Copenhagen Burnout Inventory: 19 questions, three sources of exhaustion
The CBI treats burnout as exhaustion and stops there. Kristensen and colleagues published it in 2005 as an argument with the Maslach Burnout Inventory, which measures exhaustion alongside cynicism and a reduced sense of accomplishment. Their complaint was that the three MBI dimensions move in different directions across studies, which makes a combined profile hard to defend. The CBI keeps exhaustion, then asks where it comes from: your life in general, your work, or the people you serve.
Nineteen items in three subscales. Each answer converts to a 0 to 100 scale (always 100, often 75, sometimes 50, seldom 25, never 0), and each subscale is scored as the mean of its items. There is no total score, and adding the subscales together would hide the distinction the scale was built to make.
| Subscale | Items | What it asks about |
|---|---|---|
| Personal burnout | 6 | Tiredness, physical and emotional exhaustion, feeling worn out, feeling weak |
| Work-related burnout | 7 | Exhaustion attributed to the job itself, including mornings before work starts |
| Client-related burnout | 6 | Exhaustion from dealing with patients, students or clients |
People with no direct client contact, such as back-office staff, can answer the first two subscales only. One item, about having enough energy for family and friends in leisure time, is reverse scored. Skipped items are handled by a rule rather than by guessing: fewer than 3 answers on personal or client burnout, or fewer than 4 on work burnout, and the response is dropped.
Where the line sits
A subscale score of 50 or more counts as a high degree of burnout, and in the Danish PUMA study that was true for 22.2% of people on personal burnout, 19.8% on work burnout, and 15.9% on client burnout. Those percentages are a useful reality check. Scoring high is not rare in a working population, and it is not evidence of a disorder.
Danish norms
| Sample | Subscale | Mean (SD) | N |
|---|---|---|---|
| Representative Danish adults, 20 to 59 | Personal | 32.7 (15.7) | 1,498 |
| Employed adults | Personal | 31.7 (14.8) | 1,186 |
| Adults without current employment | Personal | 36.6 (18.3) | 312 |
| PUMA study, human services staff | Personal | 35.9 (16.5) | 1,898 |
| PUMA study | Work | 33.0 (17.7) | 1,910 |
| PUMA study | Client | 30.9 (17.6) | 1,752 |
Two details stand out. People without a job scored higher on personal burnout than people with one, which fits the idea that the subscale measures exhaustion rather than work stress. And in the PUMA sample women averaged 36.9 on personal, 33.4 on work and 30.0 on client burnout, while men averaged 30.8, 31.3 and 35.6. The gender gap reverses on the client subscale, so a headline about women burning out more would be half the story.
Reliability
The three subscales reported alphas of 0.87, 0.87 and 0.85, with the range across studies given as 0.85 to 0.87. Over a three-year follow-up, the same subscale correlated 0.59, 0.51 and 0.59 between measurements. Those are moderate numbers for a three-year gap, and they say something real about burnout: it is stable enough to predict things, and changeable enough that today's score is not a life sentence.
What it relates to
Higher burnout goes with lower vitality (correlations from -0.46 to -0.75 with the SF-36 vitality scale) and worse mental health scores (-0.39 to -0.67). In an Italian validation with 915 physicians, alphas ran 0.90 to 0.96, model fit was CFI 0.94, TLI 0.93, RMSEA 0.09 and SRMR 0.04, and burnout correlated 0.76 with a depression measure and 0.73 with an anxiety measure. That last pair is worth sitting with. Exhaustion and depression travel together closely enough that no questionnaire can sort one from the other, and anyone whose exhaustion comes with low mood, lost interest or hopelessness should be assessed for depression rather than handed a burnout label.
The same study found that burnout predicted sickness absence, sleep problems and painkiller use three years later, and that a cutoff of 69 out of 90 on the combined score separated clinical from non-clinical cases with 0.91 sensitivity and 0.85 specificity.
Chinese versions
Two Chinese revisions have been published. A Taiwanese one with employees supported the three-factor structure, explaining 67.98% of the variance with subscale alphas above 0.80. A mainland revision with healthcare staff reported a content validity index of 0.791 and subscale alphas above 0.80. The scale is public domain, which is a large part of why so many translations exist, and it is on the US National Academy of Medicine list of recommended instruments for measuring clinician wellbeing.
What this page is not
Burnout is a description of a state, not a diagnosis, and the CBI measures one part of it. Exhaustion has many causes, including sleep debt, depression, illness and a job that genuinely asks too much, and the questionnaire cannot tell them apart. Read your three subscale scores next to each other: high personal burnout with low work burnout points somewhere other than a high work score. If exhaustion has been running for months, that is worth a conversation with a clinician. This page is not a diagnosis and does not replace one.