"A tool to support, not replace": patient and general practitioner perceptions of digital decision support tools for back pain.

We conducted a qualitative study using focus groups [ 13 ]. The protocol was prospectively registered on the Open Science Framework ( https://doi.org/10.17605/OSF.IO/M5TFJ ). The study followed the Declaration of Helsinki, received approval from the University of New South Wales Research Ethics Committee (iRECS6241), and was reported in line with the Consolidated Criteria for Reporting Qualitative Research (see Supplementary Data ) [ 14 ]. The authors include a medical doctor (O.B.), a physiotherapist (R.R.N.R.), a psychologist (J.H.M.), computer scientists (G.B. and T.X.), exercise physiologists (A.G., I.M., M.R., and A.C.), and people with lived experience of back pain (J.M. and S.M.). The study was oriented by relativism (i.e. the belief that reality is subjective) and interpretive orientations (i.e. acknowledging that contextual interactions construct knowledge) [ 15 , 16 ]. R.R.N.R., a male physiotherapist-researcher with experience in qualitative studies and A.G., a male honors exercise physiologist student, moderated the focus groups. We included individuals with back pain and Australian-accredited GPs. The eligibility criteria and recruitment strategy are shown in Table 1 . Purposive sampling ensured diversity in age and gender. The study was advertised online, and interested participants accessed study information, completed eligibility and consent forms via REDCap, and were contacted by a researcher (A.G.) to schedule focus groups. Participation was reimbursed with AUS$40 gift cards for individuals with back pain and AUS$180 for GPs. Eligibility criteria and recruitment strategy. Adults (18 years or older) experiencing back pain of any duration Participants with low back pain living in Australia Participants who understand and speak English Access the internet and a device that can operate Microsoft Teams Advertised via social media Emailed existing lists of individuals who had expressed interest in participating in future studies at Neuroscience Research Australia (NeuRA) Working as an accredited general practitioner in Australia Having provided care to at least five patients with low back pain in the past year Access the internet and a device that can operate Microsoft Teams External recruitment company (TKW) using general practitioners’ networks. Snowballing approach (recommendations and referrals from participants or professional networks) The research team determined the number of participants when a sufficient understanding of the phenomenon to inform clinical practice was achieved through data collection, analysis, and discussions [ 17 ]. A.G and R.R.N.R completed data collection and analysis concurrently. Based on relevant published research, a minimum of 10 individuals with back pain and 10 GPs were planned to be included [ 18 ]. We collected qualitative data through one-hour online focus groups on Microsoft Teams, held separately with individuals with back pain and GPs (3–6 participants per group). A.G. and R.R.N.R., with revisions from co-authors (including researchers, clinicians, and individuals with lived experience), developed the interview guides based on the Theoretical Domains Framework [ 19 ] (see Supplementary Data ). We defined digital decision support tools or decision aids as any digital material or tool accessed by patients or health professionals (e.g. smartphones, email, computers) to support decision-making. Examples might include fact sheets, webpages, search engines, and conversational agents that patients or health professionals can access at any point in the consultation (before, during, or after) to expand on information and facilitate decision-making for back pain management [ 20 ]. A.G. and R.R.N.R. conducted the focus groups. A.G. introduced the study, the participants (i.e. their names, locations, and roles in the focus group), and provided the ground rules for the focus group, then initiated the discussion with guiding questions. R.R.N.R provided technological support, took notes, and conducted complementary queries and summaries. We summarized the demographic data descriptively. We analyzed the qualitative data inductively using Reflexive Thematic Analysis [ 21 , 22 ]. Two analysts (A.G. and R.R.N.R.) first coded data from one focus group with individuals with back pain and one with GPs independently in NVivo 14, then compared and discussed the codes to ensure consistency. A.G. coded the remaining transcripts and developed preliminary themes, which were refined through discussions with co-authors (R.R.N.R., I.M., M.R., and AC). All authors, including those with lived experience of chronic pain (S.M. and J.M.) and a medical doctor (O.B.), reviewed and refined the themes to ensure a transparent and coherent presentation.

Overall, 50 individuals with back pain expressed interest in the study. Nineteen did not provide consent, and 18 were unable to participate due to timing conflicts or an inability to contact them. Thirteen individuals with back pain were included. We also recruited 10 GPs—one via snowball sampling and nine through an external recruitment company. Focus group moderators had no prior contact with participants, except for one GP who had participated in a previous study. We conducted a total of five focus groups: three with individuals with back pain and two with GPs. The focus groups were conducted with three to five individuals with back pain, or four to six GPs each. Most were female ( n = 8/13, 61%) and identified as White ( n = 11/13, 85%). The median (range) age was 44 (24–81) years, with a median (range) duration of back pain of 7 (0.2–50) years. Participants reported moderate to severe pain, with a median (range) score of 6 (3–10) on a 0–10 numerical rating scale. GPs were female (5/10, 50%), with a median (range) age of 51 (31–69) years and a median (range) clinical experience of 23.5 (6–31) years. The ethnicity of GPs was predominantly Asian (7/10, 70%). Most GPs were based in NSW (8/10, 80%), with only one (10%) working in rural communities. The demographic and clinical characteristics of participants are presented in Table 2 . Participants characteristics. Four themes emerged among individuals with back pain, and three themes among GPs. The themes and supporting quotes are summarized in Tables 3 and Table 4 . Themes and sub-themes for individuals with back pain. Themes and sub-themes for GPs. Individuals with back pain reported limited experience using digital decision support tools in general practice. Participants commonly used Google and webpages primarily to gain information about diagnosis and treatments for back pain. Those tools were used independently of GPs, as reported by a 24-year-old man (Q1). Two older participants in the study reported that they had not previously used digital materials to support their decision about back pain management (Q2). A 24-year-old participant explained that digital decision support tools are rarely used because few are available for back pain in general practice (Q3). However, most participants expressed a desire to use digital materials to support back pain management, primarily driven by unmet needs in the current practice (Q4). Participants noted that digital technologies could be beneficial before and after consultations to support discussions and decision-making (Q5). An 81-year-old male with a 50-year history of back pain suggested that educational materials could address persistent post-consultation questions (Q6). A 34-year-old female highlighted the value of complementary online searches for prevention and symptom monitoring (Q7). Similarly, a 32-year-old female noted that GPs could use credible online resources to update knowledge and improve care (Q8). Most participants were concerned about the trustworthiness of online resources and technologies in supporting health decisions, including their reliability, accuracy, and the credibility of the information sources, especially when used without a GP's endorsement. An 81-year-old male was skeptical about whether digital tools, such as Google, can reliably answer and explain health-related questions (Q9). The same participant recognized that his skepticism might be associated with his age and limited familiarity with digital technologies (Q10). A few participants questioned whether digital resources could personalize information for patients (Q11) and expressed concerns that the potential costs of digital technologies might limit access to such tools (Q12). The final challenge raised was participants’ discussions about GPs relying too heavily on digital technologies to make decisions for their patients. An 81-year-old male expressed that he would lose confidence in his GP if he were to use digital decision support tools during the consultation to check treatment options (Q13). However, GPs’ use of digital technologies was perceived as more trustworthy when it supported the interpretation of findings, confirmed results, and clarified information. A 54-year-old female stated that she would be comfortable if GPs used support tools to help interpret and verify or reinforce information during the consultation (Q14). Participants discussed the importance of incorporating digital materials to support decisions for back pain management into routine practice. A 54-year-old woman noted that embedding decision tools into routine care requires patients to possess digital and scientific literacy; without this, such tools may go unused or be used ineffectively (Q15). A 32-year-old woman suggested that providing details on pharmacological, surgical, and non-pharmacological options could motivate patients with back pain to continue using digital tools (Q16). Some participants indicated that integrating digital tools into consultations could help GPs organize and simplify treatment plans. For example, a 44-year-old woman noted that such tools could support self-management rather than temporary or symptom-based solutions (Q17). GPs reported experience with non-digital decision support tools, such as printed education materials (i.e. fact sheets), to support patients in following treatment and lifestyle recommendations. They were largely unaware of specific digital decision-support resources available to help GPs and patients manage back pain in general practice (Q18). Most GPs considered that digital tools would be valuable for pre-consultation screening (Q19). A 31-year-old male GP noted that digital tools could assist patients in deciding whether to seek care and in preparing for consultations through enhanced health literacy (Q20). A 59-year-old male GP similarly noted that digital tools could help GPs screen for red flags before consultations, freeing up time for meaningful discussions with patients that commonly are not possible due to time constraints (Q21). However, a 54-year-old female GP stressed that not all patients would be willing to complete forms before the consultation (Q22). Many GPs expressed concerns about the impact of digital decision support tools on consultations and clinical interactions in the management of back pain. A few GPs felt uncomfortable with the possibility that patients might use the tools to check whether clinicians had provided the right approach. For example, a 43-year-old male GP mentioned that inconsistencies between the tool and clinicians’ recommendations could undermine the therapeutic relationship (Q23). A 51-year-old female GP expressed concern about the independent use of digital decision-support tools, suggesting that this could undermine GPs’ professional role (Q24). A few GPs reported not requiring decision support tools for managing back pain, citing confidence in their existing knowledge and clinical skills (Q25, Q26). Others believed that current technology is not ready to address complex cases, particularly when patients present with multimorbidity or back pain due to uncommon causes (Q27). These concerns seemed to diminish when decision support tools were designed for patient use under GP supervision (Q28) and when the information provided was transparent, evidence-based, and trustworthy (Q29). GPs discussed the importance of embedding digital decision-making tools within routine care to support effective implementation. Overall, GPs agreed that decision support tools should be seamlessly integrated into the existing general practice workflow and not be intended to replace clinical interaction (Q30). A few GPs suggested that digital tools designed to facilitate patients’ and GPs’ decisions should be integrated into existing practice management systems (e.g. pre-appointment platforms that already capture patient information) to support implementation in clinical settings (Q31/Q32). They further highlighted that these tools should be user-friendly, efficient, and add value to general practice—for example, by monitoring symptoms or reducing administrative workload (Q33). Finally, they emphasized the importance of informing the public on how to use these tools (Q34).

Back pain constitutes a substantial burden for individuals and society, and for the past three decades, it has remained the leading cause of years lived with disability worldwide [ 1 ]. The condition generates significant healthcare costs, with expenses for consultations, diagnostic procedures, treatments, and hospital admissions estimated at €2.3 to €2.6 billion annually [ 2 ]. Clinical practice guidelines offer recommendations for managing back pain in general practice [ 3 ]. The current recommendation involves assessing patients, screening for specific causes, providing referral when needed, and, for most individuals, offering advice and information, selecting appropriate non-drug treatments, and, for some , prescribing drug treatments for acute or chronic back pain [ 4 ]. Treatment decisions should reflect shared discussions that consider the evidence, patient preferences, and contextual factors [ 5 ]. However, implementing these recommendations remains challenging [ 6 ]. Patients often arrive with misconceptions about back pain, multisite pain, and comorbidities [ 7 ], which can be difficult to address within a single consultation [ 8 ], and within the constraints of limited consultation time [ 9 , 10 ]. The use of decision-support tools, accessed by patients or health professionals via smartphones, email, or computers, has been proposed to support evidence-based care [ 11 , 12 ]. Examples include digital fact sheets, webpages, search engines (e.g. Google), and conversational agents that patients can access themselves or that are directed by health professionals. These tools might inform management and provide education, offering convenient, fast access to evidence-based information in clinical practice. Although the use and delivery of digital materials in the general practice context are growing, the perspectives of individuals with back pain and health professionals on digital decision aids for back pain management are unknown. To address this, this study aims to understand the perspectives of individuals with back pain and GPs on digital decision support tools for back pain management.

Our study combined GP and patient perspectives to generate insights for implementing digital decision-support tools in primary care. We highlight their potential to support back pain management and the need to consider context, patient, and GP views for successful implementation.

This study explored the perceptions of individuals with back pain and GPs regarding digital materials or tools accessed by patients or health professionals at any point in the consultation to support decision-making. Separate themes revealed complementary views to inform future implementation, with both groups often relying on generic tools. The discussion was structured to facilitate the research and implementation of digital materials across three stages of the clinical practice: before, during, and after the consultation. Individuals with back pain reported using search engines such as Google to understand the cause of their back pain and gain information about treatment options before and after the consultation. However, most of them recognize uncertainty regarding the trustworthiness of information sources, which, according to a systematic review, more than half of websites provide inaccurate or unclear treatment recommendations [ 23 ]. This suggests that, despite their desire for reliable information about treatment options, patients often lack access and therefore turn to untrustworthy digital resources. Access to trustworthy digital materials could be encouraged by clinicians, practices, and health insurance providers. In this study, we found that GPs, and especially individuals with back pain, believe that pre-consultation decision-support tools could increase patients’ knowledge and enable productive discussions. However, low evidence-based literacy and maladaptive pain beliefs may hinder patients’ use of digital tools and affect decision-making. These barriers can limit their understanding of treatment options and reduce their ability to engage in collaborative discussions with clinicians [ 24 ]. For some patients and professionals, education materials and time for assimilation may be required to support implementation. Another facilitator is the integration of education materials and tools into primary care workflows, as consistently reported by both patients and GPs. For example, digital pre-consultation tools could provide an opportunity to educate and prepare patients for the consultation [ 25 ]. A recent randomized controlled trial demonstrated that a pre-appointment form, which facilitated the delivery of information about back pain management, improved patients’ perceptions of how well they were prepared to engage in discussions with their GPs and to make informed health decisions [ 26 ]. GPs indicated that tools to screen for red flags before the consultation would be valuable, as they could help validate clinical decisions and allow more time during the consultation to address patients’ questions and provide education. For instance, Traeger et al. [ 11 ] developed a tool to estimate the risk of chronic back pain (PICKUP; https://myback.neura.edu.au/ ), which incorporates red flags as indicators of prompt GP visits. Although the prognostic model requires further external validation and refinement, it has been suggested that it could prevent up to 40 unnecessary GP visits per 100 patients presenting to primary care [ 11 ] and could be easily implemented in pre-consultation tools for back pain. Advancements in digital technologies, such as artificial intelligence (AI), may facilitate the identification of urgent cases and enable timely advice to patients before or during a general practice visit [ 27 ]. For example, large language models can identify and analyze patients’ self-reported symptoms in electronic health records, flag potentially serious conditions, and provide suggestions for further investigations, self-care, or referral [ 28 ]. Nevertheless, testing is needed to evaluate their acceptability among patients and clinicians, the reliability of their outputs, and their feasibility for integration into routine care [ 12 ]. These pre-appointment strategies become especially relevant in the context of prolonged waiting times for accessing health services, when patients are at a greater risk of relying on inaccurate information or missing timely reassurance and evidence-based advice. Interestingly, GPs demonstrated confidence in making treatment decisions for back pain, even though evidence of non-guideline concordant care. A systematic review [ 6 ] found that GPs prescribed second-line interventions, such as NSAIDs, in 35% of acute and chronic back pain, compared with only 20% providing advice to stay active or referring to a physiotherapist, which are considered first-line options [ 29 ]. One possible explanation for this discrepancy may be the entrenched medication culture within general practice, as well as the lack of a standardized system for prescribing non-pharmacological interventions [ 30 ]. Recent initiatives have aimed to promote the prescription of non-drug treatments in general practice by increasing access to national clinical guidelines [ 5 , 31 ] and high-quality systematic reviews and overviews [ 32 ]. For example, the Cochrane Database of Systematic Reviews has recently published a series of overviews on pharmacological, non-pharmacological, and non-surgical interventions for back pain [ 33 , 34 ], which are accessible via the Cochrane Library [ 32 ]. Nonetheless, most evidence-based information resources have yet to be systematically integrated into general practice management systems. In Australia, the e-HANDI project ( https://e-handi.com/ ), an online prescription platform, allows GPs to enter a health condition and generate evidence-based non-pharmacological options that can be emailed directly to patients, including information on benefits, contraindications, adverse effects, and practical details, such as frequency, dose, and community availability. Research on mobile app interventions for back pain is also promising, suggesting that GPs could prescribe evidence-based apps during consultations [ 35 ]. Co-design of future decision-making tools with patients, clinicians, and policymakers might increase acceptability, feasibility, and ultimately implementation. For example, interviews with individuals with back pain and health professionals were used to co-create a decision-support tool for lumbar fusion [ 36 ]. While promising, all these tools need to address challenges related to digital health literacy, age-associated barriers, internet access, information reliability, acceptability, and effectiveness [ 37 ]. Over time, the influence of these barriers may diminish as digital technologies become increasingly accessible through inclusivity initiatives [ 38 ]. Importantly, both patients and clinicians acknowledge the complementary role of digital technologies within GP-led care, emphasizing their potential as a “tool to support, not replace.” A medical doctor (O.B.) is among the authors, and the study was discussed with GPs during the conceptualization. However, no actively practicing GP was involved. A more extensive member-checking process with participants could have further strengthened confidence in the findings. Future studies with more participants living/practicing in rural and remote areas, and culturally and linguistically diverse backgrounds can enhance transferability.