Clinician Workforce Strategy: Retain Hours Before Recruiting More

About the Author

Elijah Fernandez, MSW, PMP, CISSP, is co-founder and CTO of CEREVITY Health, Inc., a private-pay concierge network of independent licensed clinicians built for high achievers: executives, founders, physicians, attorneys and finance leaders. He holds an MSW from the University of Southern California, is not a licensed clinician, and spent about 20 years in health care IT, including as director of IT operations at a national mental health company.

The shortage leaders can act on this year

The Association of American Medical Colleges projects a shortage of up to 86,000 physicians by 2036 and reports that 42 percent of the clinical physician workforce is 55 or older.1 The standard response is to expand training. That is necessary and slow: a student who matriculates this fall will not practice independently for 7 or more years. In the meantime, some of the clinical capacity a health system loses leaves through physicians who stay on its roster but work fewer clinical hours.

This commentary argues that clinician workforce strategy should count clinical hours, not only heads, and that retaining those hours requires a broad organizational strategy in which confidential access to mental health care is one component among several. A note on perspective: I hold a master of social work degree but am not a licensed clinician and do not provide clinical care. I write from about 20 years in health care IT and operations, including as director of IT operations at a national mental health company, and from my current role as co-founder and chief technology officer of CEREVITY Health, Inc., a private-pay therapy network. That background shapes the attention given here to records and data flows, and it is also a commercial interest, disclosed below. Where this piece offers my own view rather than published evidence, it says so.

What the evidence shows about lost hours, and what it does not

In a survey of 20,665 health care workers at 124 institutions, 31.4 percent of physicians said they intended to reduce their work hours within 12 months and 23.8 percent said they intended to leave their practice within 2 years. Burnout, workload, fear of exposure and COVID-19-related anxiety or depression were each independently associated with these intentions, and feeling valued by the organization was protective.2 These figures need careful reading. The survey was fielded between July and December 2020, during the pandemic, with a median organizational response rate of 34 percent, and it measured stated intentions rather than actual behavior. Intentions reported during a crisis may overstate what people later do.

Evidence on actual behavior is narrower but more direct. In a longitudinal study at Mayo Clinic that used administrative and payroll records rather than self-report, higher emotional exhaustion and lower satisfaction among physicians in 2011 were associated with actual reductions in full-time equivalent (FTE) effort over the following 24 months.3 That is one organization and an association, not proof of cause, but it shows that burnout-related loss of clinical effort can be measured in a health system's own data. Nationally, burnout has eased from its pandemic peak, yet 45.2 percent of US physicians still reported at least one symptom in 2023.4

Cost estimates are modeled rather than observed. Han and colleagues estimated that about 4.6 billion dollars a year in costs related to physician turnover and reduced clinical hours is attributable to burnout in the United States, roughly 7,600 dollars per employed physician per year.5 The American Medical Association puts the organizational cost of replacing one physician at 500,000 dollars to more than 1 million dollars.6 CEREVITY's own company-produced white paper, Burnout and the Physician Shortage: The Link Leaders Miss, argues the strategic version of this case, that retention is the workforce lever with the shortest lead time.7 It is offered here as the author's organizational perspective, not as independent evidence.

Two planning implications follow. First, reduced hours are a workforce loss that a turnover report does not show: five physicians who each drop a clinical day represent roughly one full-time physician without a vacancy being posted. Second, because the cost of inaction is already being paid, a retention intervention should be judged against the value of the hours it keeps.

Retention is a systems problem first

Burnout is driven mainly by the conditions of work. The National Academies of Sciences, Engineering, and Medicine concluded in 2019 that clinician burnout requires a systems approach addressing workload, work processes, technology, organizational culture and leadership.8 The evidence for specific drivers is substantial. In a time and motion study across 4 specialties, physicians spent 27.0 percent of their office day on direct clinical face time with patients and 49.2 percent on electronic health record and desk work, with another 1 to 2 hours of after-hours work most nights.9 In a survey of 2,813 physicians at one organization, each 1-point increase in a supervisor's composite leadership score was associated with a 3.3 percent decrease in the likelihood of burnout.10

A retention strategy that counts hours should therefore start with the levers that shape everyday work: staffing ratios and coverage, schedule control and predictable call, administrative and documentation burden, compensation models that do not penalize reduced but sustainable effort, and the selection and training of physician leaders. Access to confidential mental health care belongs inside that strategy. It addresses a different problem, the physician who is already struggling and not seeking help, and it cannot substitute for fixing the work itself.

Why existing support goes unused

Most health systems already offer an employee assistance program, peer support or a wellbeing office, and physician use is often low. One measured reason is fear of professional consequences. In a national sample of 5,829 physicians, nearly 40 percent said they would be reluctant to seek formal care for a mental health condition because of concerns about repercussions to their medical license, and reluctance was higher in states whose licensure applications asked about more than current impairment.11

The Federation of State Medical Boards recommended in 2018 that boards reconsider whether probing mental health questions are needed and, where retained, limit them to current impairment.12 Hospitals face the same issue through credentialing. As of September 25, 2026, the Dr. Lorna Breen Heroes' Foundation reported that 1,087 hospitals had verified that their credentialing applications are free of intrusive mental health questions.13 Reviewing these applications is one of the least expensive retention steps a system can take.

Confidentiality: three separate design questions

Physicians weighing whether to use a program are, in effect, asking who could ever learn that they did. In my view, that question is best answered by separating three design choices that are often blurred together. None of them, alone or combined, guarantees confidentiality.

1. Whether an insurance claim is created. Care billed through the organization's health plan generates claims data containing diagnosis and procedure codes, which in a self-insured plan is administered on the sponsor's behalf. Care that is paid directly or funded in aggregate can avoid creating a claim for that service. It does not eliminate records: the treating clinician still keeps a clinical record, as licensure and state law require. Providers who conduct standard electronic transactions are HIPAA covered entities;14 a provider who never bills electronically may fall outside that definition, although state confidentiality law and professional ethics still apply. Leaders should ask any provider, internal or external, which rules govern its records.

2. Whether clinical records are separated from employer systems. Care delivered by colleagues, in the organization's own electronic health record, inside the same credentialing structure, carries a different exposure from care documented in a separate system the employer cannot access. External care contracted outside the organization's reporting lines is one way to achieve that separation. Others include a dedicated, access-restricted record for an internal faculty and staff assistance program, peer support programs with clear confidentiality protections, and referral arrangements with community clinicians. Each has trade-offs in cost, continuity and access.

3. Whether identifiable utilization information reaches the employer. Eligibility files and single sign-on into a vendor portal can create logs, inside the employer's own identity systems, showing who accessed a mental health benefit. A utilization report broken out by department can identify an individual when a cell holds one or two people. Contracts can require eligibility to be verified once by the vendor, aggregate-only reporting, and suppression of small cells; the Centers for Medicare and Medicaid Services, for example, does not permit any cell containing a value of 1 to 10 to be reported directly.15 The wellbeing office should not hold individual-level data.

Even with all three addressed, legal exceptions remain, including subpoenas and court orders, mandatory reporting and duty-to-protect obligations, and questions a licensing board may ask. Programs should state these limits plainly rather than promise absolute confidentiality.

Measuring clinical FTE retention and program effect

Health systems can measure hours retained with data they already hold, without touching anyone's clinical information.

1. Retained clinical FTE. At a fixed baseline date, take each physician's contracted clinical FTE from payroll or scheduling records. Twelve months later, sum the current clinical FTE of the same cohort and divide by the baseline total. Report two components separately: FTE lost to departure and FTE lost to reductions among physicians who stayed. Repeat by department, keeping small-cell suppression.

2. Intent to reduce hours and intent to leave. Survey at least annually with the same items used in the published literature, administered by a third party with anonymous responses, so results can be compared over time and with peers. Treat intentions as an early signal, then check them against the retained-FTE measure.

3. Burnout and professional fulfillment. Use a validated instrument consistently from year to year, reported only in aggregate.

4. Access measures for any mental health resource: time to first appointment and continuity with the same clinician, reported by the provider as aggregate counts.

5. Credentialing alignment: an audit of every application, privileging form and peer review questionnaire against current-impairment standards, with the result communicated to the workforce.

Evaluating whether a given program changes these outcomes is harder, and claims should be modest. No published independent evidence yet shows that confidential external care programs, private-pay or otherwise, improve physician retention. A system can still learn something. Staggered rollout by department, with later departments serving as comparisons, or an interrupted time series against a pre-launch baseline, allows aggregate comparison of retained FTE and survey measures. The confidentiality rule for evaluation is simple: never link individual program use to individual HR, FTE or performance records. Compare groups, not people, and have any analysis that requires identifiable data performed by an independent evaluator under a data use agreement. Participation counts and webinar attendance should not be headline measures, because a program can report strong engagement while the physicians most at risk never use it.

Conclusion

Training more physicians remains essential, but it cannot replace clinical hours being withdrawn now by physicians who remain employed. Health systems control much of that variable. The strongest levers are the conditions of work: staffing, schedules, administrative burden, compensation and leadership. Confidential access to care is a necessary complement for physicians already struggling, and its value depends on honest design: deciding whether claims are created, separating clinical records from employer systems, restricting identifiable utilization data, and stating the limits that remain. Measuring retained clinical FTE alongside headcount would let leaders see the loss and test what reduces it.

Declarations

Conflict of interest: The author is co-founder and chief technology officer of CEREVITY Health, Inc., a private-pay therapy network that offers confidential care to physicians and medical groups. Reference 7 is a company-produced white paper and has not been peer reviewed. No other conflicts are declared.

Funding: None.

Author contributions (CRediT): Elijah Fernandez: conceptualization, writing (original draft), writing (review and editing).

Data availability: No new data were generated or analyzed for this commentary. All figures are drawn from the published sources cited.