General practitioner experiences using a low back pain management booklet aiming to decrease non-indicated imaging for low back pain

Abstract

Background

Imaging is overused in the management of low back pain, resulting in overdiagnosis, increased healthcare utilisation, and increased costs. Few effective interventions to decrease inappropriate use have been developed and have typically not been developed using behaviour change theory. An intervention to reduce non-indicated imaging for low back pain was developed using behavioural change theory, incorporating a novel low back pain management booklet to facilitate patient education and reassurance. The aim of this study was to assess the adoption and feasibility of use of the developed intervention within clinical practice and to determine appropriate implementation strategies to address identified barriers to use.

Methods

Fourteen general medical practitioners were recruited and trained to use the booklet with low back pain patients over a minimum 5-month period. Quantitative data on use of the booklet were collected and analysed descriptively. Qualitative data on use of the booklet and training session were collected in general medical practitioner interviews and thematically analysed. Barriers to use were identified and mapped to suitable implementation strategies using the Behaviour Change Wheel.

Results

Practitioners used the booklet with 73 patients. The booklet was used with 63% of patients presenting with low back pain. Facilitators for using the booklet included patient’s requesting imaging and lower practitioner confidence in managing low back pain. Barriers included accessible storage and remembering to use the booklet. Implementation strategies were identified to increase adoption and feasibility of use, including development of a digital version of the booklet.

Conclusions

General medical practitioners reported that the low back pain management booklet and training were useful for clinical practice, particularly with patients requesting imaging. Barriers to use were identified and implementation strategies to address these barriers will be incorporated into future effectiveness studies. This study forms one of a series of studies to thoroughly develop and test an intervention to reduce non-indicated imaging for low back pain; a successful intervention would decrease healthcare costs and improve patient management.

Contributions to the literature

• The adoption and feasibility of use of a unique intervention (including a low back pain management booklet) designed to reduce non-indicated imaging for low back pain were assessed.

• General medical practitioners found the booklet most useful when patients were requesting non-indicated imaging or needed more reassurance.

• Key barriers to use included a lack of storage space and remembering to use the booklet.

• Implementation strategies to improve adoption and feasibility of use of the intervention were developed using the Behaviour Change Wheel and the Theoretical Domains Framework, a novel approach in the management of low back pain.

Background

Imaging is overused in the management of low back pain (LBP), with approximately one third of imaging referrals inconsistent with clinical guidelines [1]. Imaging is indicated when there is suspicion of serious underlying pathology such as infection or cancer but does not generally improve outcomes for patients with non-specific LBP [2, 3]. Overuse of imaging may lead to inappropriate diagnoses, further unnecessary investigation or treatment, and unnecessary radiation exposure [2,3,4,5]. Decreasing non-indicated imaging for LBP in general practice is challenging: few effective interventions have been demonstrated to date [6] and few have been developed using behaviour change theory [7].

An intervention was recently developed [8] to help general medical practitioners reduce non-indicated imaging for LBP. The research team identified clinician and patient behaviours within a clinical consult which may lead to an overuse of imaging for LBP and developed an intervention to address these behaviours using the Behaviour Change Wheel [9] and the Theoretical Domains Framework [10] (Additional file 1). The intervention included clinician training and provision of a LBP management booklet designed to be used during clinical interactions. The novel booklet (available online at https://tinyurl.com/lowbackpaineducation [11]) was purpose-designed by the research team to address behaviours leading to overuse of imaging. The booklet can be used to screen the patient for indicators for imaging, educate and reassure the patient about LBP and the need for imaging, and provide a customised patient management plan.

Initial feedback on the booklet was sought from general medical practitioners and health care consumers [8]; however, they did not have any actual experience using the booklet in clinical practice. General feedback on the acceptability and appropriateness of the booklet was positive, although some practitioners raised implementation concerns related to the potential adoption and feasibility of use of the booklet in clinical practice. Before effectiveness studies are undertaken, it is important to assess the use of the booklet in clinical practice to ensure that implementation is achievable.

Therefore, the aim of this study was to assess the adoption and feasibility of use of the developed intervention within clinical practice and to determine appropriate implementation strategies to address identified barriers to use.

Methods

General medical practitioners from metropolitan Sydney, Australia, were asked to use the intervention within their clinical practice. Semi-structured interviews were used to explore practitioners’ experiences using the intervention and alignment with the proposed theoretical model (Additional file 1). Barriers to use were described using the Theoretical Domains Framework. Implementation strategies to address identified barriers were developed using the Behaviour Change Wheel. Ethics approval was granted by Macquarie University Human Research Ethics Committee, reference number: 5201600298. The study and intervention were reported in accordance with the consolidated criteria for reporting qualitative studies (COREQ) and template for intervention description and replication (TIDieR) checklists (Additional files 2 and 3).

Practitioner recruitment

To enable diversity in participant characteristics that may impact use of the intervention (socioeconomic region of practice location, further education or special interest in LBP, years in clinical practice, sex), purposive sampling of general medical practitioners was performed between May to October 2017. Practitioners or practices known to the research team, and those recommended by participating practitioners, were approached. To be eligible, practitioners needed to self-report that they were currently seeing patients with LBP to ensure that practitioners would have sufficient opportunity to use the intervention during the study period. Based on the sample size needed for a related qualitative study on the development of the booklet [8], we estimated a minimum of 10 practitioners of adequate diversity would be required for this study to generate sufficient data to meaningfully explore the research aims [12]. Further practitioners were recruited as required to ensure practitioners reflected the desired diversity.

Study procedure

Practitioners attended a 20-min face-to-face training session with one of the research team (HJ) to instruct them in the study aims and requirements and to deliver the training session developed for the intervention (Additional file 4). Demographic information and beliefs about the usefulness of imaging for LBP were obtained (Additional file 5).

The study period ran from May 2017 to April 2018. In the training session, practitioners were asked to use the booklet with patients presenting with LBP who they considered were not indicated for imaging and complete a de-identified record sheet of all low back patients. Recorded data included if and how the booklet was used, imaging referral, LBP characteristics, and suspicion of underlying pathology (Additional file 6).

At the conclusion of the study period, practitioners participated in a 15-min audio-recorded semi-structured interview with one of the researchers (HJ). Questions related to practitioner behaviour were developed using the theoretical domains framework [13, 14] (Additional file 5). Practitioners were given an AUD$60 gift voucher for their time in attending the training session and participating in the end of study interview.

Quantitative data analysis

Data from the de-identified patient record sheets (Additional file 6) were used to assess how general medical practitioners used the booklet, including the proportion of patients the LBP management booklet was used with; characteristics of patients with whom the booklet was used (e.g. previous history of back pain or imaging); and how the booklet was used with each patient (e.g. within the consult, as a handout).

Qualitative data analysis

Interviews were transcribed by one researcher (HJ) and imported into NVivo qualitative data analysis software, QSR International Pty Ltd. Version 12, 2018 for analysis. Coding was performed for each study aim prior to performing thematic analysis [15]. To assess how the booklet was used in clinical practice, initial coding was aligned with the theoretical model underpinning the development of the intervention (Additional file 1). Aims relating to clinician behaviour were initially coded using the domains outlined in the Theoretical Domains Framework [13]. Coded data were collated based on similarity, leading to the generation of common themes related to each study aim.

Two researchers (HJ and NM), both with prior experience in coding and using the theoretical domains framework, independently coded three interviews. Coding was compared and discussed, and sufficient consistency was observed between the two researchers after two rounds of discussion to allow one researcher (HJ) to code the remaining interviews. Themes were initially developed by HJ, before discussion with MH, NM, and SF to reach consensus. The resultant themes were then sent to all authors for overall discussion and final consensus.

Mapping of implementation strategies to address identified barriers

The Behaviour Change Wheel [16] was used to map the identified barriers to using the booklet to appropriate implementation strategies designed to increase use of the booklet in clinical practice. In this process, integration of the COM-B model (Capability, Opportunity, Motivation – Behaviour) and the 14 behavioural domains in the theoretical domains framework were used to map identified barriers to specific behavioural domains requiring change. Appropriate behavioural change techniques and implementation strategies were selected to address each domain, with techniques/strategies prioritised according to the APEASE criteria (Affordability, Practicability, Effectiveness and cost-effectiveness, Acceptability, Side-effects and safety, Equity) [16] and suggestions from practitioners to improve implementation of the intervention. Proctor’s specifications were used to specifically define, describe, and justify the implementation strategies selected [17]. One researcher (HJ) performed the initial mapping, which was then discussed and finalised with the research team.

Results

Twenty-one general medical practitioners were approached to participate. Of these, four (19%) declined as they either reported that they infrequently saw patients with LBP (N = 3) or did not want to participate (N = 1). Of the 17 practitioners who participated in the study, 14 (82%) completed the interview at the end of the study. Practitioners had on average 16 years clinical experience, tended to have completed continuing education in LBP (64%), and did not agree that imaging is useful for LBP (Table 1).

Table 1 General medical practitioner characteristics (N = 14)

Adoption and use of the intervention by general medical practitioners

Practitioners participated in the study for between five to 11 months (mean, SD: 8.4, 2.2), depending on their date of recruitment into the study. All practitioners attended the training session. Practitioners used the booklet with 73 LBP patients (mean, SD: 5.2, 4.1). Practitioners reported seeing 99 patients with LBP during the study period; however, only seven of the practitioners (50%) reported completing the patient record form for all LBP patients. For these practitioners, they used the booklet on average with 62.5% (SD: 38.2; 95% CI: 27.2–97.8) of patients presenting with LBP, with reported use ranging from 14.3 to 100.0%. The other practitioners only completed the patient record form when they used the booklet, so percentage of use could not be calculated. During the interviews, these practitioners estimated using the booklet, on average, with 13.6% (SD: 13.1; 95% CI: 1.5–25.7) of patients presenting with LBP, with use ranging from 0.0 to 50.0%.

The patient record form was fully completed for 71% of patients (52/73) with whom practitioners used the booklet, with partial data available for the rest. Most patients had previous episodes of LBP (68%), but the current episode duration was less than 2 weeks (58%). Serious pathology was rarely suspected (7%) and 11% were referred for imaging (Table 2). When practitioners used the booklet, they commonly provided the booklet to patients to take home (56/60, 93.3%; 95% CI: 84.1–97.4), with the patient management plan completed as directed in the training session for most of these patients (52/56, 92.9%; 95% CI: 83.0–97.2). For the remaining patients, practitioners discussed the booklet with patients who subsequently declined to take it home (4/60, 6.7%; 95% CI: 2.6–15.9).

Table 2 Characteristics of patients with whom practitioners used the booklet

Themes related to how practitioners used the booklet are presented in Table 3. Practitioners did not always discuss the booklet with patients during the consult if they were running short of time or felt they could adequately reassure and educate the patient without the booklet. However, when appropriate, they provided the booklet for the patient to read at home to present further information or reinforce messages delivered during the visit.

Table 3 Themes related to ‘How general medical practitioners used the booklet’

Most practitioners (11/14) reported that they found the booklet useful and would be likely to continue using it in the future, particularly with specific patients, that is, those that requested imaging or required more reassurance or information about their LBP.

“I genuinely think it’s [the booklet] really useful and I’ll continue to use it” (GP10)

“Not everybody like this [continue to use the booklet moving forward], but those who are not easy to convince so those who need more information about back pain who aren’t aware what’s, yes.” (GP13)

One practitioner did not use the booklet during the study, and two practitioners reported that they would be unlikely to continue to use the booklet. These three practitioners reported that they already felt confident that patients would follow their advice without additional resources.

“I think it [the booklet] would be reassuring for lots of clinicians but for me personally I think I can communicate my confidence to the patient and I might be wrong but I feel they’re OK with me just explaining why they don’t need anything” (GP1)

“I’m pretty confident that I don’t need to do the imaging in the first place, so I don’t know whether it [using the booklet] makes a tremendous difference for me really” (GP7)

Feasibility of using the intervention in clinical practice

Themes relating to barriers and facilitators impacting on general medical practitioners’ use of the LBP management booklet are presented in Table 4. Key barriers included the ability to conveniently store and remember to use the booklet and a lack of time during the consult. A digital version of the booklet was suggested as more convenient to store and remember to use. Facilitators included the ease of use of the booklet and the usefulness of the booklet to help educate and reassure the patient in a time efficient manner. The request for imaging by the patient acted as a reminder to use the booklet.

Table 4 Themes related to ‘Barriers and facilitators impacting use of the intervention’

Mapping of barriers to implementation strategies

The mapping of the identified barriers to implementation strategies is presented in Table 5 with definitions of the implementation strategies outlined in Additional file 7. The key behavioural domains addressed were those of psychological capability and reflective motivation. Additional implementation strategies selected in this process included the following: development of a digital version of the booklet to allow for easy storage, hardcopy booklets available for patients in the reception area, reminders to use the booklet through the practice management software, audit and feedback of imaging referral behaviour to clinicians, and selection of a local opinion leader to champion use of the booklet.

Table 5 Mapping barriers to using the intervention to implementation strategies

Discussion

This study found that general medical practitioners varied in their use of the developed intervention to reduce non-indicated imaging. Low users of the booklet were more likely to be confident in their management of LBP and reported not needing additional resources. Higher use was reported when patients requested non-indicated imaging or needed more reassurance. The booklet was feasible to use in clinical practice; however, important barriers to use were identified, including available storage and remembering to use the booklet. A digital version of the booklet was strongly favoured by all practitioners.

Strengths of this study included the use of both quantitative and qualitative methods to assess the adoption and feasibility of use of the intervention in clinical practice. Quantitative data showed variable use of the booklet by general medical practitioners and qualitative analysis used the Theoretical Domains Framework to identify and explore barriers and facilitators influencing use. Implementation strategies to address identified barriers were selected using the Behaviour Change Wheel and described using Proctor’s specifications.

A limitation of this research was the lack of feedback from patients regarding their experience in receiving the booklet. Future research would benefit from exploring patient feedback to assess how useful they found the booklet. The use of the booklet varied between practitioners and could not be accurately measured due to incomplete data from general medical practitioners; however, qualitative responses allowed us to explore the barriers limiting use of the booklet and address these for future implementation. Importantly, practitioners tended to remember to use the booklet with patients who requested imaging or needed more education or reassurance, thus, using the booklet in the cases where it is needed.

Study generalisability and the relatively small sample size of practitioners needs to be considered as a possible limitation. This was an exploratory study, and it is possible that with broader sampling or longer interviews, additional barriers and subsequent implementation strategies may have been identified. Sampling of practitioners was performed to achieve diversity in socioeconomic region of practice location, further education or special interest in LBP, years in clinical practice, and sex. Diversity was not achieved in beliefs about the need for imaging, with all practitioners reporting that imaging is not typically useful in the management of acute LBP. However, previous research has shown that whilst medical practitioners commonly disagree about imaging being important for LBP management [18], they still frequently order imaging. Barriers such as patient pressure for imaging and limited time in a consult (which this intervention was designed to directly address [8]) are thought to be the main drivers of imaging overuse.

The results of this study will be used to further inform the development and implementation of an intervention to reduce non-indicated imaging for LBP in general medical practice. The identified implementation strategies to increase intervention use will be incorporated into the planned studies to assess the effectiveness of the intervention in clinical practice.

Conclusion

General medical practitioners had variable adoption of a LBP management booklet in clinical practice. Low use was more common in practitioners who were confident in their ability to educate and reassure patients with LBP. Practitioners were more likely to use the booklet if patients requested imaging or required more reassurance about their LBP. Barriers impacting the use of the intervention were identified and strategies to increase use will be incorporated into future implementation measures. This study is one part of a series designed to develop and test an intervention to reduce non-indicated imaging for LBP; a successful intervention would decrease healthcare costs and improve patient management.