Workforce burnout
Can the system replace burned-out clinicians, or is the training pipeline itself throttled?
Physician burnout hit a record 62.8 percent in 2021 and eased to 45.2 percent in 2023, still above pre-pandemic levels. Specialties with the heaviest electronic-record load report some of the highest burnout, and health care workers die by suicide at rates above the general working population.
The problem
Workforce burnout is a national capacity risk for hospitals and clinics because exhausted clinicians cannot be replaced quickly when training pipelines are constrained. Documentation load, staffing stress, faculty shortages, and pay gaps combine into an operating threat that affects access, quality, and financial performance.
The recommendation
Treat workforce sustainability as a retention and pipeline strategy. The recommended approach is to reduce documentation burden, improve staffing models, invest in faculty and training capacity, and monitor whether burnout interventions actually preserve clinical hours in underserved markets.
How bad, and who
The national trajectory, from a record 62.8 percent down to 45.2 percent, and which specialties carry the heaviest load now versus which are easing.
Physician burnout, national trend 2011 through 2023
Share of physicians reporting at least one symptom of burnout, AMA / Mayo Clinic / Stanford study.
Read it this way The spike to 62.8 percent in 2021 and partial retreat to 45.2 percent by 2023 shows burnout does not climb steadily, it swings with events, here the pandemic. The 2023 level still sits above every pre-pandemic reading except 2014, so the retreat has not returned burnout to its historical floor. Use this chart to see whether the workforce risk is retention, workload, or replacement capacity, and why the recommendation addresses both burnout and the training pipeline.
⊞ data table⬇ CSV
| Year | Percent reporting burnout |
|---|---|
| 2011 | 45.5 |
| 2014 | 54.4 |
| 2017 | 43.9 |
| 2020 | 38.2 |
| 2021 | 62.8 |
| 2023 | 45.2 |
Shanafelt et al., Mayo Clinic Proceedings (AMA / Mayo Clinic / Stanford national study) · 2011-2023 · source
Burnout by specialty, 2023 survey wave, ranked
Bars above the line are more burned out than the national physician average. Medscape opt-in panel, directional per specialty.
Read it this way The ten specialties above the 45.2 national reference line, led by Emergency Medicine at 63 percent, are where burnout concentrates. The ten below it, down to Plastic Surgery at 37 percent, are relatively less affected. This is a different, opt-in survey than the national trend line, so use it to compare specialties against each other, not to update the national figure. Use this chart to see whether the workforce risk is retention, workload, or replacement capacity, and why the recommendation addresses both burnout and the training pipeline.
Caveat Medscape is a large opt-in online panel using a single self-report item, not comparable to the AMA / Mayo national instrument in the trend chart. The reference line is the AMA / Mayo national figure, shown for orientation only.
⊞ data table⬇ CSV
| Specialty | Burnout %, 2023 |
|---|---|
| Emergency Medicine | 63 |
| Ob/Gyn | 53 |
| Oncology | 53 |
| Pediatrics | 51 |
| Family Medicine | 51 |
| Radiology | 51 |
| Internal Medicine | 50 |
| Anesthesiology | 50 |
| Pulmonary Medicine | 50 |
| Gastroenterology | 50 |
| Critical Care | 45 |
| General Surgery | 45 |
| Neurology | 44 |
| Endocrinology | 44 |
| Orthopedics | 44 |
| Otolaryngology (ENT) | 43 |
| Pathology | 41 |
| Ophthalmology | 39 |
| Psychiatry | 39 |
| Plastic Surgery | 37 |
Medscape Physician Burnout and Depression Reports 2023 and 2024 · 2022-2023 · source
Burnout by specialty, 2022 versus 2023 survey waves
Medscape opt-in panel. Directionally informative, not a nationally representative estimate per specialty. The 15 specialties with a point in both waves.
Read it this way Every one of these 15 specialties shows a lower or flat 2023 bar next to its 2022 bar, so the easing in burnout was broad rather than concentrated in one field. Internal Medicine's drop from 60 to 50 percent is the largest swing shown. This pairs only the specialties present in both survey waves, so it cannot speak to the three specialties dropped from the comparison. Use this chart to see whether the workforce risk is retention, workload, or replacement capacity, and why the recommendation addresses both burnout and the training pipeline.
Caveat Medscape is a large opt-in online panel using a single self-report item, not comparable to the AMA / Mayo national instrument in the trend chart. Three specialties with only a 2022 point are excluded from this paired view.
⊞ data table⬇ CSV
| Specialty | 2022 % | 2023 % |
|---|---|---|
| Emergency Medicine | 65 | 63 |
| Ob/Gyn | 58 | 53 |
| Oncology | 52 | 53 |
| Pediatrics | 59 | 51 |
| Family Medicine | 57 | 51 |
| Radiology | 54 | 51 |
| Internal Medicine | 60 | 50 |
| Anesthesiology | 55 | 50 |
| Pulmonary Medicine | 54 | 50 |
| Gastroenterology | 52 | 50 |
| Critical Care | 55 | 45 |
| General Surgery | 51 | 45 |
| Neurology | 55 | 44 |
| Endocrinology | 51 | 44 |
| Orthopedics | 45 | 44 |
Medscape Physician Burnout and Depression Reports 2023 and 2024 · 2022-2023 · source
The cause and the cost
The measured driver, documentation and electronic-record load, paired with the human toll of suicide risk above the general working population.
Electronic record time versus burnout, by specialty
Each point is a specialty. Two independently sourced measures from the same window, not a computed correlation.
Read it this way Family Medicine and Internal Medicine sit at both the highest electronic-record time, 7.3 hours per 8 scheduled patient hours, and among the highest burnout, 57 and 60 percent, while Anesthesiology logs far less record time (2.5 hours) alongside comparable burnout, so the pattern is not uniform across every specialty. The two measures come from separately fielded surveys matched only by specialty label, so this shows an association worth noting, not a proven cause. Use this chart to see whether the workforce risk is retention, workload, or replacement capacity, and why the recommendation addresses both burnout and the training pipeline.
Caveat EHR time (Nov 2021 to Apr 2022) and burnout (Medscape wave fielded Jun to Oct 2022) are separately sourced and paired by specialty label. No correlation coefficient is claimed.
⊞ data table⬇ CSV
| Specialty | EHR hours per 8 patient hours | Burnout % |
|---|---|---|
| Infectious Disease | 8.4 | 58 |
| Endocrinology | 7.7 | 51 |
| Nephrology | 7.5 | 44 |
| Family Medicine | 7.3 | 57 |
| Internal Medicine | 7.3 | 60 |
| Anesthesiology | 2.5 | 55 |
| Orthopedics | 3.3 | 45 |
Holmgren et al., J Gen Intern Med 2024 (EHR time); Medscape 2023 (burnout) · 2022 · source
Suicide rate, health care workers versus the general working population
Deaths per 100,000, national cohort 2008 through 2019. Comparator is non health care workers.
Read it this way All three health care groups sit above the general working population's rate of 12.6 per 100,000, with support workers highest at 21.4, so elevated suicide risk is not limited to physicians or nurses alone. The comparator group is other workers, not the full US population, which is a narrower baseline than it might first appear. Use this chart to see whether the workforce risk is retention, workload, or replacement capacity, and why the recommendation addresses both burnout and the training pipeline.
Caveat All four figures are from the same Olfson et al. study, whose comparator is non health care workers, not the total US population.
⊞ data table⬇ CSV
| Group | Suicide rate per 100k |
|---|---|
| Health care support workers | 21.4 |
| Nurses | 16 |
| Physicians | 13.1 |
| General working population | 12.6 |
Olfson et al., JAMA 2023, Suicide Risks of Health Care Workers in the US · 2008-2019 · source
Nurse suicide rate within each sex, 2005 to 2016
Deaths per 100,000. The elevated nurse risk holds within each sex, which aggregate comparisons understate.
Read it this way Within both sexes, nurses' suicide rate exceeds the sex-matched general population: 10.0 versus 7.0 for women and 33.0 versus 27.0 for men, which rules out the elevated nurse rate elsewhere on this page being just an artifact of nursing's mostly-female workforce. This uses a different comparator and time window than the Olfson figures on this page, so the two should not be combined into one number. Use this chart to see whether the workforce risk is retention, workload, or replacement capacity, and why the recommendation addresses both burnout and the training pipeline.
Caveat Davidson et al. compares nurses to the sex-matched general population over 2005 to 2016. Do not combine with the Olfson figures, which use a different comparator and window.
⊞ data table⬇ CSV
| Sex | Nurses per 100k | General population per 100k |
|---|---|---|
| Female | 10 | 7 |
| Male | 33 | 27 |
Davidson et al., Worldviews on Evidence-Based Nursing 2020, A Longitudinal Analysis of Nurse Suicide in the US · 2005-2016 · source
Can we replace them
Whether the training pipeline can replenish a burning-out workforce, and the faculty-compensation bottleneck throttling it.
Qualified nursing-school applicants turned away, 2014 to 2023
Every one of these applicants met admission standards but found no seat.
Read it this way The line never drops below 56,000 turned away in any year and peaked at 91,938 in 2021, showing a decade-long, persistent bottleneck rather than a one-time problem. The 2023 figure of 65,766 is an improvement from the 2021 peak but still well above the 2017 low. Use this chart to see whether the workforce risk is retention, workload, or replacement capacity, and why the recommendation addresses both burnout and the training pipeline.
⊞ data table⬇ CSV
| Year | Qualified applicants turned away |
|---|---|
| 2014 | 68938 |
| 2015 | 68936 |
| 2016 | 64067 |
| 2017 | 56000 |
| 2018 | 80407 |
| 2019 | 80521 |
| 2020 | 80407 |
| 2021 | 91938 |
| 2022 | 78191 |
| 2023 | 65766 |
American Association of Colleges of Nursing, Enrollment and Graduations in Baccalaureate and Graduate Programs in Nursing · 2023 · source
Why qualified applicants are turned away
Faculty shortage alone accounts for 64 percent, dwarfing every other constraint.
Read it this way Faculty shortage's 64 percent share dwarfs every other listed constraint, including insufficient clinical sites at 14 percent, so a fix aimed only at classroom space or clinical placements would leave the largest driver untouched. These five categories are the reported drivers of capacity limits and sum to 100 percent, so there is no unlisted cause hiding outside this breakdown. Use this chart to see whether the workforce risk is retention, workload, or replacement capacity, and why the recommendation addresses both burnout and the training pipeline.
Caveat Shares are the reported drivers of program capacity limits and sum to 100 percent.
⊞ data table⬇ CSV
| Driver | Share % |
|---|---|
| Faculty shortage | 64 |
| Insufficient clinical sites | 14 |
| Insufficient classroom or lab space | 9 |
| Insufficient preceptors | 8 |
| Budget cuts | 5 |
American Association of Colleges of Nursing, Enrollment and Graduations in Baccalaureate and Graduate Programs in Nursing · 2023 · source
Nurse faculty pay versus clinical practice, 2023
A master's-prepared nurse earns about $37,355 more in clinical practice than in teaching, which feeds the faculty shortage.
Read it this way The $37,355 gap between what a nurse earns teaching ($87,325) versus practicing clinically ($124,680) is a direct incentive against becoming faculty, which lines up with faculty shortage being the dominant driver of turned-away applicants in the donut chart above. This is one salary comparison for one faculty rank, not a full accounting of every cost or benefit of an academic nursing career. Use this chart to see whether the workforce risk is retention, workload, or replacement capacity, and why the recommendation addresses both burnout and the training pipeline.
⊞ data table⬇ CSV
| Role | Median salary |
|---|---|
| Nurse faculty (associate professor) | 87325 |
| Nurse practitioner in practice | 124680 |
American Association of Colleges of Nursing, Salaries of Instructional and Administrative Nursing Faculty · 2023 · source
Geography
The same question, state by state and then county by county. Pick a state in the filter above to drill into its counties.
Primary-care physicians per 100k
County · direct countEach tile is a state. Pick a state in the Scope control above to drill into its counties.
HRSA Area Health Resources File · 2023 · source
Registered nurses per 100k
County · direct countEach tile is a state. Pick a state in the Scope control above to drill into its counties.
HRSA Area Health Resources File · 2023 · source
Why this matters
Electronic health record time tracks with burnout in several specialties: Family Medicine and Internal Medicine log the most EHR hours per scheduled patient hours (7.3) and also report among the highest burnout (57 and 60 percent), though low-EHR-time Anesthesiology still reports comparable burnout, so documentation load is one driver, not the only one. On the pipeline side, faculty shortage accounts for 64 percent of why qualified nursing applicants are turned away, and nurse faculty earn $37,355 less than nurses in clinical practice, a direct financial disincentive against replenishing the teaching pipeline.
Recommended actions
- Target documentation and inbox-burden reduction first in Family Medicine, Internal Medicine, and other high-EHR-time specialties.
- Screen for suicide risk across all health care worker tiers, not physicians alone, since support workers show the highest rate measured.
- Close or narrow the nurse faculty pay gap to relieve the primary bottleneck, 64 percent of rejections, in the nursing pipeline.
- Expand clinical-site and preceptor capacity as secondary levers, since they account for a combined 22 percent of turned-away applicants.
- Track burnout by specialty and survey wave to catch reversals early, since the 2023 easing was broad but not universal.
The recommendation
Therefore, treat workforce sustainability as a retention and pipeline strategy. The recommended approach is to reduce documentation burden, improve staffing models, invest in faculty and training capacity, and monitor whether burnout interventions actually preserve clinical hours in underserved markets.
Demographic slice none. tracks provider burnout, not patient demographics.
Sources
- Shanafelt et al., Mayo Clinic Proceedings (AMA / Mayo Clinic / Stanford national study) · 2011-2023
- Medscape Physician Burnout and Depression Reports 2023 and 2024 · 2022-2023
- Holmgren et al., J Gen Intern Med 2024 (EHR time); Medscape 2023 (burnout) · 2022
- Olfson et al., JAMA 2023, Suicide Risks of Health Care Workers in the US · 2008-2019
- Davidson et al., Worldviews on Evidence-Based Nursing 2020, A Longitudinal Analysis of Nurse Suicide in the US · 2005-2016
- American Association of Colleges of Nursing, Enrollment and Graduations in Baccalaureate and Graduate Programs in Nursing · 2023
- American Association of Colleges of Nursing, Salaries of Instructional and Administrative Nursing Faculty · 2023