In a systematic review and meta-analysis of 99 randomized controlled trials involving 9,330 health care professionals, investigators found differential effectiveness of burnout interventions across professional roles, with professional coaching showing probable benefit among physicians and mindfulness-based interventions demonstrating potential effectiveness for nurses, midwives, and mixed health care professionals.
Role-Specific Intervention Effects
Among physicians, professional coaching interventions lasting more than 4 weeks produced a small reduction in emotional exhaustion (standardized mean difference = −0.37, low certainty) and probably reduced depersonalization (standardized mean difference = −0.30, moderate certainty). However, mindfulness-based interventions may not effectively reduce emotional exhaustion in this population (standardized mean difference = −0.46, very low certainty).
Among nurses and midwives, mindfulness-based interventions demonstrated potential for large reductions in emotional exhaustion (standardized mean difference = −0.90, low certainty), while job-role training showed moderate certainty evidence for small improvements in personal accomplishment (standardized mean difference = 0.20).
Among mixed health care professional groups, mindfulness-based interventions may produce small-to-moderate reductions in emotional exhaustion (standardized mean difference = −0.40, low certainty), small reductions in depersonalization (standardized mean difference = −0.33, low certainty), and probably result in moderate improvement in personal accomplishment (standardized mean difference = 0.48, moderate certainty).
Clinical Significance of Effect Sizes
Supplementary analysis of unstandardized mean differences on the Maslach Burnout Inventory emotional exhaustion subscale (0 to 48 scale) revealed reductions of more than 9 units for nurses and midwives receiving mindfulness interventions, just under 5 units for mixed health care professional groups, and greater than 3.5 units for physicians receiving professional coaching.
"[A] 1-unit increase in MBI-EE is associated with an at least 5% increased risk for self-perceived major medical errors among physicians and surgeons," noted lead study author George Collett, PhD, and colleagues, suggesting clinical meaningfulness despite modest effect sizes.
Methodologic Limitations and Evidence Quality
Among the 93 individual randomized controlled trials and six cluster randomized controlled trials included, 91 of the individual randomized controlled trials and all of the cluster randomized controlled trials received an overall high risk of bias ratings, driven predominantly by the unblinded nature of interventions. Most subgroups were rated as having a very low grade of evidence using Grading of Recommendations Assessment, Development and Evaluation criteria.
Substantial heterogeneity persisted within intervention-type subgroups even following stratification by professional role, with I² statistics ranging from about 70% to 87% for several primary analyses. Meta-regression analyses suggested that longer intervention duration (more than 4 weeks) and higher baseline burnout scores were associated with a greater magnitude of effect.
Evidence Gaps and Future Directions
The investigators identified several limitations in the current evidence base. Despite including studies across multiple health care professions, total numbers of both studies and participants remained modest, particularly for professions other than physicians and nurses. All included studies focused exclusively on individual-level interventions, with no organizational-level strategies evaluated in experimental designs.
The review was funded by Barts Charity and prospectively registered in PROSPERO (CRD42024552385). The investigators searched PubMed and Scopus through May 14, 2025, including only English-language publications that evaluated modifiable interventions vs no active intervention using validated continuous burnout outcome measures.
The shortest interventions included a single 5-hour multifaceted workshop and 2-day programs, while others ranged from 2 weeks to 10 months duration. Most studies reported immediate postintervention follow-up, with only eight studies conducting assessments at extended time points after intervention completion.
Source: Annals of Internal Medicine, Editorial
