19 Dec 2019: Østerås N, Moseng T, van Bodegom-Vos L, Dziedzic K, Mdala I, et al. (2019) Correction: Implementing a structured model for osteoarthritis care in primary healthcare: A stepped-wedge cluster-randomised trial. PLOS Medicine 16(12): e1002993. https://doi.org/10.1371/journal.pmed.1002993 View correction
To improve quality of care for patients with hip and knee osteoarthritis (OA), a structured model for integrated OA care was developed based on international recommendations. The objective of this study was to assess the effectiveness of this model in primary care.
Methods and findings
We conducted a cluster-randomised controlled trial with stepped-wedge cohort design in 6 Norwegian municipalities (clusters) between January 2015 and October 2017. The randomised order was concealed to the clusters until the time of crossover from the control to the intervention phase. The intervention was implementation of the SAMBA model, facilitated by interactive workshops for general practitioners and physiotherapists with an update on OA treatment recommendations. Patients in the intervention group attended a physiotherapist-led OA education and individually tailored exercise programme for 8–12 weeks. The primary outcome was patient-reported quality of care (OsteoArthritis Quality Indicator questionnaire; 0–100, 100 = optimal quality) at 6 months. Secondary outcomes included patient-reported referrals to physiotherapy, magnetic resonance imaging (MRI), and orthopaedic surgeon consultation; patients’ satisfaction with care; physical activity level; and proportion of patients who were overweight or obese (body mass index ≥ 25 kg/m2). In all, 40 of 80 general practitioners (mean age [SD] 50  years, 42% females) and 37 of 64 physiotherapists (mean age [SD] 42  years, 65% females) participated. They identified 531 patients, of which 393 patients (mean age [SD] 64  years, 71% females) with symptomatic hip or knee OA were included. Among these, 109 patients were recruited during the control periods (control group), and 284 patients were recruited during interventions periods (intervention group). The patients in the intervention group reported significantly higher quality of care (score of 60 versus 41, mean difference 18.9; 95% CI 12.7, 25.1; p < 0.001) and higher satisfaction with OA care (odds ratio [OR] 12.1; 95% CI 6.44, 22.72; p < 0.001) compared to patients in the control group. The increase in quality of care was close to, but below, the pre-specified minimal important change. In the intervention group, a higher proportion was referred to physiotherapy (OR 2.5; 95% CI 1.08, 5.73; p = 0.03), a higher proportion fulfilled physical activity recommendations (OR 9.3; 95% CI 2.87, 30.37; p < 0.001), and a lower proportion was referred to an orthopaedic surgeon (OR 0.3; 95% CI 0.08, 0.80; p = 0.02), as compared to the control group. There were no significant group differences regarding referral to MRI (OR 0.6; 95% CI 0.13, 2.38; p = 0.42) and proportion of patients who were overweight or obese (OR 1.3; 95% CI 0.70, 2.51; p = 0.34). Study limitations include the imbalance in patient group size, which may have been due to an increased attention to OA patients among the health professionals during the intervention phase, and a potential recruitment bias as the patient participants were identified by their health professionals.
In this study, a structured model in primary care resulted in higher quality of OA care as compared to usual care. Future studies should explore ways to implement the structured model for integrated OA care on a larger scale.
Why was this study done?
- Hip and knee osteoarthritis is a common chronic joint disease in the adult population causing significant pain and disability.
- Non-surgical treatment modalities including patient osteoarthritis education, exercise therapy, and weight management represent core treatments recommended in professional guidelines. However, they are currently underutilised in people with hip and knee osteoarthritis.
- It is not established to what extent a structured osteoarthritis care model can change this and improve the quality of care.
What did the researchers do and find?
- A cluster-randomised trial was conducted to compare a structured osteoarthritis care model with usual care with respect to appropriate care delivery in people with hip and knee osteoarthritis.
- Forty general practitioners and 37 physiotherapists working in primary care attended workshops to get an update on recommendations for osteoarthritis care and were trained in the core elements of the structured care model: osteoarthritis education in groups, an individually tailored 8- to 12-week exercise programme, and a dietary intervention, if needed.
- Of the 393 patient participants, 284 were allocated to the intervention group and 109 to the usual care group.
- At 6 months, patient-reported quality of care and satisfaction with care were greater, more patients were referred to physiotherapy and fewer to orthopaedic surgeons, and more patients fulfilled physical activity criteria in the intervention group as compared to the usual care group.
What do these findings mean?
- A structured osteoarthritis care model provided by trained primary care general practitioners and physiotherapists resulted in the provision of osteoarthritis care that was more in line with current care recommendations and in higher patient-reported quality of care and satisfaction as compared to usual care.
- A structured and well-planned approach, in line with evidence-based treatment recommendations for hip and knee osteoarthritis and executed in primary care, has the potential to improve patients’ health and reduce disability. In doing so, it may also reduce the risk of sick leave and may thereby reduce the direct and indirect costs of osteoarthritis for the individual and the society.
- Although a stepped-wedge cluster-randomised controlled trial design is appropriate to conduct an effectiveness study in a clinical practice setting, strategies to prevent selection bias and differences in recruitment rates in the control and intervention periods are needed.
Citation: Østerås N, Moseng T, Bodegom-Vos Lv, Dziedzic K, Mdala I, Natvig B, et al. (2019) Implementing a structured model for osteoarthritis care in primary healthcare: A stepped-wedge cluster-randomised trial. PLoS Med 16(10): e1002949. https://doi.org/10.1371/journal.pmed.1002949
Academic Editor: Christelle Nguyen, Univ. Paris Descartes, PRES Sorbonne Paris Cité, Hôpital Cochin, Assistance Publique - Hôpitaux de Paris, FRANCE
Received: May 8, 2019; Accepted: September 24, 2019; Published: October 15, 2019
Copyright: © 2019 Østerås et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Due to the ethical approval and requirements of the data protection legislation, the statistical code and data set will only be made available after July 1st 2021 on a restricted basis according to the data sharing policies (https://nsd.no/nsddata/utlaansrutiner_en.html) at the Norwegian Centre for Research Data (www.nsd.no). Applications for access to anonymised data can be found at: https://nsd.no/nsd/english/order.html or by sending an e-mail to email@example.com.
Funding: This study was funded by The Research Council of Norway (No. 229079). The Norwegian Fund for Post-Graduate Training in Physiotherapy funded a PhD student (TM) (No.62458). KD was part-funded by a Knowledge Mobilisation Research Fellowship (KMRF-2014-03-002) from the NIHR, the NIHR Collaborations for Leadership in Applied Health Research and Care West Midlands and is a NIHR Senior Investigator. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Abbreviations: BMI, body mass index; cluster-RCT, cluster-randomised controlled trial; GP, general practitioner; ICC, intraclass correlation coefficient; MRI, magnetic resonance imaging; OA, osteoarthritis; OA-QI v[number], OsteoArthritis Quality Indicator questionnaire version [number]; OR, odds ratio; PT, physiotherapist; QI, quality indicator; SD, standard deviation
Osteoarthritis (OA) is one of the leading causes of pain and disability in the adult population worldwide and a major contributor to years lived with disability [1,2]. Prevalence of OA increases with age, and with an aging population and the epidemic of obesity, it is set to rise . The costs of treatment and work-related losses represent a considerable economic burden [4,5]. Recommended first-line core treatments include patient education, self-management, exercise, and weight reduction [1,6–8]. When non-pharmacological and pharmacological care fail, joint replacement offers an effective approach, although it is costly and associated with medical and surgical risks [9–12]. The demand for joint replacement is expected to accelerate and quadruple by 2030 with the increasing prevalence of OA . Decisions on joint replacement involve conventional radiographs, whereas magnetic resonance imaging (MRI) is usually considered unnecessary .
An evidence-to-practice gap for OA care has been identified internationally, with poor uptake of non-pharmacological approaches such as patient education and exercise treatment in contrast to surgical treatment [15,16]. Furthermore, as 22%–68% of joint replacements are considered inappropriate , it is important to improve the uptake of non-surgical care. A small number of best practice initiatives to improve the quality of OA care have shown promising but somewhat diverging results [18–24]. Inspired by these previous initiatives [18–25] and based on international recommendations for OA care [1,6–8], the SAMBA model for integrated care for patients with hip and knee OA  was developed by the research team for evaluation in a randomised controlled trial. SAMBA is an acronym formed from the Norwegian project title, ‘SAMhandling for Bedre Artrosebehandling i kommunehelsetjenesten’, which can be translated as ‘improved management of patients with hip and knee osteoarthritis in primary healthcare’.
The main aim of the present study was to assess the effectiveness of the SAMBA model in primary healthcare. We hypothesised that compared to usual OA care, the SAMBA model would increase the uptake of best practices for OA, demonstrate higher patient satisfaction with care, and increase beneficial lifestyle characteristics (physical activity, healthy weight). We hypothesised that compared with usual care, the SAMBA model would offer improvements in referral pathways, e.g., more general practitioner (GP) referrals to physiotherapy, more discharge reports from physiotherapy to referring GPs, and fewer GP referrals to MRI and orthopaedic surgeons.
Design, setting, and participants
We performed a cluster-randomised controlled trial (cluster-RCT) with a stepped-wedge cohort design between 15 January 2015 and 20 October 2017. The study was conducted in 6 neighbouring municipalities (clusters) north of Oslo, Norway, with approximately 100,000 inhabitants in total. The stepped-wedge design is explained in Fig 1 and in the published study protocol . The study was prospectively registered at ClinicalTrials.gov (NCT02333656) and is reported according to the CONSORT (S1 Text) and TIDieR (S2 Text) checklists [27,28].
All 6 municipalities (clusters) started the trial simultaneously with a control phase (general practitioners and physiotherapists providing usual care). At predefined time points about every second month, one by one the municipalities crossed from the control to the intervention phase (use of the SAMBA model) in a randomised order. Light cells in the figure represent control periods, and dark cells represent intervention periods. The asterisks indicate the timing of the interactive workshops before switching to the intervention phase. Patients recruited to the study during the control phase in any cluster constituted the control group, whereas patients recruited during the intervention phase constituted the intervention group. All patients responded to the baseline questionnaire and follow-up questionnaires at 3, 6, 9, and 12 months post-baseline. aLarge municipalities (clusters) had >20,000 inhabitants.
The Regional Committee for Medical and Health Research Ethics issued a letter of exemption for the current study (Ref. No: 2014/1739 REK south-east C). The Data Inspectorate/Data Protection Official of Oslo University Hospital approved the study on 22 December 2014. Written informed consent was obtained from patients upon inclusion.
All GPs and physiotherapists (PTs) working in private practice or healthy life centres in the 6 municipalities were invited to participate. Healthy life centres provide primary-care-based services aiming to support a healthy lifestyle for people with chronic diseases .
Potential eligible patients were identified at clinical visits by the GPs and PTs using the following inclusion criteria: age ≥ 45 years with symptomatic hip and/or knee OA diagnosis verified clinically or radiologically by the GP. Patients who did not understand Norwegian or who had 4 joint replacements (hip + knee), inflammatory rheumatic disease, malignant illness, or any other major condition that restricted their ability to adhere to the intervention were excluded . A study coordinator performed the eligibility screening and inclusion procedure.
Randomisation and blinding
Immediately before study initiation, the municipalities were randomly allocated to 1 of the 6 sequences for time of crossover from the control to the intervention phase (Fig 1) using a computer-generated list of random numbers provided by a statistician. To ensure a mix of municipality sizes in the randomised sequence, stratification on the number of inhabitants (less than versus more than 20,000) was performed. The randomised order was concealed to the clusters until soon before the switch. It was not possible to blind the involved GPs, PTs, or patients, but a statistician blinded for group allocation performed the statistical analyses of the primary outcome.
The SAMBA model for integrated OA care was developed by the research team and comprised a structured pathway for patients with OA through the healthcare system (Fig 2). The model included a GP consultation, a PT-led OA education and exercise programme (ActiveA), an optional healthy eating program, and a GP review consultation. The GPs were instructed to explain the OA diagnosis and treatment alternatives, provide pharmacological treatment when appropriate, and suggest referral to physiotherapy. The PT-led patient OA education programme was group-based and lasted 3 hours. This was followed by an 8–12 week exercise programme with twice weekly 1-hour supervised group sessions (5–10 patients per PT). Based on patient examination, the PT prescribed individually tailored resistance exercise programmes to increase muscular strength. The pool of recommended exercises was selected from previous OA exercise studies [21,22,30,31]. Dose recommendations were based on acknowledged international guidelines , and included gradually increasing the dose towards 2–4 sets with 8–12 repetitions and 60%–70% of 1 repetition maximum, or more if tolerated. The PTs were instructed to closely monitor the patients’ exercise performance and regularly provide appropriate individual adjustments of the exercise programme for progression. When the patient could perform 2 extra repetitions in the last set, the resistance was increased (‘the 2+ principle’). The patients were encouraged to add a third home-based session consisting of 30–60 minutes of cardiorespiratory exercise like brisk walking, running, or bicycling.
GP, general practitioner; OA, osteoarthritis; PT, physiotherapist.
The model intends to improve the quality of OA care through delivery of evidence-based recommendations for OA care, focusing on the core treatments, first-line analgesics, and facilitating multidisciplinary collaboration. Informed by theoretical models [33,34] and previously developed behaviour change interventions , a strategy  was designed to facilitate the use of the model among GPs and PTs in primary healthcare in the intervention phase (S3 Text). Tailored interactive workshops for GPs and PTs were arranged in close proximity to the time of crossover (Fig 1) and reflected the main intervention, ensuring the use of the model in the intervention phase. Other strategies to facilitate the use of the SAMBA model in the intervention phase included summarised treatment recommendations, regular telephone reminders, quarterly letters with feedback, and biannual outreach visits known to be effective in primary care (see details in the published protocol ).
The PT workshop.
The PT workshop included a 1-day (9 hours) workshop-based education programme organised by the Active with osteoArthritis (ActiveA) programme , which builds on the similar Swedish  and Danish  programmes. The workshop included an update on OA epidemiology, clinical features, and treatment recommendations. Education in delivery of a patient OA education programme, individually tailored semi-standardised exercises, performance testing, and healthy eating and weight reduction strategies was given. The PTs received access to the ready-to-use patient OA education programme (PowerPoint file and manuscript) and access to a database with recommendations for resistance exercises and dose.
The multidisciplinary workshop.
The 1.5-hour multidisciplinary (GPs and PTs) workshops were conducted within the general practices at established meeting time points in order to facilitate high GP attendance. The workshops included an update on current treatment recommendations. An orthopaedic surgeon presented views on when to consider referral to consultation with an orthopaedic surgeon, emphasising the importance of exploiting conservative treatment first. The research team presented the SAMBA model and facilitated a multidisciplinary discussion regarding OA care.
Attendance at workshops, patient adherence to the OA education and exercise programme, and potential adverse effects were captured from study records and patient-reported exercise diaries. A comprehensive analysis of fidelity has been published in a separate article , and the other secondary outcomes will be reported separately.
During the control phase, the GPs and the PTs delivered usual care and were naïve to the SAMBA model. Usual care may include infrequent GP visits, pharmacological therapy, and occasionally a referral to physiotherapy. Patients included during the control phase were allowed to receive physiotherapy, but not to attend the patient OA education programme nor the individually tailored exercise programme, prior to 12 months post-baseline. Physiotherapy provided to patients with knee OA may include exercise, but also often includes several other treatment modalities (e.g., massage, traction/mobilisation, stretching, and electrotherapy) showing moderate or low quality of evidence, or no evidence, in systematic reviews .
Patients self-reported at baseline (shortly after the GP consultation) and at 3 months (T3) and 6 months (T6) using an electronic questionnaire or mailed paper questionnaire returned in a prepaid envelope. The primary time point was T6, except for referral to physiotherapy (T3). Data on long-term follow-up (9 and 12 months) will be published later. Information on the patients’ age, sex, previous or planned joint replacements, and comorbidity was collected by the study coordinator during telephone screening. Other patient characteristics and OA-disease-related information, including the Knee injury and Osteoarthritis Outcome Score/Hip disability and Osteoarthritis Outcome Score Activities of Daily Living subscale (KOOS/HOOS ADL subscale), were self-reported at baseline. The GPs and PTs self-reported demographics and practice information in a questionnaire.
Primary outcome measure
The primary outcome was patient-reported quality of OA care at T6 measured with the OsteoArthritis Quality Indicator questionnaire version 2 (OA-QI v2) (S4 Text) . OA-QI v1 was developed in 2010 based on published quality indicators (QIs) for OA care identified in a literature search and further refined via expert panels and patient interviews , and was slightly revised in 2015 . OA-QI v2 reflects current OA care guideline recommendations [1,6–8] and includes 16 QI items related to patient OA education and information, regular provider assessments, referrals, and pharmacological treatment. An example of an item with response alternatives is as follows: ‘Have you been given information about osteoarthritis from a health professional? Yes/No/Don’t remember’.
All items have ‘Yes’, ‘No’, and ‘Not applicable’/‘Don’t remember’ as response options. Each QI item was considered passed if the patient had checked ‘Yes’ and was considered ‘eligible’ if the patient responded ‘Yes’ or ‘No’ for that item. On the patient level, the QI pass rate was calculated as the total number of items passed divided by the number of eligible items for each patient (in percentage), ranging from 0 to 100, with 100 representing the best quality of care score. On the group level, the mean total pass rate was calculated.
Previous applications of the questionnaire have showed acceptable measurement properties including reliability, validity, responsiveness, and interpretability [39,40]. The test–retest reliability for the total pass rate was acceptable (intraclass correlation coefficient [ICC] 0.89) . All predefined hypotheses to assess construct validity were confirmed, and responsiveness was acceptable, with 3 of 4 predefined hypotheses confirmed . Minimal important change for the total pass rate was assessed to be 20.4 on the 0 to 100 scale . OA-QI v1 has been previously tested in UK primary care in a cluster-randomised trial and has been shown to be responsive to the use of national recommendations for OA care .
Secondary outcome measures
GP referrals to PTs, MRI, and orthopaedic surgeons were patient self-reported as ‘Yes’ or ‘No’.
The Norwegian Health Economics Administration provided data on the total number of registered discharge reports for all patients for participating PTs during the control and intervention phases. This number may also include discharge reports related to non-participating patients and should be interpreted with caution.
Patients’ satisfaction with OA care was assessed using 1 item, with 5 response alternatives ranging from ‘Very satisfied’ to ‘Very dissatisfied’, from a previous study .
Physical activity was reported using 3 items on frequency (never/less than once per week/once per week/2 to 3 times per week/4 or more times per week), duration (less than 15 minutes/15–30 minutes/31–60 minutes/more than 60 minutes), and intensity (no sweat/sweat [moderate]/exhausted [vigorous]) . Using an index from a previous study , we calculated the proportion of patients ‘fulfilling’ versus ‘below’ recommendations. Corresponding to the physical activity recommendations at the time of that study, ‘fulfilling’ was 150 minutes of moderate-intensity activity or 60 minutes of vigorous-intensity activity each week, or a combination of these.
The proportion of patients who were overweight (body mass index [BMI] ≥ 25 kg/m2) or obese (BMI ≥ 30 kg/m2) was defined using patients’ self-reported body height at baseline and body weight at follow-ups.
Other patient-level secondary outcomes (e.g., symptoms, function, quality of life, and performance tests) and an economic evaluation will be reported later.
Sample size calculation
Based on previous research , we estimated the ICC to be <0.01. We estimated that a minimum of 194 individuals in each group among the 6 clusters, with an average of 50 individuals per cluster, would achieve 80% power to detect a 10-unit difference between the group means on the primary outcome measure, where standard deviation for the primary outcome measure was 24 units and intra-class correlation was 0.01, using a 2-sided test with a significance level of 0.05. To account for 30% patient dropout, we aimed to include 388 patients in total.
The primary outcome analysis was performed on an intention to treat basis by comparing OA-QI v2 mean total pass rate in the control versus the intervention group. Multilevel mixed models with random intercepts were fitted to adjust for the effect of clustering (municipality), participant (patient), and repeated measures over time. This model accounts for dropout under a missing at random assumption. The primary outcome was assessed with a linear model, whereas the secondary outcomes were assessed with logistic models except for patient satisfaction with OA care, which was assessed applying multilevel ordered logistic regression models. All regression models included an interaction term of follow-up time point and group, and were adjusted for age, sex, and secular time (number of months between study initiation and the patient entering the study). Statistical analyses were performed with STATA/IC 14.
Patient and public involvement and engagement
Patient research partners were involved in all stages of this trial, from grant application, development of study materials (including questionnaire and consent procedure) and intervention package, and interpretation of the results to final dissemination of results. Two patient research partners (ØA and JNH) were members of the trial steering committee and are co-authors of the present article.
Forty (50%) of the 80 GPs and 37 (50%) of the 64 PTs in the 6 municipalities (clusters) attended the workshops. Of the 531 patients identified by these GPs and PTs, 393 (74%) fulfilled the inclusion criteria and were willing to participate (Fig 3). Patients who were excluded (n = 138) were similar to patients who were included in terms of sex and age. In total, 109 patients (control group) were recruited during the control periods across the clusters, and 284 patients (intervention group) were recruited during interventions periods (Fig 1). Baseline characteristics of the patients, GPs, and PTs are provided in Table 1. In total, 89% of the patients (n = 349) completed the T3 questionnaire and 88% (n = 346) completed the T6 questionnaire (Fig 3). Patients who completed versus did not complete these questionnaires did not differ with regards to baseline characteristics.
ITT, intention to treat; OA, osteoarthritis.
In total, 27 PT-led patient OA education and exercise groups were arranged, with 92% (n = 261) of the patients in the intervention group attending the OA education programme, and 64% (n = 184) completing ≥8 weeks of the exercise programme. The completers versus non-completers among the intervention patients did not differ regarding sex or age.
Seven (6%) patients in the control and 3 (1%) in the intervention group received joint replacement surgery between baseline and T6. Four patients in the intervention group experienced increased prolonged knee pain and/or swelling and discontinued the exercise programme at the halfway stage. Two patients in the control group receiving physiotherapy (usual care) erroneously attended the PT-led education and exercise programme after their PT had attended the workshop.
At baseline, the OA-QI v2 mean total pass rate was similar for the intervention (39%) and the control group (37%). At T6, the pass rate was higher (60%) for the intervention as compared to the control group (41%) (Table 2). This was due to a higher uptake of non-pharmacological treatment recommendations in the intervention group, and in particular the core treatments: patient education about the disease and treatment alternatives, self-management, and exercise (Table 2). Adjusted multilevel mixed model analyses showed a statistically significant difference in mean total pass rate at T6, with higher quality of care in the intervention as compared to the control group (mean difference 18.9; 95% CI 12.7, 25.1; p < 0.001) (Table 2; Fig 4). The crude ICC was 0.016.
Mean patient-reported quality of care with 95% confidence interval. Patient-reported quality of care captured by OsteoArthritis Quality Indicator questionnaire version 2 (0–100, 100 = best score).
At T3, a significantly higher proportion of patients in the intervention group reported PT referrals compared to the control group (36% versus 25%; odds ratio [OR] 2.5; 95% CI 1.08, 5.73; p = 0.03) (Table 3). The number of physiotherapy discharge reports was 59 for the control and 127 for the intervention periods. At T6, a negligibly lower proportion was referred to MRI (5% versus 8%; OR 0.6; 95% CI 0.13, 2.38; p = 0.42), but a significantly lower proportion was referred to an orthopaedic surgeon (6% versus 13%; OR 0.3; 95% CI 0.08, 0.80; p = 0.02), in the intervention as compared to the control group (Table 3).
Compared to the control group, the intervention group had a significantly higher OR for reporting satisfaction with OA care at T6 (OR 12.1; 95% CI 6.44, 22.72; p < 0.001). A significantly higher proportion in the intervention as compared to the control group (67% versus 45%) fulfilled the recommendation for weekly physical activity at T6 (OR 9.3; 95% CI 2.87, 30.37; p < 0.001) (Table 3). The proportion of patients who were overweight or obese remained similar in the 2 groups (69% versus 67%; OR 1.3; 95% CI 0.70, 2.51; p = 0.34) (Table 3).
The evidence-to-practice gap for OA care and the diverging results in previous studies aiming to improve OA care highlight the need for care models that increase adherence by practitioners to recommendations for OA care. This cluster-RCT assessing a structured model in primary healthcare is, to our knowledge, among the first to show an increased uptake of core OA treatment recommendations among GPs and PTs. Patients in the intervention group reported significantly higher quality of care than patients in the control group, and had better outcomes for 4 of the 6 secondary outcomes related to satisfaction with care, referral pathways, and beneficial lifestyles.
The observed between-group difference in OA-QI v2 mean total pass rate at 6-month follow-up (T6) indicates that the use of the structured OA care model successfully improved the delivery of OA care among GPs and PTs in primary healthcare in this study. While the mean total pass rate at baseline in this study was comparable to that in previous studies [15,16], the mean total pass rate at T6 for the intervention group was higher than the pass rates in most previous studies. The increase in mean total pass rate in the intervention group was comparable to the increase observed in a Norwegian longitudinal patient cohort study including a patient OA education programme . Despite the significant between-group difference, the increase in mean total pass rate in the current study did not meet the minimal important change for this outcome measure, but was very close.
Among previous studies aiming to improve OA care, a UK cluster-RCT  implementing a model OA consultation in general practice and a Dutch randomised controlled trial  with interactive clinical workshops for PTs resulted in improved guideline adherence. The UK cluster-RCT  resulted in increased referral rates to physiotherapy, as did the current study. However, in a Dutch cohort study implementing a stepped OA care strategy in general practice, the provided care became only modestly consistent with the strategy . This conflicting result may be due to differences in design, settings, and models to improve OA care. The current study and the UK cluster-RCT  had a multidisciplinary approach, whereas the 2 Dutch studies focused on either PTs  or GPs . Targeting more than one health profession may add beneficial effects including improved multidisciplinary collaboration, integrated care, and consistent patient information.
The mean total pass rate in the intervention group at T6 indicates that there is a potential for even further improvements in provided care. Looking at individual quality of care items in OA-QI v2, the improvement was particularly evident for core treatment elements related to patient information, self-management, and exercise. This would be expected as these core elements were the main focus of the study intervention, and an improvement on all individual OA-QI v2 items would not be realistic in this study. The improvement for provision of core treatments is in line with the UK cluster-RCT, in which the intervention also resulted in higher uptake of core guideline recommendations . By scrutinising the individual QI item pass rates, ideas for further improvements of the care model may be generated, e.g., regarding advice and support on weight reduction.
The study intervention had a positive effect on patient satisfaction with care, patient physical activity, and referral rates to PTs and orthopaedic surgeons. The complex intervention in this study means that there are many factors that could have contributed to the higher patient satisfaction with care, e.g., the structured approach to OA care or the updated knowledge among GPs and PTs on current OA care recommendations. Unfortunately, it is not possible to disentangle the effect of the different factors in this study. The increase in PT referrals and the decrease in referrals to orthopaedic surgeons demonstrate that the GPs in this study adhered to the structured OA care model and that the patient participants exploited the core treatments before considering referral for surgical consultation.
However, the intervention showed negligible effects on MRI referrals and on the proportion of patients who were overweight or obese. Regarding MRI, fewer referrals at T6 as compared to baseline were observed in the intervention group, but there was also a decrease in the control group. The reason for the decrease in the control group is unknown, but may be related to the relatively high referral rate at baseline for both groups, which reduced the number of candidates for MRI referrals at follow-up time points. Although the structured care model included an optional healthy eating program, an intensive diet may have been required to observe an effect on the patients’ BMI and the proportion of patients who were overweight or obese over the time period studied here .
Strengths and limitations
The study has several strengths including the robust but pragmatic study design, the multidisciplinary approach, and the large patient sample. The stepped-wedge design was chosen since it allowed all clusters to test the intervention, and the GP and PT training could be done in one cluster at the time over a longer period. However, the complexity of the stepped-wedge design makes reporting of study methods and results more challenging as compared to more traditional designs.
The current study also has some limitations. One is the unbalanced group size caused by a higher patient recruitment rate during the intervention phase. This could be related to the stepped-wedge design and the GPs’ and PTs’ workshop participation, which likely increased the attention towards OA patients and access to PT-led OA education and exercise groups. We do not think this has influenced the generalisability of the sample, but because the patient participants were identified by their GPs and PTs, recruitment bias may potentially exist. Furthermore, self-reported QI pass rates, referrals, and body weight may be somewhat inaccurate due to recall bias and misconception, but the inaccuracy is likely to be similar across the groups. However, patients in the intervention group may have overestimated their physical activity levels more than those in the control group, potentially leading to bias.
Clinically effective and cost-effective treatments applied to large numbers of people with OA could result in substantial population health gains and reduced costs. We have demonstrated the uptake of recommended treatments in the SAMBA model in routine clinical practice. Clear estimates of the potential for clinical effectiveness and cost-effectiveness of the SAMBA model of care from the patients’ and societal perspectives are still needed.
Implications for clinicians, policy makers, and future research
People with hip and knee OA represent a large, common patient group for GPs and PTs, and the provision of local, interdisciplinary workshops may facilitate multidisciplinary collaboration and ensure delivery of consistent patient information. When GPs and PTs are guided in the steps of structured, evidence-based care models, people with hip or knee OA may receive care that is more in line with current recommendations for OA care. Previous research has shown that people with knee OA have an almost 2-fold increased risk of sick leave and a 40%–50% increased risk of disability pension , and that the costs of hip and knee OA are substantial . Hence, policy makers could consider facilitation of early secondary prevention strategies aiming to reduce the burden of the disease, which may potentially reduce the direct and indirect costs for the individual and the society. In addition, facilitation of treatment in groups (e.g., patient education and exercise sessions) represents an opportunity to provide access to care for more patients and to lower the costs compared to treating individual patients. A further improvement of the structured model may be included in future research, as well as exploring ways to implement the structured model for integrated OA care on a larger scale. The recent focus worldwide on healthcare overuse behaviours could be utilised to strengthen the work on preventing overuse of unnecessary treatments in OA care (e.g., opioids and imaging).
This study demonstrated that a structured care model among GPs and PTs in primary healthcare improved the quality of care for patients with hip and knee OA. The model may be adapted to other chronic diseases treated in primary healthcare.
S3 Text. Strategy to facilitate the use of the SAMBA model.
The authors would like to thank the health professionals, their leaders, and all the patients who participated in this study.
The views expressed in this paper are those of the authors and not necessarily those of the UK National Health Service, National Institute for Health Research, or Department of Health and Social Care.
- 1. National Institute for Health and Care Excellence. Osteoarthritis: care and management in adults. Clinical guideline (CG177). London: National Institute for Health and Care Excellence; 2014.
- 2. Vos T, Flaxman AD, Naghavi M, Lozano R, Michaud C, Ezzati M, et al. Years lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2013;380:2163–96.
- 3. Bijlsma JW, Berenbaum F, Lafeber FP. Osteoarthritis: an update with relevance for clinical practice. Lancet. 2011;377:2115–26. pmid:21684382
- 4. Bitton R. The economic burden of osteoarthritis. Am J Manag Care. 2009;15:S230–5. pmid:19817509
- 5. Ackerman IN, Pratt C, Gorelik A, Liew D. Projected burden of osteoarthritis and rheumatoid arthritis in Australia: a population-level analysis. Arthritis Care Res (Hoboken). 2018;70:877–83.
- 6. Hochberg MC, Altman RD, April KT, Benkhalti M, Guyatt G, McGowan J, et al. American College of Rheumatology 2012 recommendations for the use of nonpharmacologic and pharmacologic therapies in osteoarthritis of the hand, hip, and knee. Arthritis Care Res (Hoboken). 2012;64:455–74.
- 7. Fernandes L, Hagen KB, Bijlsma JW, Andreassen O, Christensen P, Conaghan PG, et al. EULAR recommendations for the non-pharmacological core management of hip and knee osteoarthritis. Ann Rheum Dis. 2013;72:1125–35. pmid:23595142
- 8. McAlindon TE, Bannuru RR, Sullivan MC, Arden NK, Berenbaum F, Bierma-Zeinstra SM, et al. OARSI guidelines for the non-surgical management of knee osteoarthritis. Osteoarthritis Cartilage. 2014;22:363–88. pmid:24462672
- 9. Daigle ME, Weinstein AM, Katz JN, Losina E. The cost-effectiveness of total joint arthroplasty: a systematic review of published literature. Best Pract Res Clin Rheumatol. 2012;26:649–58. pmid:23218429
- 10. Dakin H, Gray A, Fitzpatrick R, Maclennan G, Murray D, KAT Trial Group. Rationing of total knee replacement: a cost-effectiveness analysis on a large trial data set. BMJ Open. 2012;2:e000332. pmid:22290396
- 11. Gademan MG, Hofstede SN, Vliet Vlieland TP, Nelissen RG, Marang-van de Mheen PJ. Indication criteria for total hip or knee arthroplasty in osteoarthritis: a state-of-the-science overview. BMC Musculoskelet Disord. 2016;17:463. pmid:27829422
- 12. Bohl DD, Ondeck NT, Basques BA, Levine BR, Grauer JN. What is the timing of general health adverse events that occur after total joint arthroplasty? Clin Orthop Relat Res. 2017;475:2952–9. pmid:28054326
- 13. Ghomrawi HM, Schackman BR, Mushlin AI. Appropriateness criteria and elective procedures—total joint arthroplasty. N Engl J Med. 2012;367:2467–9. pmid:23268663
- 14. Sakellariou G, Conaghan PG, Zhang W, Bijlsma JWJ, Boyesen P, D’Agostino MA, et al. EULAR recommendations for the use of imaging in the clinical management of peripheral joint osteoarthritis. Ann Rheum Dis. 2017;76:1484–94. pmid:28389554
- 15. Basedow M, Esterman A. Assessing appropriateness of osteoarthritis care using quality indicators: a systematic review. J Eval Clin Pract. 2015;21:782–9. pmid:26083547
- 16. Hagen KB, Smedslund G, Osteras N, Jamtvedt G. Quality of community-based osteoarthritis care: a systematic review and meta-analysis. Arthritis Care Res (Hoboken). 2016;68:1443–52.
- 17. Quintana JM, Arostegui I, Escobar A, Azkarate J, Goenaga JI, Lafuente I. Prevalence of knee and hip osteoarthritis and the appropriateness of joint replacement in an older population. Arch Intern Med. 2008;168:1576–84. pmid:18663171
- 18. Allen KD, Oddone EZ, Coffman CJ, Datta SK, Juntilla KA, Lindquist JH, et al. Telephone-based self-management of osteoarthritis: a randomized trial. Ann Intern Med. 2010;153:570–9. pmid:21041576
- 19. Smink AJ, Dekker J, Vliet Vlieland TP, Swierstra BA, Kortland JH, Bijlsma JW, et al. Health care use of patients with osteoarthritis of the hip or knee after implementation of a stepped-care strategy: an observational study. Arthritis Care Res (Hoboken). 2014;66:817–27.
- 20. Smink AJ, Bierma-Zeinstra SM, Schers HJ, Swierstra BA, Kortland JH, Bijlsma JW, et al. Non-surgical care in patients with hip or knee osteoarthritis is modestly consistent with a stepped care strategy after its implementation. Int J Qual Health Care. 2014;26:490–8. pmid:24845068
- 21. Thorstensson CA, Garellick G, Rystedt H, Dahlberg LE. Better management of patients with osteoarthritis: development and nationwide implementation of an evidence-based supported osteoarthritis self-management programme. Musculoskeletal Care. 2015;13:67–75. pmid:25345913
- 22. Skou ST, Roos EM. Good Life with osteoArthritis in Denmark (GLA:D): evidence-based education and supervised neuromuscular exercise delivered by certified physiotherapists nationwide. BMC Musculoskelet Disord. 2017;18:72. pmid:28173795
- 23. Dziedzic KS, Healey EL, Porcheret M, Afolabi EK, Lewis M, Morden A, et al. Implementing core NICE guidelines for osteoarthritis in primary care with a model consultation (MOSAICS): a cluster randomised controlled trial. Osteoarthritis Cartilage. 2018;26:43–53. pmid:29037845
- 24. Allen KD, Oddone EZ, Coffman CJ, Jeffreys AS, Bosworth HB, Chatterjee R, et al. Patient, provider, and combined interventions for managing osteoarthritis in primary care: a cluster randomized trial. Ann Intern Med. 2017;166:401–11. pmid:28114648
- 25. Porcheret M, Grime J, Main C, Dziedzic K. Developing a model osteoarthritis consultation: a Delphi consensus exercise. BMC Musculoskelet Disord. 2013;14:25. pmid:23320630
- 26. Osteras N, van Bodegom-Vos L, Dziedzic K, Moseng T, Aas E, Andreassen O, et al. Implementing international osteoarthritis treatment guidelines in primary health care: study protocol for the SAMBA stepped wedge cluster randomized controlled trial. Implement Sci. 2015;10:165. pmid:26631224
- 27. Hemming K, Taljaard M, McKenzie JE, Hooper R, Copas A, Thompson JA, et al. Reporting of stepped wedge cluster randomised trials: extension of the CONSORT 2010 statement with explanation and elaboration. BMJ. 2018;363:k1614. pmid:30413417
- 28. Hoffmann TC, Glasziou PP, Boutron I, Milne R, Perera R, Moher D, et al. Better reporting of interventions: template for intervention description and replication (TIDieR) checklist and guide. BMJ. 2014;348:g1687. pmid:24609605
- 29. Norwegian Directorate of Health. [Guide for healthy life centres in primary health care. Establishment, organising and services.] Oslo: Norwegian Directorate of Health; 2016.
- 30. Fernandes L, Storheim K, Nordsletten L, Risberg MA. Development of a therapeutic exercise program for patients with osteoarthritis of the hip. Phys Ther. 2010;90:592–601. pmid:20185613
- 31. Stensrud S, Roos EM, Risberg MA. A 12-week exercise therapy program in middle-aged patients with degenerative meniscus tears: a case series with 1-year follow-up. J Orthop Sports Phys Ther. 2012;42:919–31. pmid:22960783
- 32. Garber CE, Blissmer B, Deschenes MR, Franklin BA, Lamonte MJ, Lee IM, et al. American College of Sports Medicine position stand. Quantity and quality of exercise for developing and maintaining cardiorespiratory, musculoskeletal, and neuromotor fitness in apparently healthy adults: guidance for prescribing exercise. Med Sci Sports Exerc. 2011;43:1334–59. pmid:21694556
- 33. Grol RW M, Eccles M, Davis D. Improving patient care: the implementation of change in health care. 2nd edition. West Sussex: John Wiley & Sons; 2013.
- 34. Graham ID, Logan J, Harrison MB, Straus SE, Tetroe J, Caswell W, et al. Lost in knowledge translation: time for a map? J Contin Educ Health Prof. 2006;26:13–24. pmid:16557505
- 35. Porcheret M, Main C, Croft P, McKinley R, Hassell A, Dziedzic K. Development of a behaviour change intervention: a case study on the practical application of theory. Implement Sci. 2014;9:42. pmid:24708880
- 36. Holm I, Risberg MA, Roos EM, Skou ST. A pragmatic approach to the implementation of osteoarthritis guidelines has fewer potential barriers than recommended implementation frameworks. J Orthop Sports Phys Ther. 2019;49:1–4. pmid:30598054
- 37. Moseng T, Dagfinrud H, Osteras N. Implementing international osteoarthritis guidelines in primary care: uptake and fidelity among health professionals and patients. Osteoarthritis Cartilage. 2019;27:1138–47. pmid:31075423
- 38. Jamtvedt G, Dahm KT, Holm I, Flottorp S. Measuring physiotherapy performance in patients with osteoarthritis of the knee: a prospective study. BMC Health Serv Res. 2008;8:145. pmid:18611250
- 39. Osteras N, Tveter AT, Garratt AM, Svinoy OE, Kjeken I, Natvig B, et al. Measurement properties for the revised patient-reported OsteoArthritis Quality Indicator questionnaire. Osteoarthritis Cartilage. 2018;26:1300–10. pmid:30231991
- 40. Osteras N, Garratt A, Grotle M, Natvig B, Kjeken I, Kvien TK, et al. Patient-reported quality of care for osteoarthritis: development and testing of the osteoarthritis quality indicator questionnaire. Arthritis Care Res (Hoboken). 2013;65:1043–51.
- 41. Skudal KE, Garratt AM, Eriksson B, Leinonen T, Simonsen J, Bjertnaes OA. The Nordic Patient Experiences Questionnaire (NORPEQ): cross-national comparison of data quality, internal consistency and validity in four Nordic countries. BMJ Open. 2012;2:e000864. pmid:22649175
- 42. Kurtze N, Rangul V, Hustvedt BE, Flanders WD. Reliability and validity of self-reported physical activity in the Nord-Trondelag Health Study: HUNT 1. Scand J Public Health. 2008;36:52–61. pmid:18426785
- 43. Moholdt T, Wisloff U, Lydersen S, Nauman J. Current physical activity guidelines for health are insufficient to mitigate long-term weight gain: more data in the fitness versus fatness debate (the HUNT study, Norway). Br J Sports Med. 2014;48:1489–96. pmid:24782484
- 44. Gronhaug G, Osteras N, Hagen KB. Quality of hip and knee osteoarthritis management in primary health care in a Norwegian county: a cross-sectional survey. BMC Health Serv Res. 2014;14:598. pmid:25422042
- 45. Peter WF, van der Wees PJ, Verhoef J, de Jong Z, van Bodegom-Vos L, Hilberdink WK, et al. Postgraduate education to increase adherence to a Dutch physiotherapy practice guideline for hip and knee OA: a randomized controlled trial. Rheumatology (Oxford). 2013;52:368–75.
- 46. Messier SP, Mihalko SL, Legault C, Miller GD, Nicklas BJ, DeVita P, et al. Effects of intensive diet and exercise on knee joint loads, inflammation, and clinical outcomes among overweight and obese adults with knee osteoarthritis: the IDEA randomized clinical trial. JAMA. 2013;310:1263–73. pmid:24065013
- 47. Hubertsson J, Petersson IF, Thorstensson CA, Englund M. Risk of sick leave and disability pension in working-age women and men with knee osteoarthritis. Ann Rheum Dis. 2013;72:401–5. pmid:22679305
- 48. Salmon JH, Rat AC, Achit H, Ngueyon-Sime W, Gard C, Guillemin F, et al. Health resource use and costs of symptomatic knee and/or hip osteoarthritis. Osteoarthritis Cartilage. 2019;27:1011–7. pmid:30922982