After duplicate removal 1,209 citations were identified from the search strategy. An additional 28 papers were identified through citation screening and 33 papers were included from the previous review (1). Subsequently 1,270 citations were identified for screening. After title and abstract screening 85 papers were sought for retrieval full paper screening. There was 98.24% agreement between the first and second reviewers (k = 0.67). Two papers were unable to be retrieved, and 61 papers were excluded of which 32 papers had an incorrect intervention and 29 papers did not include AHP. This resulted in 22 studies being eligible to be included in this review (See Fig. 1 for PRISMA Flow Diagram (15).
Studies were conducted in seven different countries: five in Canada, five in Australia, four in the United States of America, four in the United Kingdom, two in Germany, one in Spain, and one in Ireland. These studies took place from 1999 to 2021. The studies were conducted in a wide range of clinical areas with seven studies in Oncology, four studies in Musculoskeletal Physiotherapy, three studies in Neurology, three studies in Paediatrics, two studies in Cardiology, two studies in multiple clinical areas and one study in Respiratory. Where studies included a single discipline, the professional disciplines cited were Physiotherapists (n = 8) followed by Occupational Therapists (n = 3), Paramedics (n = 1) and Radiographers (n = 1). The remaining nine studies included only partial sample populations of AHPs. Included within these nine studies was a combination of occupational therapists, physiotherapists, and a speech and language therapist (n = 1); an undefined combination of AHPs and other healthcare professionals (n = 1); and undefined hospital-wide healthcare professionals likely to incorporate AHPs (n = 7). See Table 4 the full study characteristics and Fig. 2 for the relevancy to the protocol’s inclusion criteria of all included studies.
Impact
Thirteen studies reported a specific impact; all of which showed positive findings in relation to healthcare performance, specifically processes of care. Nine studies reported a broad impact, all of which showed positive findings in relation to processes of care (n = 6) and health outcomes (n = 3).
Type of research engagement intervention
Ten studies were judged to meet the protocol’s inclusion criteria for intervention with clear ‘engagement in research’ (5, 16–24). Examples of engagement in research within these studies included:
(a) being directly involved in delivering an intervention within a clinical research study;
(b) hospitals which were involved in the participation of a clinical research; and
(c) self-reports of participation in research that they had participated in research.
Of these, 5 of these studies were judged to meet the protocol’s inclusion criteria for population (5, 16, 17, 22, 23) and of these only 2 papers were judged to meet the protocol’s inclusion criteria for study type to indicate an association between engagement in research and health outcomes (5, 16). These two studies were judged to be of high importance. Both of these studies had mixed-positive (16) and positive (5) outcomes on processes of care with both specific and broad impacts respectively.
A further six studies were judged to meet the broadened inclusion criteria for intervention: research engagement with a mixed scenario of ‘engagement in research’ and ‘engagement with research’ (25–30). Examples of mixed scenarios of research engagement included:
knowledge translation implementation including research activity such as scoping reviews, research implementation, reading research articles, evidence-based learning
- knowledge translation implementation including research activity such as scoping reviews, research implementation, reading research articles, evidence-based learning
- knowledge translation programme including research studies, journal clubs, and critically appraised topics
- participatory action research cycles
- hospitals which were involved in a research network group
- participation in a Partnering for Change stakeholder group to transform service delivery
- participation in a clinical trial quality assurance programme
Of these, five studies met the protocol’s inclusion criteria for population (25, 26, 28–30) and of these only one met the criteria of study type (25) and therefore was evaluated to be of high importance. This study had positive findings in relation to processes of care which had a specific impact.
The remaining six studies were judged to meet the broadened inclusion criteria for intervention that described ‘engagement with research’ (31–36). Examples of engagement with research included:
- knowledge translation toolkit programme
- engagement with a e-learning modules containing evidence base and clinical cases
- engagement with a research facilitator
- hospital implementation of guidelines as part of a clinical trial
- participation in education for actionable knowledge translation
Of these, four met the protocol’s inclusion criteria of population (31, 32, 35, 36) and of these, one met the inclusion criteria of study type (31). This was judged to be of high quality and therefore of high importance due to the high relevance to the inclusion criteria. This study showed positive findings for processes of care with a specific impact.
Studies reporting positive findings
Of the 22 included studies one randomised controlled trial (high quality) (31), one quasi-experimental study (high quality) (25) and one cross-sectional study (low quality) (5) were judged to meet the protocol’s inclusion criteria for both population of interest and study type. However, the degree of research engagement was questionable and was not explicitly described in two of the studies (25, 31). Despite this, these three studies were identified to be of high importance because of the combined relevance to the protocol’s inclusion criteria for study type and overall quality, population, and intervention. All three studies reported a positive impact of research engagement on specific (25, 31) and broad (5) processes of care.
Four mixed methods studies were included and were judged to meet the protocol’s inclusion criteria for the population of interest (16, 17, 26, 32). Only two studies used an intervention which would meet the protocol’s inclusion criteria of engagement in research (16, 17). None of the four studies used a study type incorporated within the protocol’s inclusion criteria; however, these studies were included following the decision to broaden the inclusion criteria (16, 17, 26, 32). Out of these four mixed methods studies, three studies were judged to be of high quality (16, 17, 32) and one was judged to be of low quality (26). Based upon relevancy to the protocol’s inclusion criteria and quality of the studies, three studies were identified to be of high importance. These studies showed both mixed-positive findings with broad impact (26) and mixed-positive findings with specific impact (16, 17) for improving processes of care. The remaining study was evaluated to be of low importance and was classified to have positive findings with broad impact for improving processes of care (32).
Seven of the cross-sectional studies included only partial populations of AHPs, (18–21, 27, 33, 34). Four of these seven cross-sectional studies were judged to meet the protocol’s inclusion criteria for the intervention in that the study made explicit reference to engagement in research (18–21). Six out of the seven studies were graded to be of high quality (18–20, 27, 33, 34) and one study was judged to be of low quality (21). Due to the mixed population of health professionals and variation of the research engagement intervention, the seven studies were judged to be of low importance. Four of the studies had positive findings with regard to whether research engagement had a specific and broad impact and improved processes of care (20, 27, 33, 34). Furthermore, three of the studies had positive findings with regard to whether research engagement had a broad impact and improved health outcomes (18, 19, 21).
Six out of the eight qualitative studies were judged to meet the protocol’s inclusion criteria for the population of interest (22, 23, 28, 29, 35, 36), the remaining two studies having a partial population of AHP (24, 30). Only two of these qualitative studies were judged to make explicit reference to engagement in research (22, 24). Furthermore, the majority of the studies were classified to be of low quality (23, 24, 28, 29, 35, 36). Thus, due to the methodological design and the variation in appropriate intervention, all eight qualitative studies were judged to be of low importance. All eight qualitative papers were positive with regard to whether research engagement improved specific (22, 23, 28–30, 35) and broad processes of care (24, 36).
Mechanisms of the intervention
The most common mechanism (Additional file 2) identified through which healthcare performance improved across the included studies was that AHP research engagement may facilitate a ‘change in human capital’. This is in both enhancement of group and individual behaviour, including more rapid uptake of new treatments and greater likelihood of following clinical guidelines (n = 16) and training/updating staff through research engagement leading to the acquisition and use of new skills, and change in attitudes towards research and research findings (n = 13). Mechanisms of ‘improvements in the processes of care related to conducting a specific trial’ were commonly identified, these were a more rigorous process of defining the standard of care (n = 6) and closer monitoring and support (n = 3).
Mechanisms of improvement were also identified at organisational and clinician levels.
‘Organisational mechanisms’ were identified in seven studies; these were a global category of conducting research to address known issues in the healthcare system, allowing AHPs time to conduct research and thus being an attractive organisation to work for, and conducting research to identify best performance targets and using research in quality improvement. Mechanisms of improvement related to ‘collaborative working between organisations, teams and individuals’ were identified in seven studies. Mechanisms were: linkage and exchange that improves the relevance of research and policy-makers’/managers’/clinicians’ willingness to use it (n = 6); and research networks (n = 2). Finally, mechanisms of improvement which were less commonly identified were ‘action and participatory research’ (n = 2) and ‘changes in the structures of institutions’ (n = 2).