Impact of direct access on the quality of primary care musculoskeletal physiotherapy: a scoping review from a patient, provider, and societal perspective

Introduction

Musculoskeletal (MSK) disorders are among the top ten leading causes of years lived with disability (YLD) (1,2). Mounting evidence suggests that the quality of care offered in primary care settings treating these MSK disorders is suboptimal and that often inaccurate diagnoses are made by primary care physicians (3,4). Physiotherapy is a frequently recommended treatment option for the management of MSK disorders (5). Responses to the latest World Confederation for Physical Therapy survey reveal direct access (DA) is available in 48 countries and there is no restriction on private practice in 77 countries (6). In countries where physiotherapists receive professional autonomy through DA, evidence suggests several benefits including a more valid diagnosis as compared to primary care physicians, better outcomes for patients, and more efficient use of resources, while maintaining high patient satisfaction (5,7).

Further benefits can be linked to DA physiotherapy, such as shorter waiting times, reduced health care costs including physician fees, medical imaging expenses, and medication costs (5,8), increased prestige for physiotherapists (9,10), and decreased workload for primary care physicians (10). However, cost reduction may be restricted to direct costs and general workload for the Physical Therapist (PT) may not necessarily be reduced (11). Also, potential disadvantages to this model of health care have been described, for instance, potential erosion of a strong patient-doctor relationship (12,13) or a robust physiotherapy-doctor connection (13), as well as concerns about overconsumption of physiotherapy services (14).

This scoping review aimed to identify, appraise, and synthesize existing literature to assess the impact of DA on primary care physiotherapy for patients presenting with various MSK disorders. The impact of DA will focus on outcomes from the perspectives of the patient, the provider, and society.

Methods

The reporting of this scoping review conforms to the PRISMA Extension for Scoping Reviews (PRISMA-ScR) guidelines (15).

MEDLINE (PubMed) and Web of Science were searched from 1990 until March 2024. The electronic search strategy used in these searches is listed in Table 1.

Titles and abstracts were independently reviewed by two reviewers up to March 2024, applying the following inclusion criteria: availability of quantitative data of at least one group that received physiotherapy through DA or direct allocation without consulting a physician and assessment of at least one of the perspectives for the patient (quality of life [QoL], well-being, satisfaction, pain, functioning, or adverse events), physiotherapists (treatment compliance, responsibility, liability, status, prestige, or job satisfaction), and society (number of referrals with and without a DA setting, amount of medical imaging, medication use, number of sessions needed for rehabilitation, and overall costs and cost-effectiveness). Articles written in English, Dutch, or French were considered. Papers not complying with the inclusion criteria were excluded. Randomized clinical trials were selected and studies retrieved by the above search string which were published after the latest systematic review were added to this scoping review.

The Risk of Bias in Systematic Reviews (ROBIS) tool was developed specifically to be used by guideline developers, authors of overviews of systematic reviews (“reviews of reviews”), and review authors who might want to assess or avoid risk of bias in their reviews (16). The ROBIS tool was utilized by two reviewers independently to assess the risk of bias in the included systematic reviews. Discrepancies between the two reviewers were resolved by discussion and if disagreement persisted, a third reviewer made the final decision.

Results

The flow of studies through the review is presented in Fig. 1.

Patient perspective

The low risk of bias review by Babatunde et al (17) found no significant differences in pain reduction between DA physiotherapy and care supervised by a general practitioner (GP). Similar findings were reported by Piscitelli et al (18) and Demont et al (19), although small differences in favour of DA physiotherapy were noticed (p = 0.76) (18). Ojha et al (20) reported a significant but small result for pain reduction in favour of DA physiotherapy (p = 0.011).

Babatunde et al (17) reported no significant results for QoL and function while Ojha et al (20) and Piscitelli et al (18) showed better outcomes in terms of QoL and function in favour of DA physiotherapy (p = 0.04 and p = 0.03, respectively). According to Ojha et al (20) and Demont et al (19), patients reported 5.0-21.5% greater satisfaction in the DA groups than in physician referral groups (p < 0.01). Additionally, Gallotti et al (21) reported equal to superior QoL and a tendency to higher patient satisfaction in DA groups.

Societal perspective

Piscitelli et al (18) and Demont et al (19) concur that DA physiotherapy can reduce the waiting time for primary consults by 4 to 63 days (p < 0.001). Similarly, Gallotti et al (21) reported shorter waiting times as well as improved management accuracy regarding the type of access to PT (i.e. by GP referral, access by consultant, or DA physiotherapy).

Four out of five systematic reviews showed consistent results regarding the amount of prescribed medication and medication use (17-20). DA physiotherapy led to 11.9-65.0% less prescribed medication (p < 0.01) and reduced pharmacological costs by $42-710 (p < 0.01) (17,20).

Ojha et al (20) reported significantly fewer physiotherapy visits in a DA setting, with a range of 1.1-13.4 visits (p < 0.01). Demont et al (19) found no consensus about the number of physiotherapy visits, with either two to three fewer physiotherapy visits needed in a DA setting (p = 0.001) or no significant difference found. Babatunde et al (17) and Piscitelli et al (18) reported that DA physiotherapy led to 2.0-21.5% fewer follow-up visits with the primary care physician (p < 0.05). Gallotti et al. (21) indicate a shorter time to discharge in DA. Demont et al (19) reported that 17% fewer patients required a primary care physician visit in a DA physiotherapy setting (p = 0.0113).

Ojha et al 2014 (n = 8)	Piscitelli et al 2018 (n = 12)	Demont et al 2019 (n = 18)	Babatunde et al 2020 (n = 26)	Gallotti et al 2023 (n = 28)
Hackett et al 1993	Gentle et al 1984	Overman et al 1988	Greenfield et al 1975	Daker-White et al 1999
Mitchell et al 1997	Hackett et al 1993	Holdsworth et al 2004	Mitchell et al 1997	Oldmeadow et al 2007
Holdsworth et al 2004	Mitchell et al 1997	Moore et al 2005	Overman et al 1988	Sephton et al 2010
Moore et al 2005	Holdsworth et al 2004	Holdsworth et al 2006	Ferguson et al 1999	Ludvigsson et al 2012
Holdsworth et al 2007	Holdsworth et al 2006	Holdsworth et al 2007	Moore et al 2005	Phillips et al 2012
Webster et al 2008	Holdsworth et al 2007	Brooks et al 2008	Holdsworth 2007/2008	Kooijman et al 2013
Leemrijse et al 2008	Leemrijse et al 2008	Leemrijse et al 2008	Bossonnaulth et al 2010	Salisbury et al 2013
Pendergast et al 2012	Brooks et al 2008	Webster et al 2008	Pendergast et al 2012	Mallett et al 2014
	Webster et al 2008	Ludvigsson et al 2012	Phillips et al 2012	O’farrell et al 2014
	Ludvigsson et al 2012	Pendergast et al 2012	Chetty et al 2012	Samsson et al 2014
	Pendergast et al 2012	Mallett et al 2014	Ludvigsson et al 2012	Bornhöft et al 2015
	Badke et al 2014	Swinkels et al 2014	McCallum et al 2012	Samsson et al 2015
	Bishop et al 2017	Bomhöft et al 2015	McGill et al 2013	Bird et al 2016
		Mintken et al 2015	Badke et al 2014	Kerridge-Weeks et al 2016
		Goodwin et al 2016	Mallett et al 2014	Samsson et al 2016
		Bishop et al 2017	Swinkels et al 2014	Bishop et al 2017
		Bomhöft et al 2019	Bornhöft et al 2015	Chang et al 2018
		Downie et al 2019	Mintken et al 2015	Bornhöft et al 2019
			Ojha et al 2015	Caffrey et al 2019
			Boissonnaulth et al 2016	Downie et al 2019
			Goodwin et al 2016	Lankhorst et al 2020
			Harland et al 2016	Ojha et al 2020
			Pearson et al 2016	Oostendorp et al 2020
			Bishop et al 2017	Peterson et al 2021
			Mant et al 2017	Ho-Henrikson et al 2022
			Denninger et al 2018	Lyons et al 2022
				Szymanek et al 2022

Shaded sections refer to primary studies that have been analysed also in previous systematic reviews.

Four systematic reviews concluded that DA physiotherapy could lead to 6.3-70.0% fewer X-rays and other medical imaging (p < 0.001) (17,18, 20,21). Babatunde et al (17) and Piscitelli et al (18) also showed lower overall health costs (p < 0.01) up to 20%. Ojha et al (20) and Demont et al (19) showed decreased costs in a DA physiotherapy setting compared to a GP referral setting (p < 0.05). This was further supported by Gallotti et al (21).

Babatunde et al (17) and Gallotti et al (21) reported less work-related absence and sick leave in DA physiotherapy. Ojha et al (20) reported an average of 17.4 days less work absence in a DA physiotherapy. Piscitelli et al (18) did not find a consensus for the return-to-work rate. They found either no difference in return-to-work rate or 14.1% less lost time from work and daily duties (p < 0.05).

First author (year of publication)	Patient perspective	Physiotherapist perspective	Societal perspective
Ojha et al (2014) (20)	Significantly less average pain was reported in DA. (VAS decreased from 5.7 to 2.7 vs. from 5.7 before treatment to 3.2 after treatment; p = 0.011.) Higher improvement in function in DA (79% vs. 60%; p = 0.04). Higher levels of satisfaction in DA (79%-93.2% vs. 74%-84.1%; p < 0.01). DA PT is not related to any adverse events (p>0.05).	/	Lower costs for radiological examinations in DA PT (5.1%-13.6%; p = NR)/(4.00 vs. 7.43; p < 0.01). Significantly fewer medication prescribed in DA PT (32.2%-48% vs. 44.1%-84%; p < 0.01). Significantly less PT visits in DA PT (5.9-20.2 vs. 7.0-33.6; p < 0.01). Decreased cost in DA (£9.55-$14.83 vs. £47.94-$63.65; p < 0.01). Mean number of days of work missed was significantly less in DA PT (10.4 vs. 27.9 days: p = NR).
Piscitelli et al (2018) (18)	No significant difference between GP referral and DA PT in terms of pain reduction (64.6% vs. 66.6%; p = 0.76). DA PT has better clinical outcomes in terms of function and health-related QoL (self-reported questionnaires: 2.4 ± 2.8 vs. 4.1 ± 4.6; p = 0.03). DA PT is not related to any adverse events (p>0.05).	DA PT provides greater treatment compliance (76%-79% vs. 58%-69%; p = 0.004).	DA PT reduces waiting time and improves accessibility of PT (5-6 vs. 9-69 days; p < 0.001). DA PT leads to less medical imaging ($44 ± 190 vs. $175 ± 541; p < 0.01)/(47 vs. 242 patients; p < 0.001). DA PT leads to less prescribed medication ($36-163 vs. $78-873; p < 0.01)/(62-79 vs. 219-276 patients; p < 0.001).
Demont et al (2021) (19)	No significant difference between DA and physician-led usual medical care for pain but tendencies to positive effects (p≥0.05). Better clinical outcomes for function in DA PT (RMQ function score 4.1 in GP referral vs. 2.4 in DA PT; p = 0.03). No significant difference in QoL, but tendencies to positive effects (p≥0.05). Significantly higher satisfaction levels in DA compared to physician-led usual medical care (74.7% vs. 53.2%; p = 0.002). DA PT is not related to any adverse events (p>0.05).	Significantly higher treatment compliance in DA (76%-79% vs. 58%-69%; p = 0.004).	Significantly shorter waiting time with DA PT (4-5 vs. 9-31 days; p < 0.001). Significantly less X-ray imaging described in DA PT (7.3% vs. 13.6%; p < 0.001). Significantly less medication prescribed in DA PT (32.2% vs. 44.1%; p < 0.01). Fewer or the same number of PT visits in DA PT (two to three fewer sessions; p = 0.001). Significant fewer patients required a primary care physician visit in DA (54% vs. 71%; p = 0.0113). Average cost saving per episode is significantly lower in DA PT (average saving = £36.42/patient/episode of care; p = 0.016)/(average costs £66.31 vs. £89.99; p < 0.05).
Babatunde et al (2020) (17)	No significant differences between patient characteristics in DA/self-referral vs. medical triage (p≥0.05) regarding age and gender. No significant differences in pain, functional outcomes, and QoL for patients who assessed MSK care via DA compared to GP-led care. No reports of adverse effects or missed red flags.		Less work-related absence and sick leave in DA PT. 10%-20% less total cost for DA PT compared to GP-led care. >65% less prescribed analgesics and NSAID. >70% less radiology. 2%-10% lower follow-up consults.
Gallotti et al (2023) (21)	Better management accuracy, less waiting time, and equal to higher QoL.	Higher rate of presence; shorter time to discharge (+6%).	Higher cost-effectiveness (€−441.9 per episode of care) Imaging −28%; medication −41.2%; referral −20.7% Less sick leave (5% less prescriptions, −37 less sick leave days prescribed)

DA = direct access; GP = general practitioner; MSK = musculoskeletal; NSAID = nonsteroidal anti-inflammatory care; PT = physiotherapy; QoL = quality of life; RMQ = Roland-Morris questionnaire; VAS = visual analogue scale.

Discussion

The purpose of this scoping review was to analyse the impact of DA on the quality of MSK primary physiotherapy care from the perspectives of the patient, the provider, and society. No differences were found for pain reduction, QoL, functioning, and well-being although some systematic reviews indicated a tendency in favour of DA physiotherapy. Higher treatment compliance and a more accurate decision-making were found concerning the providers’ perspective and, finally, differences in favour of DA were found for waiting time, prescribed medication, and medical imaging. Less work-related absence and a clear reduction in health care costs were reported in some studies. Although some of the included reviews had a high risk of bias, their findings are in full agreement with the recent low-bias review of Babatunde et al (17).

Proponents of the DA system argue many advantages of this system. For patients, the most important advantage is that the physiotherapist becomes much more accessible, and patients lose less valuable time in the diagnostic process (26). Other advantages were expected to be found in terms of QoL, functioning, and patient satisfaction. From a clinical aspect, DA physiotherapy performs at least as well as physiotherapy by referral. However, all primary studies had relatively short follow-up of about 1 month, and the added benefits of DA may not be captured in the absence of a long-term follow-up.

For physiotherapists, an advantage could be found in the prestige of their profession and the greater responsibility involving an interesting challenge for physiotherapists in the functional evaluation of the patient (9,10). Moreover, several other health professions, such as chiropractors and osteopaths, are directly accessible, while their training courses are less focused on the diagnosis and screening of red and yellow flags (8). Red flags may indicate the presence of a serious underlying cause explaining the current symptoms. However, this must be put in perspective, as the evidential value for red flags has proved insufficient to state that they are excellent predictors of serious underlying disorders (12). Yellow flags identify underlying patient characteristics that could potentially lead to a slower recovery process or ending up in chronicity. It is very important to identify the presence of yellow flags to avoid non-response to the treatment of patients (27).

Furthermore, DA physiotherapy could be a step forward in the physiotherapists’ autonomy and development as a diagnostician. However, it should not become compulsory as some physiotherapists may not consider themselves competent or do not support it. In addition, it would be useful for older physiotherapists who did not follow the most modern training, especially in clinical reasoning and diagnosis, to be given further training in the field of diagnosis. However, no information emerged from this review regarding the responsibility, liability, status, prestige, and job satisfaction of the physiotherapists.

Evidence suggests greater treatment compliance of patients and fewer missed appointments in a DA physiotherapy setting, allowing the physiotherapist to spend their time optimally (21). Some diagnoses made by a GP do not provide added value in the clinical reasoning of the rehabilitation plan. Some diagnoses made by a GP are actually superfluous, as the pattern of symptoms may not have a clear pathophysiological foundation (28,29). The best example of this is the well-known non-specific low back pain phenomenon. In many cases, no underlying pathophysiological mechanism can be found. Treatment by the physiotherapist is then based on the pattern of symptoms and not on the prescribed medical diagnosis (29). Some studies (13) reported that the decision-making ability of physiotherapists is great, but they do not consistently recognize the need for immediate referral. Physiotherapists with MSK specialization were more likely to make correct decisions for patients with MSK conditions and critical medical conditions (13). But overall, the reported results may indicate that training might have to be adapted for this purpose.

Consensus on the benefits of the society perspective has not been found in all areas. DA physiotherapy can decrease the workload and pressure of GPs because a large number of patients with MSK disorders proceed directly to a physiotherapist. In this way, GPs have more time to focus on their other patients. It can also lead to less waiting time, less medical imaging, and fewer prescription and use of medication (17-21). Some studies (17, 19-21) reported less work-related absence and sick leave in DA physiotherapy but others did not find a consensus for the return-to-work rate (18). And, in general, there is evidence of an important cost reduction in health care for MSK disorders (18-21).

As a final argument in favour of DA physiotherapy, it can be noticed that there is no reporting of countries in which DA physiotherapy has been introduced where it was subsequently rejected. This shows that the advantages at least counterbalance, but presumably overweigh the possible disadvantages (30).

Limitations of this scoping review

The information gathering was restricted to English, Dutch, and French and two different databases (MEDLINE and Web of Science), possibly causing some relevant articles to be missed. A specific additional search in the PEDro database for systematic reviews did not reveal any additional publications.

Recommendations for future research

Different types of studies, preferably high-quality RCTs, should be conducted, focusing on various perspectives that remain unanswered or unclear. Economic evaluations could be performed from the societal perspective, and several options exist for designing studies from both the patients’ and therapists’ perspectives.

Conclusions

This scoping review suggests that DA physiotherapy can offer multiple advantages over GP referral physiotherapy. Although no significant effects were found for pain and QoL, strong evidence from one unbiased study, and supported by some lower quality evidence, indicates that DA does not result in a significant decrease in functional outcome. As such DA physiotherapy seems to be as beneficial to the patient as physiotherapy by referral. Moreover, evidence indicates that it does reduce the use of medical imaging and leads to less prescription and use of medication, resulting in costs. The small significant differences in favour of DA physiotherapy from the patients’ perspective, combined with no loss in terms of pain reduction, suggest at least an equal level of care quality. Moreover, it seems that DA physiotherapy does not have any adverse effects on patients. This, coupled with predominantly positive benefits from DA physiotherapy from a societal perspective, suggests that the advantages of DA physiotherapy are more situated in the societal domain.

Disclosures

Conflict of interest: The authors declare no conflict of interest.

Financial support: This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Authors’ contributions: All authors actively participated in the preparation of this manuscript. EC: data curation, formal analysis, investigation, methodology, project administration, supervision, validation, writing original draft, writing review and editing. JVdB: data curation, formal analysis, investigation, supervision writing original draft. RP and AT: data curation, formal analysis, writing original draft. AS: formal analysis, validation. EVT: conceptualization, methodology, supervision, writing review and editing.

Data availability statement: Additional material may be requested directly from the responsible author. The information generated and analysed during the current study are publicly available by means of the mentioned search engines and publications in the reference list