from all respondents who completed the survey were included in the analyses. Descriptive statistics were used to describe baseline characteristics of the respondents and to report the answers to the barrier and facilitator items of the questionnaire.
The data from the questions concerning barriers and facilitators were dichotomized into "disagree" (grouping the answering categories "totally disagree" and "partly disagree") and "agree" (grouping "totally agree" and "partly agree") because of little observations in some cells.
The values of some baseline characteristics were also dichotomized: The number of years of work experience was dichotomized into "0-10 years" and ">10 years," and the annual number of treated patients was dichotomized into "0-50 patients" and ">50 patients." Two groups of respondents were defined: the surgeons who indicated that they intended to stop performing routine radiographic imaging at 6 and 12 weeks after trauma or operative fixation if proven not clinically effective (hereafter referred to as the "intend-to-stop" group) and the surgeons who indicated that they did not intended to do so (hereafter referred to as the "intend-to-continue" group).
The background characteristics, current usage of radiography, and response to each barrier and facilitator were compared between the intend-to-stop and intend to continue groups. Differences between groups were tested with chi-square. The Fisher exact test was used when the number of observations in a cell was less than 6. These analyses were stratified for distal radius and ankle fractures.
Barriers and facilitators with a statistically significant difference between groups (P < 0.05) were considered as potential predictors.
Next, as individual barriers and facilitators may be related to others, we included all potential predictors in a multivariate logistic regression model (P < 0.05), using a backward stepwise, likelihood ratio method.
The intention to stop performing routine radiographs was analysed as the dependent variable, and the barriers and facilitators were analysed as the independent variables. All analyses were performed using IBM SPSS Statistics 23 (IBM Corp, Armonk, NY).

| Respondent characteristics
Of the e-mail invitations sent to 236 Dutch orthopaedic trauma surgeons, seven failed to be delivered and one surgeon indicated that he or she did not work as an orthopaedic trauma surgeon anymore, resulting in 228 invitations. The questionnaire was completed by 130 orthopaedic trauma surgeons (response rate 57%). The reason for nonresponse was not verified. Table 1 shows the baseline characteristics of the respondents. The vast majority (95%) were male, and the mean age was 48 years; 55% of the respondents treat over 50 distal radius fractures, and 34% treat over 50 ankle fractures annually. There were no differences in baseline characteristics between the intend-to-stop and intend-to-continue groups (data not shown).
In total, 71% of the orthopaedic trauma surgeons had the intention to stop taking radiographs routinely for both distal radius fractures and ankle fractures if these radiographs were proven not to be effective in the Warrior-trial (Table 2).

| Current radiographic follow-up
The current radiographic follow-up strategy used by the responding orthopaedic trauma surgeons for conservatively and operatively treated distal radius fractures is depicted in Figure 1. Results are reported separately for both groups. The current follow-up strategy for ankle fractures is highlighted in the same manner in Figure 2. Overall, the majority of respondents indicated to order radiographs after approximately 6 weeks for both conservatively (75%) and operatively treated distal radius fractures (84%), as well as for both conservatively (81%) and operatively treated (88%) treated ankle fractures. Respondents from the intend-to-stop group were significantly less likely to obtain radiographs as a routine part of their current practice. For distal radius fractures, less radiographs were obtained at 6 weeks when treated conservatively (71% vs 95%, P < 0.05). For operatively treated distal radius fractures, less radiographs were obtained at week 2 (16% vs 45%, P < 0.05) and week 6 (81% vs 100%, P < 0.05). For operatively treated ankle fractures, less radiographs were ordered at week 2 (16% vs 32%, P < 0.05) and week 12 (27% vs 47%, P < 0.05). At other time points, there were no differences between groups. Table 3 shows the barriers and facilitators in the questionnaire for each domain of the framework according to Grol         patients: The notion that reducing the number of radiographs in the follow-up of distal radius and ankle fracture leads to cost savings for the health care system, and the need of incorporation of the trial's findings in the regional protocol. A future de-implementation strategy, assuming that the WARRIOR-trial will provide evidence for the reduction of the number of routine radiographs without compromising the quality of care, should focus on changing the current protocols into protocols with fewer radiographs on a regional level. Besides that, a thorough cost-effectiveness analysis needs to be performed, in order to confirm the assumption that implementation of such a protocol will lead to a reduction in cost.

| The influence of barriers and facilitators
To our knowledge, no previous studies on barriers and facilitators for de-implementation of routine radiographs have been conducted.

Voorn et al assessed barriers among orthopaedic surgeons and
anaesthesiologists for the intention to stop the use of erythropoietin (EPO) and blood salvage in total hip and total knee arthroplasty. 24 They found that the intention to stop EPO and blood salvage was

| Strengths and limitations
By conducting semistructured interviews, a complete set of barriers and facilitators based on an established framework was provided for the survey, which can be seen as one of the strengths of this study.
While orthopaedic trauma surgeons with an interest in development or revision of protocols would have been more likely to participate, it is questionable whether their responses would be any different than those of surgeons who do not have an interest in this area. With a response rate of 57%, which is much higher than the responses found in other surveys among surgeons, 24,30,31 the chance of response bias is moderate. Additionally, the number of respondents was large (n = 130), further reducing the risk of response bias. The facilitators that were independently associated with the intention to stop performing the routine radiographic imaging are likely to be relevant to convince surgeons to stop performing routine radiographs.

| CONCLUSIONS
Identifying barriers and facilitators among orthopaedic trauma surgeons regarding the use of a protocol with fewer radiographs is crucial for successful de-implementation of routine radiography for distal radius and ankle fractures. The majority of orthopaedic trauma surgeons intend to follow newly published evidence on the reduced use of routine radiographs. When comparing the intend-to-stop and intend-to-continue groups, several independently associated facilitators can be identified. The identified facilitators can be of value for the development of a tailored de-implementation strategy. In this particular case, the strategy should focus on adjusting the current regional protocols into protocols with less routine radiographs and local, regional, and national education. This education should target the potential benefits of the implementation of these protocols in the terms of cost savings and time efficiency. The education on these protocols will also create familiarity with the study outcomes, and a higher awareness among orthopaedic trauma surgeons.