Developing and evaluating clinical leadership interventions for frontline healthcare providers: a review of the literature

Background The importance of clinical leadership in ensuring high quality patient care is emphasized in health systems worldwide. Of particular concern are the high costs to health systems related to clinical litigation settlements. To avoid further cost, healthcare systems particularly in High-Income Countries invest significantly in interventions to develop clinical leadership among frontline healthcare workers at the point of care. In Low-Income Countries however, clinical leadership development is not well established. This review of the literature was conducted towards identifying a model to inform clinical leadership development interventions among frontline healthcare providers, particularly for improved maternal and newborn care. Methods A structural literature review method was used, articles published between 2004 and 2017 were identified from search engines (Google Scholar and EBSCOhost). Additionally, electronic databases (CINHAL, PubMed, Medline, Academic Search Complete, Health Source: Consumer, Health Source: Nursing/Academic, Science Direct and Ovid®), electronic journals, and reference lists of retrieved published articles were also searched. Results Employing pre-selected criteria, 1675 citations were identified. After screening 50 potentially relevant full-text papers for eligibility, 24 papers were excluded because they did not report on developing and evaluating clinical leadership interventions for frontline healthcare providers, 2 papers did not have full text available. Twenty-four papers met the inclusion criteria for review. Interventions for clinical leadership development involved the development of clinical skills, leadership competencies, teamwork, the environment of care and patient care. Work-based learning with experiential teaching techniques is reported as the most effective, to ensure the clinical leadership development of frontline healthcare providers. Conclusions All studies reviewed arose in High-Income settings, demonstrating the need for studies on frontline clinical leadership development in Low-and Middle-Income settings. Clinical leadership development is an on-going process and must target both novice and veteran frontline health care providers. The content of clinical leadership development interventions must encompass a holistic conceptualization of clinical leadership, and should use work-based learning, and team-based approaches, to improve clinical leadership competencies of frontline healthcare providers, and overall service delivery.


Background
Clinical leadership by frontline healthcare providers is a critical part of bedside care [1]. Clinical leadership is recommended for the potential impact on clinical practice and on the clinical care environment, and contributes to safe and quality patient care, and to job satisfaction and retention of frontline healthcare providers [1][2][3][4][5][6]. Frontline healthcare providers are well placed to identify work inefficiencies, motivate other members of the care team to act on patient care, and lead change initiatives to correct problems that arise in the clinical setting. Frontline healthcare providers can also identify inefficiencies related to organizational structures and work flows, and to poor policies and procedures for the delivery of optimal patient care [2,5,[7][8][9][10]. Conversely, poor frontline clinical leadership in the clinical setting has been associated with adverse events and clinical litigation settlements, prompting many healthcare systems, particularly in High-Income Countries (HICs), to invest significantly in interventions that support clinical leadership development [3,11].
However, in Low-and Middle-Income Countries (LMICs), clinical leadership development is not well established. As an example, in South African maternity services, maternal and perinatal deaths have been associated with deficiencies in frontline clinical leadership [12][13][14][15][16][17][18]. Albeit the need for clinical leadership development interventions has been identified, there is little evidence to support the planning, implementation and evaluation of such interventions, particularly among frontline healthcare providers, in LMICs [13][14][15][16][17][18].
Towards identifying a model to inform clinical leadership development among frontline healthcare providers in LMIC, including maternity services in South Africa, a literature review was conducted. The purpose of the literature review was to synthesize published evidence on frontline clinical leadership development and its evaluation and included multiple frontline-care contexts. A database was constructed to extract important dimensions of the clinical leadership development interventions. Further, to synthesize the reported findings on the evaluation of the effectiveness of clinical leadership interventions, Kirkpatrick's evaluation approach was used [19,20]. Kirkpatrick's approach to evaluation comprises four levels, presented as a sequence, and includes evaluating the: (1) Reaction: what participants think and feel about the intervention (2) Learning: the resulting increase in knowledge or skills, and changes in attitude (3) Behaviour: change in practice because of the intervention (4) Result: the final result that occurs as a result of the intervention (e.g. service delivery, or patient outcomes) [19,20].
The findings of the literature review will contribute to the design and evaluation of interventions to improve clinical leadership at the bedside in LMICs generally, and in the maternity services of South Africa specifically.

Aim
The aim of the literature review was to describe the characteristics and the evaluation of clinical leadership development interventions targeting frontline healthcare providers.

Design
A structured approach, the systematic quantitative literature review method [21], was used to search and identify the literature, and extract information on interventions for clinical leadership development.

Search methods
The searches were conducted using Google Scholar and EBSCOhost search engines. Additionally, electronic databases including CINHAL, PubMed, Medline, Academic Search Complete, Health Sources: Nursing/Academic Edition, Science Direct and Ovid®), were searched using the following keywords: 'clinical leadership' , 'frontline leadership' , 'nursing leadership' , 'ward leadership' , 'medical leadership' , 'clinician leadership' in combination with: 'development' , 'programme' , 'interventions' , 'evaluation' and 'training'. A manual search was conducted to trace sources in the reference list of retrieved published articles.

Eligibility criteria
Inclusion criteria Papers meeting the following criteria were included for review: (1) original research published in peer-reviewed journals; (2) grey literature; (3) reporting the implementation or evaluation of interventions for clinical leadership development; (4) published in English between 2004 and 2017.
Exclusion criteria Papers exploring the implementation and/or the evaluation of interventions or approaches for the development of health service or organizational leadership, or development of senior healthcare leaders were excluded.

Assessment of publications
The database search generated 1600 records; grey literature (health services reports, research reports, theses, and dissertations) generated 75 records; of which 1558 were duplicate. On a review of abstracts 117 papers were excluded [related to developing or evaluating organizational or health services leadership]. On screening 50 potentially relevant full-text papers, 24 were excluded [did not report on developing or evaluating clinical leadership for frontline healthcare providers], and 2 [did not have full text available]. Twenty-four papers met the inclusion criteria and were captured in the database. Figure 1 presents the search algorithm indicating the number of identified studies, included and excluded studies, and reasons for exclusion.
The quality of the studies reviewed was appraised using the Standards for Reporting Implementation Studies (StaRI) [22]. Against the StaRI criteria, the studies reporting the interventions and the evaluation of interventions for clinical leadership development did not provide adequate descriptions of the interventions themselves, of the methods used in implementing the interventions, and of the evaluation of the interventions. However, they included sound descriptions of the aims and the target groups for which the interventions were designed. Two studies provided sufficient descriptions of the intervention, the implementation and the evaluation in order to produce transferable findings [23,24]. Overall, the studies included in this review were of poor quality. However, the shortcomings identified did not detract from the purpose of the present literature review.

Constructing the database
A database was constructed to summarise the studies identified for the review. The following information was captured in the database: the country where the intervention/ evaluation was implemented, the aim of the intervention/ evaluation, the target population for which the intervention was designed, the content areas of the intervention, the educational approach used, the educational techniques used, the time frame of the intervention, how the impact of intervention was measured, the outcomes and limitations of the intervention as reported in the papers.

Results
The aim of this literature review was to establish, from the published corpus, how clinical leadership was developed among frontline healthcare providers. Interventions for clinical leadership were summarized and synthesized. A total of 24 papers exploring the implementation and the evaluation of interventions for clinical leadership development met the inclusion criteria. The interventions are summarized below.

Aims of the interventions
The emphasis of most interventions was on developing clinical skills. Some interventions were designed to develop leadership competencies, to promote succession planning, to enhance the contribution of frontline healthcare providers to patient experiences, and to ensure quality and safe health services [23-31, 36, 37, 39-42, 44-46] Table 2). Other interventions focused on preparing nursing students, medical students, and novice frontline healthcare providers for future leadership roles and for ensuring quality care and patient safety [33,35,43]. Some interventions were developed to transform managers into leaders [32].

Target group for which interventions were implemented
Interventions for clinical leadership development targeted a variety of frontline health care providers (Table 3). Only a few interventions included frontline healthcare providers for maternal and child health [28-30, 32, 34, 36, 37, 41], while the remainder of interventions included early career nurses, qualified nurses, medical doctors, and allied healthcare professionals in hospital settings, including primary and secondary, acute, academic, community and regional hospitals, and mental health and geriatric wards [23, 26, 27, 31, 35, 38-40, 44, 45]. Other target groups include novice students, senior level nursing students, senior registrars, and postgraduate medical and dental students [24,25,42,43].

Educational approaches
Primarily the interventions for clinical leadership development were offered in the form of in-service training using a work-based learning (WBL) educational approach within the clinical settings [28, 29, 31-34, 36-38, 42, 45, 46]. Classroom-based learning (CBL) conducted in classrooms outside of clinical settings [30,43] or a combination of both were also used [27,41,44]. Some interventions were offered as postgraduate training programmes, using a combination of WBL and CBL [24,25,30,35] (Table 5).

Time frame of interventions for clinical leadership development
Most interventions for clinical leadership development were offered as multiple contact sessions of varying duration, ranging from a few days, to a few weeks, or to lasting several months [24,26,30,33,[37][38][39][40][44][45][46] (Table 7). Other interventions were offered as multiple contact sessions in postgraduates programmes [24,30,35]. One intervention was offered as a full-time master degree programme with no detail of the contact sessions provided [25].

How interventions were measured
Best practice in measuring an intervention is to use pre-post evaluation. Nine out of twenty-four studies used pre-and post-test methods to measure the learning attainment, behaviour, and impact of the intervention  [23,24,28,35,38,41,42,44,46]. Fifteen studies used only post-test methods to measure the effectiveness of the interventions (Table 8).
To categorize how the different articles evaluated their interventions, Kirkpatrick's approach was used. Only one study included an evaluation at all four levels namely, the reaction, learning attainment and behaviour, and impact of the intervention on service delivery [32]. Measuring participant reactions to the interventions was common to most interventions [23-27, 29-32, 35, 38-40, 43]. Learning attainment, and the behavior of participants were also measured. The tools used to collect evaluation data included self-report questionnaires, online surveys, evaluation sheets, structured evaluation forms. Additional tools included in-depth-interviews, group interviews, FGDs, observations of action learning sets and document review.

Outcomes of the interventions as reported in the papers
The outcomes of the interventions recorded in the papers include: personal development [increased self-awareness and confidence, feelings of empowerment, time management, development of emotional intelligence skills and increased learning ability] [27,32,34,37,38,45]; enhanced leadership knowledge and skills [communication, willingness to lead teams, delegation, ability to empower others, problem solving, decision making, ability to inspire a shared vision, team management] [24, 26-29, 32, 34-37, 41, 43-46]; improved clinical knowledge and skills [enhanced basic nursing knowledge and skills, improved clinical practices, understanding of contribution to patient care] [42], improved teamwork [ability to work as part of multi-disciplinary teams, ability to manage teams] [23,25,30,31,37,40,43], improved patient care [increased focus on patient  Table 9).

Limitations of the interventions
Of studies that reported the limitations of interventions the following were identified: difficulty in gaining consent from patients to be observed while care was being provided and some trainers may not be skilled enough to observe using direct observation [39]; interventions that were too intensive and demanding, affecting the motivation and ability of participants to attend all sessions [27,30,40]; time away from clinical duties, resistance from colleagues to implement changed practices, and nurses or midwives taking clinical leadership roles and lack of support from health service managers [23,25,34,41]; short timeline for progamme implementation which did not allow for assessing the impact of interventions on participants, service users and on service delivery [23,29,34,36]; and challenges with sustainability of gains made through the interventions [31,38]. A lack of a control group in evaluating interventions was also considered a limitation in attributing changes to the intervention [24]. The transferability of the intervention was also questioned [42] (Table 10).

Discussion
This literature review of the implementation and evaluation of interventions for clinical leadership development was conducted towards identifying a model to inform clinical leadership development among frontline healthcare providers in Low-and Middle-Income Countries (LMICs) generally, and for the delivery of optimal maternal and perinatal care in South Africa specifically. All descriptions of interventions for clinical leadership development derive from studies implemented in HICs. This would limit the transferability of study findings to LMICs, where clinical leadership is still underdeveloped and healthcare systems are faced with different contextual challenges [8]. Studies are required to explore appropriate interventions to improve clinical leadership in LMICs, including South Africa.
Of note, clinical leadership development programmes targeted novice to veteran frontline healthcare providers,  [34,38,45]. Other interventions were more selective, based on checklists of whether participants manifested certain clinical skills. Interventions that embrace a holistic conceptualization of clinical leadership are more detailed, and can produce well trained and skilled clinical leaders. However, they may be expensive, and may require longer training periods, as they include multiple dimensions of clinical leadership. Interventions based on a selective understanding of clinical leadership may be shorter in nature, as they may focus on fewer dimensions of clinical leadership. However, these interventions may not be able to produce skilled clinical leaders.
Most interventions for clinical leadership development used work-based learning as an educational approach to improve, develop, maintain or increase practicing professionals' competence in the clinical setting [47,48]. Work-based learning (WBL) has been shown to promote practical learning and to help practitioners relate new knowledge to their work environment [49,50]. Classroom-based learning takes participants away from their work environment, a feature often considered as a major weakness of this approach [49,50]. A systematic review evaluating in-service training suggests that WBL is the most appropriate approach to improve not only the knowledge of participants but also the skills, behaviors and attitudes of participants [51][52][53]. WBL with In many interventions, the actual length of exposure to contact sessions, and the balance of time between the delivery of training content, and hands-on activities, were not detailed. The paucity of information poses a challenge when trying to replicate the interventions to other settings. In the interventions that did indeed describe the length of exposure to the intervention, multiple contact sessions, over varying periods of time, were used to deliver the interventions. Intensive once-off training sessions are shown to have a negative impact on participants' motivation [27,40]. Multiple time-spaced contact sessions appear to be the most suitable approach to delivering in-service training programmes, as they provide participants with sufficient time and space to engage, reflect on the content of the training programme, and apply knowledge and skills to the work place [52,54]. While designing interventions for clinical leadership development, there is a need to ensure that a reasonable timeframe tailored to participants' needs is provided.
Most studies used only post-test evaluation to measure the effectiveness of the interventions. Post-test evaluation is outcome oriented and is concerned with the results of the intervention. The absence of pre-test observations and a lack of a control group in post-test evaluations limits the ability to attribute observed changes to the intervention [55]. Nonetheless, post-test is used in most interventions because of the logistical difficulties in obtaining pre-test observations due to time constraints [56].
Pre-post-test evaluation may be the most accurate way to provide a full picture of changes in participants over the course of the training programme. [56]. However, many interventions were implemented as once-off short interventions, over a couple of weeks. A short implementation timeline may not be sufficient to allow change to occur, and may not permit sufficient time to measure Kirkpatrick's approach to evaluation recommends four levels of evaluation to objectively measure the effectiveness of training programmes [19]. Most papers did not provide thorough descriptions of evaluation methods. Only one study reviewed included an evaluation at each of the four levels suggested by Kirkpatrick's approach [32]. Most papers reporting the evaluation of interventions for clinical leadership development focused on the reactions of participants and learning attainment [27,32,34,37,38,45]. Participants reported positive experiences, and indicated the acquisition of leadership knowledge and skills as result of the intervention [27,32,34,37,38,45]. Some studies reported improved clinical knowledge and skills improved teamwork as the behavior of participants [23,25,29,33,34,40,42,43]. The impact of interventions include improved patient care, improved patient outcomes, and change in care processes [24,28,33,39,41,45].
Although some interventions used validated tools to evaluate the interventions, most outcomes recorded in this review used self-reported changes. Tools that elicit self-reported learning attainment and behaviour changes are considered to provide weak evaluation evidence and are of variable accuracy [57]. Factors that affect accuracy include information bias, influenced by recall bias and social desirability bias, and design bias, influenced by questionnaire design and mode of data collection [57]. To move beyond the weaknesses of to self-reported changes, the literature suggests the use of 360°assessments [58][59][60]. This method involves an individual and several other people (e.g. peers, supervisors, assessors, and managers) provide a comprehensive feedback on an individual's behaviour and effectiveness [60]. It is suggested that used in combination with training programmes or interventions, 360°feedback can be an effective assessment tool [58][59][60]. Adequate descriptions of interventions, and rigorous description of methods used in implementing, and evaluating the interventions are required to ensure transferability of findings of interventions to other settings.
Most studies did not discuss the limitations of the interventions, or the sustainability of gains made through the intervention. One strategy to ensure sustainability of interventions for clinical leadership development is the team training approach [23]. A team training approach to clinical leadership development may serve a dual purpose: the transfer of skills and teambuilding. Teambuilding is an integral part of clinical leadership development, as well as an outcome of clinical leadership. A team training approach allows multiple professionals to be trained together, reduces resistance to change, and reduces the resistance to frontline healthcare leaders taking clinical leadership roles [23].

Strengths of the review
This review highlights the diversity, extent, and gaps of interventions for the development of clinical leadership among frontline healthcare providers. The review also highlights the conceptualizations of clinical leadership embedded in the interventions, and the challenges encountered in the implementation of interventions for clinical leadership development.

Limitations of the review
Although rigorous steps were carried out in this review, we are also aware of some limitations -Studies may have been omitted from the review if they were not published in the databases searched, or if they were published in languages other than English.   focus on clinical skills and on competencies that support optimal clinical care. Thirdly, interventions for clinical leadership development should use work-based learning approaches, and experiential and practice-based learning techniques, as these are more likely to contribute to the sustainable development of clinical leadership among frontline healthcare providers, and to the improvement in overall service delivery. Fourthly, team-based approaches to clinical leadership development, implemented through multiple contacts over a period of time, allow the acquisition and the transfer of skills, and teambuilding. Fifthly, assessment of the expected learning and evaluation of expected outcomes need to be carefully planned in the design of clinical leadership development interventions, and measured preferable through pre-post assessments, and 360 0 assessments. Lastly, adequate description of the implementation setting, of the intervention model, and of the methods used in implementing and evaluating the interventions are necessary to ensure transferability of an intervention to other settings. These guidelines established from this review of the literature, must be incorporated in the design of interventions for clinical leadership development in LMIC settings.