Background

Musculoskeletal disorders (MSDs) are one of the most common health conditions experienced worldwide and are costly to the healthcare system [1], with one in six adults (15.6%) reporting chronic MSDs [2]. The most common chronic MSDs are osteoarthritis (OA), neck pain (NP), and low back pain (LBP). The World Health Organization estimates that 10% of individuals 60 years or older have significant clinical problems (e.g., functional limitation) that are attributed to OA [3]. Likewise, over 80% of the population experience LBP and NP during their lifetime [4,5,6]. Importantly, MSDs are associated with a high economic burden globally [7,8,9]. In 2010, the Public Health Agency of Canada highlighted that MSDs are associated with a higher economic burden than any other group of diseases, estimated at $37 billion [10].

Rehabilitation practitioners such as physiotherapists (PTs), occupational therapists (OTs), chiropractors (DCs) deliver care to over 11 million Canadians with MSDs, with estimates pointing to an increase to 15 million patients seeking care by 2031 [11]. Despite the availability of clinical practice guidelines to inform practice in rehabilitation [12,13,14,15,16], substantial research-practice gaps among rehabilitation clinicians persist [17,18,19,20]. The lack of adherence to recommended care can lead to negative effects on the health outcomes of individuals and communities and lead to inefficient use of limited health care resources [21, 22]. Knowledge Translation (KT) aims to promote the use of research evidence in healthcare systems [23]. Although, KT researchers evaluated the relative effectiveness of different KT interventions in changing healthcare professionals’ practice behaviour [24, 25], there is uncertainty regarding which KT interventions are likely to be effective in increasing the use of research findings [24, 26]. As KT interventions are multilevel interventions (e.g., professional, patient, or organizational level), other factors that may have impacted the success of the intervention [27, 28]. Low participant enrollment, adherence and retention rates (i.e., maintaining clinicians’ engagement throughout the course and up to the end of a trial) are major factors that can contribute to the success or failure of KT intervention [29] and influence the estimation of the effectiveness of any intervention [30] in healthcare research [31]. Difficulties in participants’ enrollment may lead to untimely delays in study initiation, financial burden, and failure to meet enrollment goals (i.e., underpowered trials) resulting in very expensive trials [31]. Thus, maximizing enrollment, adherence, and retention rates requires thoughtful planning, and specific strategies embedded in the trial process, and careful monitoring [31]. Assessing those rates may help researchers develop more appealing KT interventions that practitioners will more easily accept and sustain into their everyday practices, and improve the design of future trials, and consequently, increase their validity and generalizability [32]. Although successful enrollment and retention strategies have been described in clinical trials focusing on adults [33] and children [34], we are not aware of prior reviews having systematically assessed the enrollment, adherence, and retention rates in KT studies. This systematic review aimed to 1) estimate the enrollment, adherence, and retention rates of KT interventions targeting rehabilitation practitioners in charge of patients with MSDs, and 2) identify factors likely to impact on the enrollment, adherence, and retention rates.

Methods

Searches

A search strategy was developed in collaboration with a health-sciences librarian to ensure that we captured the maximum number of studies in rehabilitation sciences (Supplementary Material 1: Appendix 1). The search strategy was adapted from a previous review from this team [35], using subject headings (MeSH), keywords, and abstract/text words for MSDs, KT, and rehabilitation, and their synonyms. We searched published literature in scientific journals in the following five databases from the inception to October 2022: OVID MEDLINE, EMBASE, PsycINFO, CINAHL, and Cochrane databases, in English language. All identified citations were exported into EndNote after removing duplicates.

Study inclusion and exclusion criteria

Three independent reviewers (DG, OE, KM) screened the titles and abstracts of studies identified by applying the eligibility criteria. The same reviewers then independently assessed full-text reports of potentially eligible studies. Reviewers met periodically to resolve disagreements and reach a consensus on the eligibility of studies at all stages.

Inclusion criteria

Study design

As recommended by the Effective Practice and Organisation of Care (EPOC) systematic reviews [36], we included the following study designs: Randomized Clinical Trials (RCTs), cluster randomized controlled trials, non-randomized controlled trials (NRCTs), or before-and-after studies.

Participants

All types of rehabilitation practitioners (e.g., PTs, OTs, Osteopaths, or DCs) managing patients with MSDs.

Intervention

KT interventions directed toward rehabilitation practitioners were selected according to the Expert Recommendations for Implementing Change (ERIC) classification [37], which provide comprehensive catalogue of KT interventions that can be used in isolation or combination in implementation research and practice (Supplementary Material 1: Appendix 2).

Outcomes

Three feasibility measures were considered: enrollment rate (defined as the proportion of participants who accepted to participate in the study over all eligible participants invited for the study, considering that the number of eligible practitioners was calculated after excluding individuals who did not meet the study’s inclusion criteria), adherence rate (defined as the proportion of participants who completed the intervention over all participants who were assigned to the intervention group), and retention rate (defined as the proportion of participants who completed through to the first follow-up point over the participants who started the study in each group “intervention group [IG] or control group [CG]”).

Exclusion criteria

Studies failing to report follow-up data for the participating practitioners were excluded. Studies published in abstract form, as conference proceedings, or protocol forms were also excluded.

Data extraction strategy

A structured extraction sheet was created to collect and extract data from the eligible studies. We extracted data related to study characteristics (i.e., year of publication, country, study design, study duration, number of the study group, study duration, number of follow-up points, number of outcomes); KT interventions (i.e., type of KT intervention based on ERIC classification, number and duration of the KT interventions, mode of delivery, intensity of the intervention); practitioners (i.e., age, profession and types of MSDs managed; number of practitioners who were approached, eligible to participate, excluded, refused to participate, accepted to participate, and assigned to each study group “if applicable”; number of practitioners who adhered to the KT intervention, and who participated at least in the first follow-up point; and reasons for refusal to participate, and for dropping-out in each study group. The data were extracted by the principal investigator (DG) and reviewed by a second reviewer (OE).

Study quality assessment

Quality assessment was not considered since the focus of this review was on enrollment, adherence, retention rates, and not the effectiveness and effect size of KT interventions likely to be impacted by criteria such as randomization process and missing data [38].

Data synthesis and presentation

Statistical analyses were performed using the Statistical Analysis Systems (SAS version 9.3) [39] guided by two KT experts (AB and SA). Descriptive statistics were conducted to describe variables that possibly affected enrollment, adherence, and retention rates as proportion (%). Moreover, variables were categorized according to their nature and the frequency of data for study characteristics and practitioners- and intervention-related variables described above.

Meta-regression weighted by the sample size [38] of included studies was used to estimate the overall enrollment, adherence, and retention rates. Meta-regression was used since the outcome had a specific range (0–100%). The study, intervention, and practitioners-related variables aforementioned were used as the factors (i.e., predictors) of the enrollment, adherence, and retention rates. The overall enrollment and retention rates were calculated for intervention and control groups, and the overall adherence rate was estimated for intervention groups only. A meta-regression model was used to assess the correlation between the potential variables and enrollment, adherence, and retention rates. These three rates were treated as continuous variables. A p-value of less than 0.05 was considered as statistically significant.

Results

Descriptive statistics

The search strategy yielded 6088 records after duplicate removal. Screening for titles and abstracts identified 105 potentially eligible articles of which 33 studies met our inclusion criteria [40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69] (See Fig. 1. PRISMA flowchart). Table 1 provides a description of the study- practitioners- and KT intervention characteristics of the included studies.

Fig. 1
figure 1

PRISMA flow diagram representing the process of study selection

Full size image
Table 1 Characteristics of the included studies
Full size table

Studies’ characteristics

The included studies were published between 2004 and 2022 in Europe (61%, n = 20), North America (27%, n = 9), Australia (9%, n = 3) or elsewhere (3%, n = 1). Most of the studies (61%, n = 20) were controlled trials, such as RCT (n = 8) [40, 43, 46, 49, 53, 55, 56, 68], cluster-RCT (n = 5) [42, 58, 59, 61, 67], pragmatic RCT (n = 1) [45], prospective controlled trials (CT) (n = 1) [52], single-blind RCT (n = 1) [47], and non-RCT (n = 1) [44], mixed methods trial (n = 1) [70, 71], and before-and-after control study (n = 1) [41]. The remaining studies were before-and-after studies (39%, n = 13) [48, 50, 51, 54, 57, 60, 62,63,64,65,66, 69, 72]. Study duration ranged from 2 to 26 months. The median of the follow-up points was one follow-up. More than half (57%, n = 19) [40,41,42, 44, 45, 47, 49, 51,52,53,54, 57, 58, 61, 65, 67,68,69, 71] of the included studies had one follow-up point or two follow-up points (30%, n = 10) [43, 46, 50, 55, 56, 59, 60, 64, 66, 72], while few studies had more than two follow-up points (12%, n = 4) [48, 62, 63, 70].

Practitioners’ characteristics

The practitioners’ mean age was provided in 21 studies [40,41,42,43, 45, 47,48,49,50, 53,54,55,56,57, 60,61,62, 65, 67, 70, 72]; ranging from 28 to 47.5 years for intervention groups (x̅ = 40 years; SD = 5.1) and 33 to 54.3 years for control groups (x̅ = 43 years; SD = 4.7). Most of the studies targeted PTs (70%, n = 23) [40, 42,43,44, 47, 48, 50,51,52,53,54,55,56, 58,59,60,61,62, 65, 66, 68, 69, 72]. The remaining studies targeted a mixed types of practitioners (PTs, OTs, DCs) (13%, n = 4) [49, 57, 63, 64], DCs (12%, n = 4) [41, 46, 67, 70], OTs (3%, n = 1) [45], or osteopaths (3%, n = 1) [71]. The types of disorders includes back pain (45%, n = 15) [41,42,43, 48,49,50, 60,61,62,63, 65, 68, 70,71,72], MSKs in general (27%, n = 9) [44,45,46, 51,52,53,54, 59, 69], NP (18%, n = 6) [40, 47, 57, 58, 66, 67], and OA (6%, n = 2) [55, 56], and rheumatoid arthritis (3%, n = 1) [64].

KT interventions’ characteristics

Based on the ERIC classification [37], the predominant type of KT interventions were educational meetings (97%, n = 32) [40,41,42,43,44,45,46,47,48, 50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72], distribution of educational materials (58%, n = 19) [40,41,42,43,44,45, 48, 49, 51, 52, 54,55,56, 58, 60, 61, 63, 64, 70], audit and feedback (30%, n = 10) [40, 42, 45, 48, 53, 54, 56, 60,61,62], local opinion leaders (24%, n = 8) [41, 51, 56,57,58,59, 62, 70], reminders (21%, n = 7) [44,45,46, 52, 53, 62, 70], facilitation (15%, n = 5) [40, 47, 55, 61, 68], educational outreach visits (12%, n = 4) [40, 41, 58, 62], ongoing consultation (12%, n = 4) [40, 44, 52, 69], and developing centralize technical assistance (12%, n = 4) [44, 52, 60, 72]. Developing educational materials [44, 64, 69], and creating a learning collaborative [45, 60, 69] were used in equal frequency (9%, n = 3). Providing clinical supervision [40, 60] and intervening with patients to enhance adherence [44, 64] (6%, n = 2), preparing patients to be active participants [52], local consensus discussions [60], conducting ongoing training [69], and using mass media [41] (3%, n = 1) were uncommon (Table 2). An equal proportion of studies combined two to three interventions (42%, n = 14) [42, 43, 46,47,48, 51, 53,54,55, 57, 59, 63, 68, 72] or more than three interventions (39%, n = 13) [40, 41, 44, 45, 52, 56, 58, 60,61,62, 64, 69, 70]. Few studies employed a single intervention (18%, n = 6) [49, 50, 65,66,67, 71]. The majority of the studies delivered the KT intervention in person (76%, n = 25) [40,41,42,43,44,45,46,47,48, 50,51,52,53,54,55,56,57,58,59,60,61,62, 65, 66, 72] or online (21%, n = 7) [63, 64, 67,68,69,70,71]. Only one study used postal dissemination (3%, n = 1) [49].

Table 2 Types of KT intervention classified according to ERIC classification
Full size table

The other groups (named as control groups) either received no intervention (n = 6) [41, 44, 49, 52, 55, 71], or educational interventions (n = 5) (e.g., external coach [61], opinion leader, audit and feedback [56], practicing skills [45], peer assessment approach [53], and outreach visit, clinical supervision, ongoing consultation [40]), educational materials only (n = 4) [42, 58, 67, 70], less frequent educational sessions (n = 3) [43, 47, 68], or a similar intervention but on different topics (n = 2) [46, 59].

Enrollment, adherence, and retention data

  1. 1)

    Enrollment rate: Overall, 7146 eligible practitioners from 27 KT studies were invited to participate, of whom 5880 agreed to participate [40, 43,44,45,46,47,48,49, 51,52,53, 55, 56, 58,59,60,61,62, 64,65,66,67,68, 70,71,72]. The overall unweighted enrollment rate was 84% and the weighted enrollment rate was 82%, ranging from 32% [51] to 100% in eleven studies [40, 43, 49, 59,60,61, 65, 67, 68, 71, 72], including 8 controlled trials (CTs) [40, 43, 49, 59, 61, 67, 68, 71] and 3 before-and-after studies [60, 65, 72].

  2. 2)

    Adherence rate: The included studies reported the number of practitioners who participated in the educational meetings only, not for every type of KT interventions separately. Thus, the adherence rate was calculated for educational meetings only, however, the calculated adherence rate can well exemplify the adherence rate for the other associated KT interventions in each study as educational meetings were mostly delivered concurrently with other intervention such as distribution of education materials, local opinion leaders, facilitation, audit and feedback, etc. Overall, 4775 practitioners were assigned to attend the educational meetings; 3537 did in fact attend as per protocol, with an overall unweighted adherence rate of 88% and a weighted adherence rate of 74% (73% for before-and-after studies [48, 51, 60, 62,63,64,65,66, 69, 72] and 78% for controlled trials [40, 43,44,45,46,47, 49, 52, 53, 55, 56, 58, 59, 61, 67, 68, 70, 71]). A 100% adherence was reached in 14 studies [40, 43, 47, 48, 50, 51, 53, 54, 57,58,59, 65, 70, 72], of which half were controlled trials [40, 43, 47, 53, 58, 59, 70]. The length of educational meetings ranged from 1 to 8 h. However, the lowest adherence (44%) was observed in a before-and-after study targeting individuals with LBP [63]. The educational sessions was available for online browsing for up to 3 weeks.

  3. 3)

    Retention rate: The number of practitioners who completed the first follow-up point was reported in 28 studies (before-and-after studies [n = 9] [51, 54, 57, 62, 63, 65, 66, 69, 72] and controlled trials [n = 19] [40, 42,43,44,45,46,47, 49, 52, 53, 55, 56, 58, 59, 61, 67, 68, 70, 71]). For before-and after studies, 1031 practitioners out of 2094 completed the first follow-up point, with a retention rate of 49% (range: 36% to 100%). For controlled trials, the retention rate was 80% for interventions groups (1672 / 2085) and 81% for control groups (1494 / 1838), with an overall rate of 81% for both groups. The overall unweighted retention rate across all studies was 85% and the weighted retention rate was 65%, ranging from 36% in a before-and-after study targeting mixed types of practitioners [63] to 100% in 7 studies, including 3 CTs [40, 43, 47] and 4 before-and-after studies [54, 65, 69, 72]. Figures 2, 3, and 4 present forest plots for the enrollment, adherence, and retention rates.

Fig. 2
figure 2

Forest plot of the enrollment rates

Full size image
Fig. 3
figure 3

Forest plot of the adherence rates

Full size image
Fig. 4
figure 4

Forest plot of the retention rates

Full size image

Reasons for refusal to participate

Only 8 studies [42, 45, 48, 55, 56, 62, 67,68,69] (27%) reported on the reasons for refusing to participate, namely lack of time (n = 4) [45, 48, 56, 68], lack of interest (n = 6) [42, 45, 48, 56, 62, 67], having a health condition preventing them from participating (n = 3) [45, 55, 67], or unavailability during the study time (n = 2) [56, 69]. Other reasons reported only once were distance from the intervention site [56], holidays [56], other priorities [56], work obligations [68], work schedule conflict [55, 68], moved out of country [67], fail to submit consent [45], retired [67], joining politics [67], and invalid address [67].

Reasons for drop outs

Thirteen studies [42, 44, 45, 52, 53, 56, 57, 59, 61, 62, 67,68,69] (43%) reported the reasons for participants’ drop out throughout the course of the studies after consenting to participate, including: personal life changes (n = 9) (i.e., change of job, retirement, maternal leave or pregnancy) [42, 45, 53, 56, 57, 59, 62, 67, 68], lack of time (n = 5) [42, 53, 56, 57, 67], unknown reason (n = 3) [42, 56, 61], work conflict (n = 3) [45, 56, 69], unable to contact the participant (n = 3) [44, 52, 57], lack of interest (n = 2) [42, 45], lack of compliance (n = 2) [62, 69], being out of town (n = 1) [67], and transportation problem (n = 1) [45].

Factors influencing enrollment, adherence, and retention rates

In general, the meta-regression showed that all the aforementioned factors (i.e., variables related to studies, interventions, and practitioners) significantly affected the rates of enrollment, adherence, and retention. All comparisons were significant at P-value < 0.0001 (Table 3).

  1. 1)

    Factors influencing enrollment rate: For study-related factors, the enrollment rate was 12% higher in Europe compared to North America. Further, before-and-after studies had about 12% lower enrollment rate compared with controlled trials; implementing a study with more than one study group was associated with a 12% higher enrollment rate, while having more than one follow-up point was associated with an 11% greater enrollment rate. For practitioners-related factors, enrollment rate was higher for mixed types of MSDs practitioners by 15% compared to delivering intervention to single type of practitioners (e.g., OTs, PTs, DCs). Also, enrollment rate was higher for practitioners managing BP and NP by 33% and 19%, respectively, compared to practitioners who manage mixed type of MSDs. As for KT intervention-related factors, enrolment rate was lower when employing two to three interventions, or more than three interventions (38% and 16%) higher respectively compared to employing single intervention; but 16% higher when delivering KT intervention online (virtually) compared to in-person; 19% higher when delivering educational meeting for more than 4 h; and 29% higher when conducting the educational meetings more than one time.

  2. 2)

    Factors influencing adherence rate: Interestingly, adherence rate for study-related factors was 28% lower in Europe when compared to North America, 13% lower when implementing a study for over 6 months; 11% lower when there was more than one follow-up point; but 27% higher when measuring more than two professional outcomes. For the practitioners-related factors, adherence rate was lower when recruiting mixed types of practitioners, DCs and DOs, OTs, compared to PTs by 32%, 17%, and 5%, respectively. Also, adherence rate was higher for practitioners managing NP by 13%, and lower by 11% for practitioners managing BP, both compared to practitioners who manage mixed type of MSDs. As for the KT intervention-related factors, adherence rate was 22% lower when employing two to three interventions; 32% lower when delivering KT intervention online compared to in-person mode; but 16% greater when implementing a meeting length more than 4 h; 12% higher when conducting the educational meetings more than one time; and 29% higher when delivering the KT intervention for a long period (e.g., 1 month up to 6 months).

  3. 3)

    Factors influencing retention rate: Similarly for study-related factors, retention rate was 17% higher in other countries (i.e., Australia) compared to North America. Retention rate was 31% lower when in before-and-after study compared to controlled trials; 22% lower when implementing a study for over 6 months; but 31% greater when implementing a study in more than one study group. The retention rate was also 28% lower when having more than one follow-up point; and 14% higher when measuring more than two professional outcomes. Concerning the practitioners-related factors, retention rate was 16% higher when recruiting OTs, but lower 15% when recruiting mixed types of MSKs practitioners compared to recruiting PTs. Also, retention rate was lower for practitioners managing BP by 12% compared to practitioners who manage mixed type of MSDs. As for the KT intervention-related factors, retention rate was 31% and 13% lower when employing two to three interventions or more than three interventions vs employing a single intervention, respectively. Compared to in-person mode of delivery, retention rate was also 35% lower for online KT interventions, but 15% higher for postal dissemination. Retention rate was 13% greater when conducting the educational meetings more than one time and 29% higher when delivering the KT intervention for a long period (e.g., 1 month up to 6 months).

Table 3 Meta-regression of factors influencing rates of enrollment, adherence and retention
Full size table

Discussion

To our knowledge, this is the first review estimating the enrolment, adherence, and retention rates of KT interventions targeting rehabilitation practitioners managing patients with MSDs. Results of the current review showed high enrolment rate (82%) in KT studies, and relatively high adherence (74%) and retention rates (65%) across studies.

This review supported that designing a study with more than one group of practitioners with a controlled arm is associated with higher feasibility rates. Similar findings were reported by Lixin Song et al. [74] when examining the enrollment and retention rates clinical trials of patients with cancer and their caregivers. Studies of shorter period (less than 6 months) with only one follow-up point with multiple outcome measures were associated with higher feasibility rates. These findings are possibly explained by the difficulties for practitioners to commit to their regular work schedule over a long period, thereby limiting their ability to report outcomes over multiple follow-up points.

We uncovered a number of appealing intervention-related features for rehabilitation practitioners that seems to promote all three feasibility measures. First, employing a single intervention for a short period of time (1 month up to 6 months) is significantly associated with the higher rates. Systematic reviews of KT studies have suggested that single active KT interventions may be as effective as multi-component intervention in changing practice [26, 75,76,77]. The complexity of interventions may dampen the key messages and diminish the ability of practitioners to digest the presented information [78]. Previous studies reported a higher enrollment rate when recruiting participants for studies with ≥ 4 months intervention duration [74, 79]. Second, implementing a long educational meeting (more than 4 h) for more than one time is associated with higher rates; this possibly means that practitioners prefer for instance to concentrate on a full day workshop offered multiple times (i.e., long-term engagement), instead of having several short meetings during their busy working day when being exposed to the content of the KT intervention. Lastly, delivering KT intervention virtually or in-person mode provided mixed results with virtual mode being associated with a higher enrolment rate, whereas in-person mode was associated with greater adherence and retention rates; these findings support that online interventions could be considered as time and effort saving modes of delivery. Feasibility rates don’t seem to be harmonically affected by the practitioners’ profession or the type of MSDs they manage. Considering those intervention-related factors may secure higher practitioners’ involvement in the KT studies for longer duration.

Strengths and limitations

This systematic review followed rigorous methodology, including a comprehensive search strategy developed in collaboration with a medical science librarian, the use of multiple electronic databases. However, this review is not without limitations. First, several studies failed to report on the number of practitioners who were eligible to participate in the study. Second, other variables that could be influential, such as practitioners’ educational backgrounds and practitioners’ beliefs in KT interventions, could not be included in the analyses as sufficient information on these variables was not available. Third, assessing the impact of each type of KT interventions separately on the feasibility rates was not possible due to the overlapping of the KT interventions in each study. Fourth, the included studies fail to reported the number of participants received each KT intervention separately. Finally, this review was restricted to KT interventions targeting MSDs rehabilitation practitioners only. Thus results may not apply to other healthcare disciplines.

Conclusion

This systematic review identified 33 studies employing KT interventions to promote the uptake of research evidence by MSDs rehabilitation practitioners, including PTs, OTs, DCs, and osteopaths. Findings showed that enrolment, adherence, and retention rates ranged between 65 and 82% across the KT studies. Moreover, this review showed that single intense (e.g., high frequency, short duration) KT intervention was more appealing for practitioners. Interventions which require less effort and less commitment, and which save participants’ time have higher feasibility rates. KT researchers should consider the time required from healthcare practitioners to participate in a KT studies to maximize the feasibility rates, and consequently increase the generalizability of their findings.