Purpose
Research in psychology or management science has shown that shared leadership (SL) enhances information sharing, fosters participation and empowers team members within the decision-making processes, ultimately improving the quality of performance outcomes. Little has been done and, thus, less is known of the value and use of SL in acute healthcare teams. The purpose of this study is to (1) explore, identify and critically assess patterns and behaviour of SL in acute healthcare teams; and (2) evaluate to what extent SL may benefit and accomplish safer care in acute patient treatment and healthcare delivery.

Design/methodology/approach
The authors conducted a review that followed the PRISMA-P reporting guidelines. A variety of sources were searched in April 2018 for studies containing primary research that focused on SL in acute healthcare teams. The outcome of interest was a well-specified assessment of SL, and an evaluation of the extent SL may enhance team performance, lead to safer patient care and healthcare delivery in acute healthcare teams.

Findings
After the study selection process, 11 out of 1,383 studies were included in the review. Studies used a qualitative, quantitative or mixed-methods approach. Emerging themes based on behavioural observations that contributed to SL were: shared mental model; social support and situational awareness; and psychological safety. High-performing teams showed more SL behaviour, teams with less seniority displayed more traditional leadership styles and SL was associated with increased team satisfaction.

Research limitations/implications
Evidence to date suggests that SL may be of benefit to improve performance outcomes in acute healthcare team settings. However, the discrepancy of SL assessments within existing studies and their small sample sizes highlights the need for a large, good quality randomized controlled trial to validate this indication.

Originality/value
Although studies have acknowledged the relevance of SL in healthcare service and delivery, a systematic, evidence-based and robust evaluation of behavioural patterns and the benefits of SL in this field is still missing.