Impact of clinical supervision on healthcare organisational outcomes: A mixed methods systematic review

Objective To review the impact of clinical supervision of post-registration/qualification healthcare professionals on healthcare organisational outcomes. Background Clinical supervision is a professional support mechanism that benefits patients, healthcare professionals and healthcare organisations. Whilst evidence is growing on the impact of clinical supervision on patient and healthcare professional outcomes, the evidence base for the impact of clinical supervision on organisational outcomes remains weak. Methods This review used a convergent segregated approach to synthesise and integrate quantitative and qualitative research findings, as per the Joanna Briggs Institute’s recommendations for mixed methods systematic reviews. Databases searched included CINAHL, Embase, PubMed, PschINFO, and Scopus. Whilst a narrative synthesis was performed to present the findings of the quantitative and qualitative studies, the evidence from both quantitative and qualitative studies was subsequently integrated for a combined presentation. The review followed the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses. Results Thirty-two studies including 27 quantitative, two qualitative and three mixed methods studies, were included in the review. The results of the quantitative analysis showed that effective clinical supervision was associated with lower burnout and greater staff retention, and effective supervisor was associated with lower burnout and greater job satisfaction. Qualitative findings showed that healthcare professionals believed that adequate clinical supervision could mitigate the risk of burnout, facilitate staff retention, and improve the work environment, while inadequate clinical supervision can lead to stress and burnout. The evidence from quantitative and qualitative studies were complementary of each other. Conclusion Clinical supervision can have a variable effect on healthcare organisational outcomes. The direction of this effect appears to be influenced by the effectiveness of both the clinical supervision provided and that of the clinical supervisor. This highlights the need for organisations to invest in high quality supervision practices if maximal gains from clinical supervision are to be attained.


Background
Clinical supervision is a professional support mechanism that benefits patients, healthcare professionals and healthcare organisations. Whilst evidence is growing on the impact of clinical supervision on patient and healthcare professional outcomes, the evidence base for the impact of clinical supervision on organisational outcomes remains weak.

Methods
This review used a convergent segregated approach to synthesise and integrate quantitative and qualitative research findings, as per the Joanna Briggs Institute's recommendations for mixed methods systematic reviews. Databases searched included CINAHL, Embase, PubMed, PschINFO, and Scopus. Whilst a narrative synthesis was performed to present the findings of the quantitative and qualitative studies, the evidence from both quantitative and qualitative studies was subsequently integrated for a combined presentation. The review followed the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Results
Thirty-two studies including 27 quantitative, two qualitative and three mixed methods studies, were included in the review. The results of the quantitative analysis showed that effective clinical supervision was associated with lower burnout and greater staff retention, and effective supervisor was associated with lower burnout and greater job satisfaction.

Introduction
Clinical supervision is widely practiced in health and social care professions across the globe owing to its beneficial effects to patients, health professionals and organisations [1,2]. Operationally, clinical supervision, for post-qualification health professionals, is viewed as a process that provides quarantined time and an opportunity to further develop the supervisee's skills and knowledge, within the context of an ongoing professional relationship, usually with an experienced practitioner (one-to-one supervision), or with peers (peer group supervision). The aim of clinical supervision is for the supervisee to engage in guided reflection on current practice in ways designed to develop and enhance that practice in the future [1,2]. This type of supervision involves reflective thinking, and discussion regarding professional development issues, caseload, clinical issues, and staff interpersonal issues. Issues in clinical supervision definition and terminologies are widely prevalent [2]. In this review, the following definition of clinical supervision has been adopted: "The formal provision, by approved supervisors, of relationship-based education and training that is work-focused, and which manages, supports, develops and evaluates the work of colleague/s" [1].
Whilst efforts are growing to strengthen the evidence for clinical supervision, there is also criticism about a vast majority of evidence on supervision, as being proof by association or tentative [3]. While there is a growing evidence base for the impact of clinical supervision on patient outcomes such as reduced risk of mortality, reduced risk of complications and more effective care [4][5][6][7], and health professional outcomes such as being better supported in their roles [8], there remains a need to systematically review the evidence for the impact of clinical supervision of post-qualification health professionals, on organisational outcomes, to further strengthen the evidence base on clinical supervision.
Determining the impact of clinical supervision on healthcare organisations, however, is difficult given the challenges in defining organisational outcomes and the overlapping nature of patient, health professional and organisational outcomes. For example, improved patient outcomes (e.g. improved morbidity and mortality) can satisfy multiple targets for healthcare organisations, as can health professional outcomes (e.g. reduction in stress and burnout), which can reduce staff sick leave, a usual key performance indicator for organisations. In determining the organisational outcomes of interest for this review, we undertook a scan of the broader literature. A recent systematic review of leadership styles and outcome patterns for the nursing workforce and work environment, grouped the outcomes into six categories: staff satisfaction and job factors, staff relationships with work, staff health and wellbeing, relations among staff, organisational environment factors and productivity and effectiveness [9]. Another systematic review on the relationship between governance mechanisms in healthcare and health workforce outcomes considered staff turnover and job satisfaction [10]. Other organisational outcomes cited in the clinical supervision literature include improved teamwork [11] and job satisfaction [12]. In considering all this, organisational outcomes in the current review will reflect the well-being of health professionals resulting from clinical supervision, that lead to better outcomes for the organisations such as recruitment and retention, intent-to-stay, intent-to-leave, job satisfaction and quality of work life, burnout and absenteeism. Furthermore, despite the benefits of supervision, to date, no review has explored health professionals' perspectives of, and the impact from, clinical supervision on organisational outcomes.
Therefore, as means of addressing these knowledge gaps, using a mixed methods design, this review aims to answer the following research questions: 1. What are the effects of clinical supervision of healthcare professionals on organisational outcomes?
2. What are healthcare professionals' experiences, views, and opinions regarding clinical supervision as it relates to organisational processes and outcomes?
3. What can be inferred from the qualitative synthesis of healthcare professionals' experiences/ views that can explain the effects of clinical supervision or inform its appropriateness and acceptability for health professionals?

Methods
This systematic review was conducted using Joanna Briggs Institute (JBI) methodology for mixed methods systematic review, specifically the convergent segregated approach to synthesis and integration [13]. The review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline [14] and was based on an a-priori published protocol [15].

Eligibility criteria
The review protocol indicated the inclusion of studies that focused on one-to-one clinical supervision rather than group supervision. However, during the screening of studies, it became apparent that there was a prevalence of studies that investigated both one-to-one and group supervision (which was facilitated by a supervisor, as opposed to peer supervision), and studies that did not specify the type of clinical supervision investigated. Given this challenge, and to reflect the reality of healthcare organisations utilising both these types of supervision regularly, the review team agreed to include any study on clinical supervision, regardless of the type (i.e. one-to-one or group). To be eligible, studies had to meet the following criteria: (1) investigated clinical supervision of qualified or registered health professionals (i.e. clinical supervision of post-qualified health professionals, where they engage in one-to-one or group supervision sessions that happen over a period of time); (2) used qualitative, quantitative or mixed-methods study design; (3) if a quantitative study, examined the effects of clinical supervision on organisational outcomes, such as staff retention and recruitment, intent to stay, intent to leave, job satisfaction and quality of work life, burnout, and absenteeism; (4) if a qualitative study, explored health professionals' experiences, views, or opinions regarding clinical supervision as they relate to organisational outcomes.

Search strategy
As means of avoiding publication and location bias, the search strategy was developed to identify black (commercially published) and grey literature. Search terms were identified based on the key concepts relating to the intervention/phenomenon of interest, i.e. clinical supervision and outcomes of interest, i.e. organisational outcomes. An initial limited search of PubMed and CINAHL was undertaken followed by analysis of text words contained in the title and abstract and the index terms used to describe the articles. The search strategy, including all identified keywords and index terms, was then adapted for each database. The search for published studies was performed from the date of inception until May 2020 in the following databases: CINAHL, Embase, PubMed, PsycINFO, and Scopus. These databases were chosen as they commonly include literature from health disciplines, a combination of discipline specific (e.g. CINAHL includes nursing and allied health literature) and multi-disciplinary (e.g. Scopus) and are routinely used in systematic reviews. The search for grey literature was undertaken in ProQuest Dissertations and Theses, Google Scholar and WorldWideScience.org. Reference lists of relevant studies were reviewed to identify additional publications. The search strategy for each database is shown in S1 Appendix.

Study selection
Following the search, all identified citations were collated and uploaded into EndNote X8.2 (Clarivate Analytics, PA, USA) [16] and duplicates removed. Two reviewers independently screened the titles and abstracts (LL and DS) against the inclusion criteria for the review. Potentially relevant articles were retrieved in full and assessed independently for eligibility by two other reviewers (PM and SK). Disagreements were resolved through discussion and consensus. Studies that did not meet the inclusion criteria were excluded and reasons for their exclusion are provided in S2 Appendix. Abstracts and full text articles did not require translation to another language to determine their eligibility. All full text articles reviewed contained sufficient information to determine their eligibility without the need for further clarification from authors. The PRISMA flow diagram of included studies is available in Fig 1.

Quality assessment
All eligible studies were assessed for methodological quality by two independent reviewers (PM and DS for quantitative studies; PM and LL for qualitative studies using the relevant JBI critical appraisal tools [17]. These tools were chosen as they assist in assessing the trustworthiness, relevance and results of published studies and are widely used. Any disagreements that arose between the reviewers were resolved through discussion. All studies, regardless of the results of their methodological quality, underwent data extraction and synthesis.

Data collection
For the quantitative component, data were extracted from quantitative and mixed methods studies (quantitative component only) and included specific details about the supervisee and supervisor characteristics (sample size, profession), characteristics of the supervision (type, frequency, duration), study design, setting, clinical supervision characteristics, outcomes measured, and results related to the organisational outcomes. For the qualitative component, data were extracted from qualitative and mixed methods studies (qualitative component only) and included specific details about the supervisee and supervisor characteristics (sample size, profession, work experience), study design and methods, setting, and findings which included participants' experiences of clinical supervision as they relate to organisational outcomes. Findings extracted from individual studies consisted of themes or subthemes reported by the authors. These findings were accompanied by a direct quotation representing a participant's voice (i.e. illustration). Findings were also assigned one of three levels of credibility according to the following criteria: (1) unequivocal: findings accompanied by an illustration that is beyond reasonable doubt and therefore not open to challenge, (2) credible: findings accompanied by an illustration lacking clear association with it and therefore open to challenge, and (3) not supported: findings are not supported by data. The review team discussed the data extraction process, established standards and consistencies on how this should occur, and those with quantitative expertise (DS) and qualitative expertise (LL) lead the extraction process, with the primary reviewer (PM) acting as the additional reviewer for validation purposes.

Data synthesis and integration
A convergent segregated approach to synthesis and integration was applied [13]. This involved an initial independent synthesis of the quantitative studies and qualitative studies followed by the integration of findings from such syntheses using configurative analysis.
Quantitative data were analysed descriptively; meta-analyses were deemed not appropriate due to heterogeneity between studies in terms of clinical supervision interventions and participants. Odds ratios (OR) of dichotomous events and standardised mean differences (SMD) for continuous measures were calculated. For experimental studies OR were converted to SMD using an online calculator [18], to assist with interpretation of effect size. For observational analytical studies the correlational coefficient (r) was calculated in addition to OR and SMD. Effect size was determined using the following reference values for SMD: small 0.2, medium 0.5, large 0.8 [19]; OR: small 1.68, medium 3.47, large 6.71 [20]; and r: small 0.1, moderate 0.3, large 0.5 [19].
Qualitative synthesis was conducted using the meta-aggregative approach [21]. Meta-aggregation is aligned with the philosophy of pragmatism, focusing on the practicality and usability of the synthesised findings and generation of statements that are useful for informing actions in clinical practice [21]. This involved assembling and aggregating the extracted findings from individual studies, based on similarity in meaning, to generate a set of statements (i.e. categories) that represented that aggregation. These categories were then subjected to meta-synthesis to produce a set of synthesised findings. The development of categories and synthesised findings was conducted via a consensus process between the reviewers (LL and PM).
The findings of each single method synthesis were juxtaposed and examined for convergence/divergence and complementarity. To explore the relationship across individual syntheses, the findings were reviewed to determine whether they were supportive or contradictory and identify which aspects of the quantitative evidence were not explored in the qualitative studies and vice-versa. The clinical supervision interventions which had been investigated in the quantitative studies were further analysed in line with the experiences of participants in the qualitative studies to explain the impact of clinical supervision on the different organisational outcomes. However, due to the heterogeneity of the quantitative studies and the lack of wellconducted trials, and the limited qualitative studies, no clear cause and effect relationships can be determined, nor in-depth analysis can be made to explain the impact of clinical supervision.

Results
The database search yielded 1266 records. Eighty-five articles were retrieved for full text review following application of the eligibility criteria to title and abstract. Thirty-four fulfilled the inclusion criteria when applied to full texts. Three of these articles were duplicate publications, resulting in a yield of 31 studies. One article was identified through pearling of references in the included studies; hence the final yield was 32 studies (Fig 1).

Methodological quality
The predominant methodological risk of bias for analytical cross-sectional cohort studies (n = 14) was the absence of strategies to deal with confounding factors [36,39,41,45,46,48,52]. For quasi-experimental studies (n = 14) it was unclear if participants received similar support interventions other than clinical supervision in 12 studies [23-29, 31, 33-35, 51], outcome measurement was not performed both pre and post intervention (i.e. multiple time points) in nine studies [24, 27-29, 31, 33-35, 51], and it was unclear if participants were similar at baseline in seven studies [24, 27-29, 33, 34, 51]. The single randomised controlled trial [22] only met five of the 13 items; notably the method of randomisation was unclear and there was no between group statistical comparison. JBI Critical Appraisal Checklists can be found in S1-S3 Tables.
The methodological quality of the five qualitative studies (including the qualitative component of mixed methods studies) was generally high. Two studies [51,52] scored 10 out of 10, while two other studies [49,50] scored eight out of 10, failing to account for the potential influence of the researcher on the research findings. One study [53] did not demonstrate congruity between their stated philosophical perspective and the research methodology used, nor was there congruence between their research methodology and their research question/objectives, methods of data collection and analysis and interpretation of results. The JBI Critical Appraisal Checklist can be found in S4 Table. Impact of clinical supervision on organisational outcomes (quantitative findings) 1. Clinical supervision compared to control. Eleven studies, including 2,965 participants, evaluated the effect of clinical supervision on organisational outcomes by comparison to a control group that did not receive clinical supervision [22, 24, 25, 27-29, 31, 33-35, 51]. Eight studies included nursing professionals [22,24,25,31,[33][34][35]51], one study included social work/psychology/counselling professionals [27], two studies included other allied health professionals [28,29] and two studies included medical professionals [22,33]. While individual studies found clinical supervision had a positive effect on organisational outcomes, there were variable results across studies for burnout (six studies, n = 776 participants) (Fig 2A-2D), job satisfaction (four studies, n = 2,020 participants), well-being (four studies, n = 444 participants), and workplace environment (five studies, n = 783 participants). Notably, a single randomised controlled trial (n = 30 participants) found that clinical supervision had a large effect on burnout ( Fig 2D) and well-being [22]. Results from individual studies can be found in S5 Table. 2. Clinical supervision compared to within-group historical control (pre/post implementation). Six studies, including 178 participants, evaluated the effect of clinical supervision on organisational outcomes by comparing post-implementation with pre-implementation [22,23,25,26,30,32]. Six studies included nursing professionals [22,23,25,26,30,32], one study included social work/psychology/counselling professionals [32] and one study included medical professionals [22]. While individual studies found clinical supervision had a positive effect on organisational outcomes, there were variable results across studies for burnout (four studies, n = 122 participants) (Fig 3A-3D), job satisfaction (four studies, n = 114 participants), well-being (five studies, n = 144 participants), and workplace environment (three studies, n = 95 participants). Results from individual studies can be found in S6 Table. 3. Association between effective clinical supervision and organisational outcomes. Eight studies, including 1,376 participants, investigated the association between effective clinical supervision and organisational outcomes [37,[38][39][40][41][42][43]52]. Five studies included nursing professionals [37,[39][40][41][42], three studies included social work/psychology/counselling professionals [38,43,52] and one study included other allied health professions [52]. There was preliminary evidence to suggest that effectiveness of clinical supervision may be negatively associated with burnout and positively associated with job retention ( Table 2). The association between effective clinical supervision and job satisfaction was unclear.
Synthesis of five studies [37,39,41,42,52], including 1,046 participants, indicated that effectiveness of clinical supervision may be negatively associated with emotional exhaustion and depersonalisation, but not associated with personal accomplishment. Three studies found a small to moderate association with emotional exhaustion [39,42,52] and four studies found small association with depersonalisation [37,39,41,42].
Synthesis of two studies [40,52], including 152 participants, indicated that effectiveness of clinical supervision may be positively associated with job retention. Both studies found a moderate association with job retention.
Synthesis of three studies [41][42][43], including 836 participants, indicated that the association between effectiveness of clinical supervision and job satisfaction was unclear. Two studies [41,42] found a small positive association and one study [43] found a small negative association. Results from individual studies are available in S7 Table. 4. Association between effective supervisor and organisational outcomes. Eight studies, including 1,600 participants, investigated the association between effectiveness of the supervisor and organisational outcomes [32,36,39,41,43,45,46,48]. Five studies included nursing professionals [32,39,41,46,48], seven studies included social work/psychology/counselling professionals [32,36,43,45,46,48,52] and one study included medical professionals [48]. There was preliminary evidence to suggest that an effective supervisor may be negatively associated with burnout, and positively associated with job satisfaction ( Table 3).
Synthesis of three studies, [39,41,48] including 901 participants, indicated that an effective supervisor may be negatively associated with depersonalisation but not associated with personal accomplishment. Two studies found a small association with depersonalisation [39,48]. The association between an effective supervisor and emotional exhaustion was unclear, with two studies finding a small negative association [39,48] and one study finding a small positive association [41].
Synthesis of five studies [32,41,43,45,46], including 1128 participants, indicated that an effective supervisor may be positively associated with job satisfaction. Studies found a small to large association with job satisfaction. Results from individual studies are available in S8 Table.

Healthcare professionals' experiences of clinical supervision as it relates to organizational processes and outcomes (qualitative findings)
Five studies, including two qualitative [49,50] and three mixed methods studies [51][52][53], explored the experiences of healthcare professional supervisees on clinical supervision as it relates to organisational outcomes. A total of 16 findings and their illustrations were extracted. Of the 16 findings, 14 were unequivocal and two were credible. The 16 findings were organised into four categories which were further deduced to two synthesised findings. Table 4 shows a summary of the qualitative findings. supported the risk of experiencing burnout or leaving the workplace was less likely. Conversely, health professionals, who felt that their supervision was inadequate, reported that clinical supervision had no positive impact or can lead to stress and burnout if they felt unsupported. This synthesised finding was developed from two categories comprising of seven unequivocal findings.

� Category 1.1 Adequate clinical supervision mitigates the risk of burnout, while inadequate clinical supervision can lead to stress and burnout.
Adequate supervision meant that health professionals experienced the opportunity to debrief challenging events with their supervisor and gain a better understanding of patient interactions which can be stressful, and cause burnout for some staff. However, participants who felt unsupported identified stress and burnout as the negative consequences. This category was supported by four findings: 1. Some respondents felt that inadequate supervision had no impact; however, others identified personal consequences in terms of stress and burnout, feeling unsupported and there being an impact on their work, the ward, and clients.   Positive association for job retention, job satisfaction, and well-being indicates effectiveness of supervision is associated with better outcome. Negative association for burnout indicates effectiveness of supervision is associated with better outcome. https://doi.org/10.1371/journal.pone.0260156.t002

Implementation of effective CS facilitates staff retention and reduces their intention to leave
The supervisor played an active role in encouraging staff to undertake career developing activities.
Participants reported that clinical supervision was a reflection that the health organisation valued their staff. Participants also indicated that supervisors encouraged staff to pursue career developments. These experiences enhanced job satisfaction and reduced staffs' intention to leave the healthcare organisation. This category was supported by three findings:  Synthesised finding 2: Clinical supervision improves the work environment through boosting of staff morale, motivation to work, staff well-being and team relationships. Health professionals indicated that clinical supervision was valuable, led to increased motivation and enthusiasm at work, and provided not only reassurance to staff but also a safe space for improved working relationships. This synthesised finding was developed from two categories comprising of seven unequivocal findings and two credible findings.

� Category 2.1 Clinical supervision enhances team relationships through improved communication.
Participants (ie. midwives) felt that clinical supervision offered an opportunity to enhance their ability to engage in difficult conversations with their team which is key in effective working relationships. This category was supported by two findings: Participants reported that having a clinical supervisor to support them and provide valuable feedback made them believe that they had a place within their organisation, increased their morale and enthusiasm at work, and improved their overall perception of their work environment. This category was supported by six findings: 1. Midwives described feeling more positive about the work environment with an increased desire to 'give back' to the unit. (Credible) 5. Receiving positive feedback was particularly valuable for workers (at the time of data collection) as they were experiencing high uncertainty in many areas including changes to their roles and the focus of the service. Feedback from supervisors provided reassurance, as well as a sense of stability amidst the evolving occupational landscape. (Unequivocal) 6. Supervision increased health professionals' sense of connection to the employing organisation, enabling supervisees to feel that they individually had a place within the organisation and therefore a sense of belonging to something greater than their immediate and often atomized local environment. (Unequivocal).

Integration of quantitative and qualitative evidence
Quantitative and qualitative findings in this review have been largely complementary and supportive of each other, especially on the impact of clinical supervision on burnout, staff wellbeing, job satisfaction, job retention and workplace environment. Burnout. Quantitative findings have provided preliminary evidence that effective clinical supervision and effective supervisor may be negatively associated with burnout. This was also supported by qualitative findings that showed that adequate clinical supervision mitigated the risk of burnout, and that inadequate clinical supervision lead to stress and burnout.
Staff well-being. Quantitative findings from a single randomised controlled trial showed a large effect on reducing burnout and enhancing well-being. Qualitative studies supported this, showing that effective clinical supervision improved staff well-being.
Job satisfaction. Although quantitative evidence from three studies showed that the association between effective clinical supervision and job satisfaction was unclear, evidence from four studies showed a positive association of an effective supervisor with job satisfaction. Qualitative findings supported this showing that effective clinical supervision strengthened team relationships and sense of belonging to the organisation, thereby enhancing job satisfaction. This was particularly true when the supervisor was effective, provided valuable feedback and encouraged staff to pursue career developments.
Job retention. Evidence from two quantitative studies showed a moderate positive association of the effectiveness of clinical supervision with job retention. Similarly, qualitative studies showed that adequate clinical supervision facilitated staff retention.
Workplace environment. Synthesis of quantitative evidence from 11 studies investigating the effect of clinical supervision, and six studies investigating post-implementation of clinical supervision with pre-implementation, showed variable results in regard to its effect on workplace environment. However, qualitative evidence highlighted that effective feedback from supervisors were considered valuable and improved supervisee perceptions of the work environment and their sense of belonging to the organisation.
In summary, both the quantitative and qualitative evidence highlight that effective clinical supervision and effective clinical supervisors may be associated with positive organisational outcomes, whereas, ineffective or inadequate clinical supervision and ineffective supervisors may have a negative impact on the well-being of the supervisee.

Discussion
This systematic review of 32 studies is the first known synthesis of quantitative and qualitative evidence to further our knowledge on the impact from, and experiences of, clinical supervision of post-qualification health professionals, on organisational outcomes. Quantitative findings indicate that clinical supervision can have variable effects on organisational outcomes. The effectiveness of both the clinical supervision and the supervisor appear to influence this effect; effective clinical supervision is associated with lower burnout and greater staff retention, and an effective supervisor is associated with lower burnout and greater job satisfaction. This is supported by the qualitative findings which show that healthcare professionals believe adequate clinical supervision can mitigate the risk of burnout, facilitate staff retention, and improve the work environment, while inadequate clinical supervision can lead to stress and burnout. Overall, qualitative synthesis highlights that the effectiveness of clinical supervision and supervisors can significantly influence the effect of clinical supervision on organisational outcomes.
Effective clinical supervision and effective supervisors may be pre-cursors for the realisation of beneficial effects of clinical supervision by healthcare organisations. This is consistent with a model of clinical supervision, for post-qualification health professionals, proposed by Gonge and Buss [42], where participation in effective clinical supervision (ie. prioritising supervision time) is a pre-requisite to beneficial clinical supervision. While clinical supervision has become increasingly mandated in many healthcare organisations, through standard policies and procedures, the subsequent challenge lies in its effective and consistent implementation and uptake. This can be achieved in several ways. Organisations can adopt/utilise evidence-informed clinical supervision frameworks to guide supervision, such as the one recently developed by Rothwell and colleagues [54]. This review by Rothwell and colleagues, based on evidence from 135 studies, encourages organisations to consider making supervision mandatory to increase the value placed on it, and provide protected time for supervisors and supervisees to engage with it. It also offers several practical strategies such as providing staff with both one-to-one and group supervision options, facilitating a person-centred supervision approach with clear boundaries, tasks, ground rules and record keeping processes, and provision of ongoing training to supervisors and supervisees [54]. Implementation and uptake of clinical supervision can be completed by building a positive organisational culture that supports engagement in and uptake of clinical supervision [54], which could be regularly monitored through routine evaluations. Such evaluations will be critical to identify and respond to what clinical supervision strategies have worked, or not worked, for whom and why. Based on our work in this field, we believe that the organisational context can have an important role, and there is no one-size fits all approach when it comes to supporting the implementation and uptake of clinical supervision within organisations.
Healthcare organisations also need to support clinical supervisors to build and foster positive supervisory relationships with their supervisees. This has commonly been reported to be the single most important factor that influences the effectiveness of clinical supervision [3,11,54], and requires investment of both time and resources. Supervisors and supervisees can also be guided by evidence-informed principles that facilitate effective clinical supervision. For example, Martin and colleagues [11] provide several practical recommendations for supervisors and supervisees to enhance the effectiveness of clinical supervision, such as the development of a supervision contract, undertaking sessions at an optimal length and frequency, utilising different modes including telesupervision, evaluating supervision, and working on skills and abilities such as open communication, flexibility, trust and availability to foster a positive supervisory relationship [11]. Health professionals can be provided with continuing professional development opportunities to upskill in evidence-informed supervision practices [3,55]. There is evidence from a longitudinal, multi-methods study to support the delivery of supervision training in various modes such as videoconference, online and blended modes, thereby catering to those that can't access face-to-face training. In this study, participants knowledge and confidence in the provision of supervision increased after training, which was also sustained at three-months post-training across all the four modes. This success was attributed to the careful design and delivery of training across different modes, which maximised participant access to training [56].
This review found various methodological concerns across many studies reviewed, which is consistent with findings from a recent survey of 20 systematic reviews on clinical supervision reported between 1995 to 2019 [3]. Methodological concerns include predominance of ex post facto, cross-sectional, correlational designs, small sample sizes, over reliance on self-report measures, lack of psychometrically sound supervision measures, and lack of experimental and longitudinal designs [3]. Incomplete provision of information (on clinical supervision parameters) seems to continue to plague supervision research, as again found in the survey of supervision reviews [3], and in the systematic review reported here, making it hard to judge the full merit of the study or replicate it. There is a need for further rigorous high-quality studies in this area that use pluralistic research approaches where experimental investigation, randomisation, and data-driven case studies are used in conjunction with ex post facto, and cross-sectional designs [3]. Studies also need to better define the specifics of the clinical supervision intervention to allow replication and identification of the clinical supervision practices that are, or are not, effective for improving outcomes.

Limitations
The final review deviated from the protocol to also include group supervision, as many studies did not specify the type of supervision investigated. However, group supervision is commonly practiced in healthcare organisations and including these studies in this review likely improves the generalisability of our findings. Although the qualitative studies included were deemed to be of good quality, there were several shortcomings in the methodologies employed by the quantitative studies, especially the lack of randomised trials and absence of strategies to deal with confounding factors in cross-sectional studies. Although there were a variety of healthcare settings and health professionals represented in this review, the majority of included studies were conducted in mental health settings with nursing and/or mental health disciplines (i.e. psychology, counselling, and social work). This may limit the generalisability of the results to other disciplines and indicates the need for further research beyond mental health settings and nursing/mental health disciplines.

Conclusions
Clinical supervision can have a variable effect on healthcare organisational outcomes. This effect appears to be influenced by the effectiveness of both the clinical supervision provided and that of the clinical supervisor. This highlights the need for organisations to invest in high quality supervision practices if they wish to benefit from clinical supervision. Without such investment, there is a risk of policy-practice gaps in this area (i.e. while there may be policies to support clinical supervision in healthcare organisations, in practice it may not be implemented well). Ongoing further research, which grows the evidence base for high quality clinical supervision and helps to unpack the black box of clinical supervision practices that have the most effect on organisational outcomes, is required.