Burnout: what it is and what the evidence says

Short answer

Burnout is a work-related syndrome that ICD-11 defines as resulting from chronic workplace stress that has not been successfully managed, with exhaustion, cynicism and reduced professional efficacy [C1]. It is common in healthcare workers [C2] and overlaps with depression [C3]. Trials of coaching, mindfulness and digital tools show modest, context-dependent reductions [C6][C7][C8], and workload is a key driver [C4].

What it is and how it shows up

Burnout is not a personal failing or a medical diagnosis: the ICD-11 includes it as an occupational phenomenon resulting from chronic workplace stress that has not been successfully managed [C1]. It is usually described as developing gradually in response to prolonged work stress rather than appearing after a single event. It can arise in any occupation, but most of the research — and some of the highest prevalences — concern health care workers [C2].

The syndrome has three dimensions: emotional exhaustion, depersonalisation or cynicism towards one's job, and a reduced sense of professional accomplishment or efficacy [C1]. These are typically measured with self-report questionnaires, most often the Maslach Burnout Inventory, which asks about feelings towards work over the past weeks.

Unlike everyday stress, which can be energising in short bursts, burnout is specifically tied to work and to the feeling that effort is not matched by accomplishment. It is usually described as a process rather than a single state: exhaustion tends to come first, then detachment and cynicism, and finally a sense of ineffectiveness [C1]. Because it is defined by the work context, burnout is not diagnosed as a disease in the ICD-11; it is an occupational phenomenon that can coexist with, but is distinct from, depressive disorders [C1][C3]. Recognising this distinction matters in practice, because the workplace factors that drive burnout are different from those addressed when treating depression [C4].

Burnout is common in healthcare. A global meta-analysis of studies of nurses reported an overall pooled prevalence of burnout symptoms of 11.23% [C2], and a meta-analysis of intensive care and emergency staff reported a pooled burnout prevalence of 46%, with high emotional exhaustion at 48%, high depersonalisation at 30% and low personal accomplishment at 47% [C2].

Burnout overlaps with depression but is treated as a distinct construct. In nursing samples the pooled correlation between burnout and depression was r = 0.403, which the review authors described as a large correlation [C3].

The strongest and most consistent correlates are organisational: in a meta-analysis of 84 relevant studies, quantitative workload, role stress and work-family conflict were positively associated with burnout [C4]. Burnout also matters for the quality of care: in a meta-analysis of 170 observational studies covering 239,246 physicians, burnout was associated with about doubled patient safety incidents (2.04) and with markedly lower job satisfaction (odds ratio 3.79) [C5].

Lifestyle factors show weaker, observational links: in a systematic review of health care workers, physical activity was often associated with a reduced risk of burnout, and reported burnout prevalence across studies ranged from 7% to 83% depending on how it was measured [C11]. Much of the prevalence research comes from nurses and physicians, so the picture in other professions is less well mapped [C2].

What research shows by approach

Intervention research splits into individual-level programmes, physician-focused trials, digital tools and broader organisational approaches. Most trials measure burnout with self-report questionnaires shortly after the programme ends, so the effects below are mostly short-term changes in self-reported symptoms rather than long-term recovery [C6]. The effects describe groups in studies, not predictions for an individual.

Coaching and individual-level programmes

In a meta-analysis of 93 randomized and cluster trials including 9330 participants, individual-level interventions reduced aspects of burnout in health professionals: professional coaching was probably effective in reducing emotional exhaustion (SMD -0.37), and mindfulness-based interventions may reduce burnout among nurses and midwives (emotional exhaustion SMD -0.90) [C6]. The certainty of these estimates was rated low by the review authors, which means future trials could shift the pooled numbers [C6].

Interventions for physicians in randomized trials

A separate meta-analysis of randomized trials of interventions to decrease physician burnout found a mean improvement of 1.11 points in emotional exhaustion, which the authors judged unlikely to produce meaningful changes in clinical burnout [C7]. The pooled estimate rests on many trials with high statistical heterogeneity, so the size of the average improvement is imprecise [C7].

Digital and web-based tools

Digital and web-based interventions showed a significant reduction in burnout scores for physicians in a meta-analysis (between-groups SMD -0.34), though the authors note the evidence is limited by small sample sizes, heterogeneous study designs and high attrition [C8].

Mindfulness programmes for nurses

Results for mindfulness are mixed: in a meta-analysis of randomized trials in nurses, mindfulness-based interventions had a minimal, non-significant effect on burnout (SMD -0.38) but significantly improved resilience (SMD 0.27) [C9]. The resilience gain suggests mindfulness may act on related resources rather than on burnout itself, but the trials were heterogeneous [C9].

Broader intervention programmes for nurses

Broader intervention programmes targeting nurses produced significant reductions in global burnout (g = -0.71), while personal accomplishment showed small, imprecise and inconsistent changes [C10]. Because very different programmes were pooled together, which component matters most is not clear from this evidence [C10].

Organisational and workload factors

Because quantitative workload, role stress and work-family conflict are positively associated with burnout, interventions that target these organisational drivers are studied as a complement to individual programmes [C4]. The pooled trial evidence reviewed here, however, mostly tested individual-level interventions rather than changes to staffing or scheduling [C6], so the relative contribution of each level is not settled by these meta-analyses [C4].

Limitations and uncertainty

Much of this evidence is cross-sectional and relies on self-report questionnaires, most often the Maslach Burnout Inventory, so associations such as those with workload or physical activity cannot show causation [C4][C11]. Prevalence estimates vary widely with the cut-off used, from 7% to 83% across studies of health care workers [C11]. Intervention trials are often unblinded and short-term, with heterogeneous results: coaching and mindfulness effects in trials were modest and sometimes non-significant [C6][C9], and the physician-trial authors judged their pooled improvement unlikely to be clinically meaningful [C7]. Digital-intervention evidence is limited by small sample sizes and high attrition [C8], and nurse programmes showed small, imprecise changes in personal accomplishment [C10]. Follow-up in most trials is short, so lasting effects are uncertain. Most studies come from high-income countries and health care settings, so generalisation to other professions and cultures is uncertain [C2][C4].

When to see a professional

Burnout itself is not a medical diagnosis, but persistent exhaustion, cynicism and reduced efficacy that impair work, relationships or daily functioning deserve attention [C1]. Because burnout and depression overlap strongly (r = 0.403 in nursing samples) [C3], low mood, loss of interest, disturbed sleep or thoughts of self-harm are reasons to seek an assessment from a qualified clinician rather than relying on self-help alone. A clinician can discuss evidence-based options and check whether the symptoms fit a broader condition. Occupational health services, a general practitioner or a psychologist are typical entry points, and asking for help early is reasonable whenever work stress feels unmanageable. The consequences reviewed here reach beyond the individual: associations with patient safety incidents and job satisfaction show that burnout is not only a personal burden [C5]. This page summarises research findings; it is not a diagnosis and not a treatment plan.

FAQ

Is burnout a mental illness?

No. The ICD-11 classifies burnout as an occupational phenomenon resulting from chronic workplace stress that has not been successfully managed, not as a medical condition [C1].

What are the signs of burnout?

The three dimensions are emotional exhaustion, depersonalisation and a reduced sense of professional efficacy [C1], and burnout symptoms are common in healthcare workers, with a pooled prevalence of 11.23% among nurses globally [C2].

How common is burnout among doctors and nurses?

Pooled estimates vary with the cut-off used: 11.23% of nurses worldwide in one global meta-analysis, and 46% burnout among intensive care and emergency staff in another [C2].

Does mindfulness help with burnout?

Evidence is mixed. Mindfulness-based interventions may reduce burnout among nurses and midwives in a meta-analysis of individual-level trials [C6], but a meta-analysis of randomized trials in nurses found a minimal, non-significant effect on burnout (SMD -0.38) alongside a significant improvement in resilience (SMD 0.27) [C9], and broader nurse programmes produced significant reductions in global burnout (g = -0.71) [C10].

Can burnout affect patient safety?

In a meta-analysis of 170 observational studies of 239,246 physicians, burnout was associated with about doubled patient safety incidents (2.04) and lower job satisfaction (odds ratio 3.79) [C5]. These are associations, not proof of causation.

When should I see a professional for burnout?

If exhaustion, cynicism or reduced efficacy persist or impair daily life, or if low mood or thoughts of self-harm appear, professional assessment is warranted; burnout and depression overlap strongly in nurses (r = 0.403) [C3], and the ICD-11 frames burnout as an occupational phenomenon [C1].

Self-help information does not replace a clinician; people with severe or persistent symptoms, or thoughts of self-harm, should contact a professional or local emergency services.