Collective leadership involves multiple professionals sharing viewpoints and knowledge. Based on the available evidence, we cannot be sure it makes much difference for professional actions, patient health care or staff well-being. Our confidence in these results varies from moderate to very low, severely limited by the low quality and number of included studies.
What did we want to find out?
We aimed to see whether experiences with collective leadership (as opposed to more centralised and hierarchical leadership styles) improve professional actions, patient health care, and staff well-being. We looked for studies where researchers compared collective leadership with centralised leadership.
What did we do?
We collected and analysed all relevant studies with collective leadership interventions characterised by sharing decisions and interactions among health professions.
What did we find?
We found three relevant studies (955 participants). The studies were carried out in hospitals in Canada, Iran and the USA. Collective leadership interventions probably improve leadership (3 studies, 955 participants), may improve teamwork (1 study, 164 participants), and may slightly decrease work-related stress (1 study, 164 participants). We do not know if collective leadership has an effect on these outcomes: clinical performance (1 study, 60 participants), inpatient deaths (1 study, 60 participants), and staff absence (1 study, 60 participants).
What are the limitations of the evidence?
We are moderately confident that collective leadership improves leadership in healthcare settings. The evidence showed that collective leadership had a large effect on leadership strategies. We are less confident in our results about teamwork and work-related stress. We are not confident in the evidence related to clinical performance, inpatient deaths and staff absence. It is possible that people in the studies were aware of which intervention they were getting. Not all studies provided data about everything that we were interested in. The evidence is based on few cases.
How up to date is this review?
We searched for studies published up to January 2021.
Collective leadership involves multiple professionals sharing viewpoints and knowledge with the potential to influence positively the quality of care and staff well-being. Our confidence in the effects of collective leadership interventions on professional practice, healthcare outcomes and staff well-being is moderate in leadership outcomes, low in team performance and work-related stress, and very low for clinical performance, inpatient mortality and staff absence outcomes. The evidence was of moderate, low and very low certainty due to risk of bias and imprecision, meaning future evidence may change our interpretation of the results. There is a need for more high-quality studies in this area, with consistent reporting of leadership, team performance, clinical performance, health status and staff well-being outcomes.
Collective leadership is strongly advocated by international stakeholders as a key approach for health service delivery, as a response to increasingly complex forms of organisation defined by rapid changes in health technology, professionalisation and growing specialisation. Inadequate leadership weakens health systems and can contribute to adverse events, including refusal to prioritise and implement safety recommendations consistently, and resistance to addressing staff burnout. Globally, increases in life expectancy and the number of people living with multiple long-term conditions contribute to greater complexity of healthcare systems. Such a complex environment requires the contribution and leadership of multiple professionals sharing viewpoints and knowledge.
To assess the effects of collective leadership for healthcare providers on professional practice, healthcare outcomes and staff well-being, when compared with usual centralised leadership approaches.
We searched CENTRAL, MEDLINE, Embase, five other databases and two trials registers on 5 January 2021. We also searched grey literature, checked references for additional citations and contacted study authors to identify additional studies. We did not apply any limits on language.
Two groups of two authors independently reviewed, screened and selected studies for inclusion; the principal author was part of both groups to ensure consistency. We included randomised controlled trials (RCTs) that compared collective leadership interventions with usual centralised leadership or no intervention.
Three groups of two authors independently extracted data from the included studies and evaluated study quality; the principal author took part in all groups. We followed standard methodological procedures expected by Cochrane and the Effective Practice and Organisation of Care (EPOC) Group. We used the GRADE approach to assess the certainty of the evidence.
We identified three randomised trials for inclusion in our synthesis. All studies were conducted in acute care inpatient settings; the country settings were Canada, Iran and the USA. A total of 955 participants were included across all the studies. There was considerable variation in participants, interventions and measures for quantifying outcomes. We were only able to complete a meta-analysis for one outcome (leadership) and completed a narrative synthesis for other outcomes. We judged all studies as having an unclear risk of bias overall.
Collective leadership interventions probably improve leadership (3 RCTs, 955 participants). Collective leadership may improve team performance (1 RCT, 164 participants). We are uncertain about the effect of collective leadership on clinical performance (1 RCT, 60 participants). We are uncertain about the intervention effect on healthcare outcomes, including health status (inpatient mortality) (1 RCT, 60 participants). Collective leadership may slightly improve staff well-being by reducing work-related stress (1 RCT, 164 participants). We identified no direct evidence concerning burnout and psychological symptoms. We are uncertain of the intervention effects on unintended consequences, specifically on staff absence (1 RCT, 60 participants).