A 44-year-old internist is sitting in her car in the hospital parking garage after her last patient of the day, googling "physician nonclinical careers" for the fourth time this month.
She is not burned out in the dramatic, crisis sense. She is mid-career, competent, respected by her colleagues, and quietly certain that she does not want to be doing exactly this for another twenty years. She has heard that people go into medical affairs at pharmaceutical companies. She has heard that some become hospital administrators. She has heard, vaguely, that a former colleague from residency now works somewhere in digital health. She does not actually know what any of those jobs are like day to day, what they pay compared to what she makes now, or whether the person who made that move regrets it.
She buys a book. She reads a few blog posts written by career coaches who are, she notices, all selling a course. She considers reaching out to her old co-resident, the one who went into industry, but they have not spoken in six years and it feels strange to cold-message someone she barely remembers to ask them to justify their career choice to her.
What she actually wants is simple and specific: a mid-career internist, from a specialty like hers, who made a move like the one she is considering, two or three years ago, who would tell her honestly what it is actually like, what it pays, and whether they would do it again. That person exists. Statistically, given how many physicians are making exactly this move right now, there are probably dozens of people who match her situation closely. She has no way to find a single one of them.
Nearly half the physicians in this country are weighing the most consequential career decision of their professional lives with no structured access to the one thing that would most improve it: a verified peer who has already made the specific move they are considering.
The scale of what is happening right now
This is not a slow-moving background trend. The numbers describe an acceleration.
Forty-four percent of physicians surveyed say they are "somewhat" or "very" likely to leave clinical practice entirely within the next two years, according to Doximity's 2026 Physician Compensation Report. Among overworked physicians specifically, 66 percent report considering a career change, and 46 percent are now contemplating early retirement, up sharply from 34 percent the prior year.
That jump, 34 to 46 percent in a single year, is the detail that separates this from a stable, long-running attrition pattern. Something is moving quickly. The same survey found that 81 percent of physicians said they would still choose medicine if starting over, and yet 63 percent would not recommend medicine as a career to their own children, a striking split that suggests the dissatisfaction driving this wave is specific to how medicine is currently practiced, not a rejection of medicine itself.
Physician turnover already runs at a median 7.3 percent annually, according to the Association of Staff Physician Recruiters' 2025 Benchmarking Report, with median time-to-fill running 118 days and offer-acceptance rates falling from 83 to 71 percent year over year. Each departure a system did not anticipate or plan for costs on the order of $770,000 to $1.2 million per search cycle in recruitment and vacancy expense, a figure this series has documented in detail elsewhere in the context of physician recruiting and staffing.
This is a different problem than burnout, and a different problem than retirement
It is worth being precise about what this article is not about, because two adjacent failures are easy to conflate with it.
This is not the burnout crisis itself. Burnout is a real, well-documented driver of physicians' desire to leave, and it has its own extensive literature and its own set of interventions aimed at the workplace conditions producing it. This article assumes burnout, overwork and administrative burden as background context, already established elsewhere, and asks a narrower question: once a physician has decided she is seriously considering leaving, what happens next, and how well is she supported through the actual decision and transition.
Nor is this about full retirement and re-entry, the loss of institutional knowledge when senior physicians retire, or the barriers facing physicians who want to return to practice after time away. That is a distinct, previously documented failure in this series, concerned with what is lost when experience exits the field entirely and how hard it is to come back.
This article is about the specific, large, and growing population of mid-career physicians moving sideways, not out of medicine's usefulness to society, but out of direct clinical practice, into industry, administration, consulting, or a hybrid non-clinical role, while still relatively early enough in their careers that this is a genuine pivot rather than a retirement. That population has almost no structured peer infrastructure at all.
What actually happens when a physician tries to make this move
The current process, reconstructed from how physicians describe it, is close to accidental by design.
A physician considering leaving clinical practice searches online, reads generic career-coaching content, and, if she has the money and inclination, buys a course or hires a physician-specific career coach. If she is fortunate, she happens to know one or two people personally who made a similar move, discovered by luck rather than by any systematic search. There is no way to search specifically for "a mid-career internist who became a medical director at a health-tech company in the last three years and is willing to talk about it."
This matters because the physician-coaching industry that has grown up around this gap has, structurally, a commercial incentive that runs against solving it for free. A course or a paid coaching program sells better when the alternative, a free conversation with someone who has already done exactly this, is unavailable. That is not a criticism of any individual coach's intentions; it is simply the economics of the market as it currently exists, and it means the industry best positioned to fill this gap has limited incentive to build the version that would make itself less necessary.
Alumni and residency networks, examined elsewhere in this series for their role in referral relationships, are built around clinical referral and donor relationships, not around tracking who left clinical practice and where they went. LinkedIn shows a current job title, which is useful for confirming someone made a move, but nothing on the platform signals whether that person is willing to talk candidly, medicine-specific, verified, about how the transition actually went.
The structural failure: the precedent exists, and nobody tracks it
The physicians who have already made a given transition are the single most valuable resource for someone considering it, and they are completely unfindable at the moment they are most needed.
Follow why each plausible source of this connection falls short.
Physician career-coaching businesses benefit from the search friction, not from removing it. A free, structured peer-connection service is a direct substitute for a paid coaching product, and no business builds the thing that cannibalizes its own revenue voluntarily.
Medical societies and residency alumni offices have no mechanism for tracking members who left clinical practice. Once a physician stops practicing clinically, she typically drops off the radar of exactly the institutions that would otherwise be positioned to reconnect her with current members facing the same decision.
Doximity and LinkedIn record a job title, not a willingness signal. Neither platform has built a verified, opt-in "ask me about my transition" field tied to medical credentials, the specific piece of infrastructure this problem requires and neither currently offers.
Employers who lose a physician to this kind of transition have no incentive to help the next one leave smoothly. If anything, the opposite: a well-informed, low-friction exit is not obviously in a current employer's short-term interest, even though a chaotic, abrupt exit is measurably more disruptive and costly than a planned one.
The result is a precedent pool that exists, provably, given how many physicians are making these moves right now, and is entirely unindexed, unverified and unreachable by the next physician who needs it.
What would actually work
A structured profile of the specific transition being considered, matched to verified peers who made it. Specialty, career stage, and target next chapter, matched against members who have already made that exact combination of moves and opted in to answer questions, not a generic "ask a career coach" form.
Candor as the explicit design goal, not an accidental byproduct. The value of this kind of connection collapses if it becomes another venue for polished, promotional success stories; the mechanism needs to protect and actively encourage honest accounts of what a transition actually paid, cost, and felt like, including regret where it exists.
A registry of actual outcomes, not just intentions. What was considered, what was chosen, and how it went financially and professionally, reported by the people who lived it, is a dataset that does not currently exist anywhere and would let future physicians see real patterns rather than anecdotes.
Positioned as a complement to paid coaching, not a replacement for it. A free verified precedent pool and a paid coaching program serve different needs, the former factual and experiential, the latter structured and process-driven, and the two can coexist without either undermining the other.
Retired and senior physicians recruited deliberately as a supply source. This population, already identified across this series as an underused source of mentoring and reference capacity elsewhere in medicine, includes many who made exactly this kind of transition years ago and have the time and distance to talk about it candidly.
Boundaries around financial and legal advice made explicit from the start. A precedent conversation is not a financial plan or a contract review; the structure needs to make clear that peers are sharing experience, not providing licensed advice, so that candid conversation is not chilled by liability concern on either side.
Built into everyday membership, not sold as a crisis add-on. A physician does not wake up one day newly needing this. She arrives at the decision gradually, and the value of a precedent network is highest when it is already part of her professional infrastructure before she needs it urgently.
What you can do now
If you are a mid-career physician considering a transition
Name the specific transition you are considering, not the vague category. "Leaving medicine" is too broad to act on. "Becoming a medical director at a health-tech company" or "moving into hospital administration in my current system" is specific enough to actually search against, even informally through your own network.
Reach out to the person you already half-remember. The old co-resident who moved into industry, the friend of a friend who left for consulting. The awkwardness of reconnecting to ask a genuine, respectful question is almost always smaller in practice than it feels in anticipation, and this series' broader argument throughout is that most of medicine's informal knowledge sharing already runs exactly this way, imperfectly, through exactly this kind of reconnection.
Ask specifically about the first year, not just the decision. What the new role actually paid in year one, what was harder than expected, and what they wish someone had told them before they made the move; this level of detail is what generic career content cannot provide and what a genuine precedent conversation can.
If you already made this transition
Consider making yourself findable, even informally. A short note on your own professional profile, or a willingness to take a call from a former colleague considering the move you already made, costs you little and is precisely the kind of information this article establishes as currently unavailable to the physicians who need it.
Be honest about the parts that did not go as planned. The value of your account is disproportionately concentrated in the candid, unflattering details, not the polished summary version.
If you lead a department or health system
Distinguish planned transitions from abrupt departures in your own retention data. If your system does not currently track whether a physician's exit was sudden or well-signaled in advance, you are missing the data that would show whether earlier, better-informed transitions reduce the disruption and cost of turnover your organization already measures at $770,000 to $1.2 million per search cycle.
Consider that a well-supported exit is not automatically against your interest. A physician who leaves on good terms, with a realistic understanding of her next chapter reached well in advance, is less likely to leave abruptly and more likely to remain a referral source, an alumnus, or even a future part-time or consulting contributor to your organization.
Frequently asked questions
What percentage of physicians are considering leaving clinical practice? Forty-four percent say they are "somewhat" or "very" likely to leave clinical practice entirely within the next two years, according to Doximity's 2026 Physician Compensation Report. Among physicians who report being overworked, 66 percent say they are considering a career change.
How many physicians are considering early retirement? Forty-six percent of overworked physicians are now contemplating early retirement, up sharply from 34 percent the previous year, according to the same 2026 Doximity survey, one of the fastest year-over-year shifts identified anywhere in this series' data.
How can I talk to a physician who left my specialty for a specific nonclinical role? Currently, mostly through personal connections discovered by chance rather than any systematic search; no verified, searchable network currently matches physicians considering a specific transition to peers who have already made that exact move and are willing to discuss it candidly.
What nonclinical career options do physicians have? Common paths include medical affairs and other industry roles at pharmaceutical or device companies, hospital and health-system administration, consulting, health-technology leadership, and part-time or hybrid clinical-nonclinical arrangements. The quality of information available about any specific path currently depends almost entirely on a physician's personal network rather than any structured resource.
Does an unplanned physician departure cost more than a planned one? Direct research comparing planned versus abrupt physician departures specifically is limited, but physician turnover generally is documented to cost $770,000 to $1.2 million per search cycle at a median 7.3 percent annual turnover rate, according to the Association of Staff Physician Recruiters' 2025 Benchmarking Report, and unplanned, burnout-driven exits are reasonably understood to carry more organizational disruption than exits signaled well in advance, though this specific comparison warrants further dedicated study.
Is this the same as physician burnout? No. Burnout and administrative overload are well-documented drivers of physicians' desire to leave clinical practice, but this article addresses a narrower and separate gap: once a physician has decided to seriously consider a transition, she has almost no structured, verified way to find a peer who has already made the specific move she is weighing, distinct from the underlying reasons she wants to leave in the first place.
The bottom line
Close to half the physicians in this country say they are likely to leave clinical practice within two years, and the share newly considering early retirement jumped from roughly a third to nearly half in a single year. This is not a slow drift. It is an accelerating, present-tense decision point for a huge share of the profession.
And at the exact moment a physician is weighing the most consequential career decision she may ever make, walking away from a decade or more of training, earning potential, patient relationships and professional identity, the single most useful resource, a verified peer who has actually made the specific move she is considering, is entirely unfindable through any structured channel. She is left with generic career-coaching content, a course somebody is selling her, and whatever she happens to remember about people she trained with years ago.
The physician-coaching industry that has grown up to fill this gap has a business reason not to close it for free. Medical societies track clinical careers, not the ones that left. Doximity and LinkedIn show a job title and nothing about whether that person would actually talk to a stranger about how the move really went.
The internist sitting in the parking garage, googling the same search for the fourth time this month, is not lacking initiative. She is missing an infrastructure that was never built. Nearly half the physicians in this country are weighing the most consequential career decision of their professional lives with no structured access to the one thing that would most improve it: a verified peer who has already made the specific move they are considering.
Part of a series on the missing professional infrastructure of healthcare. Previously: The Non-Clinical Curbside
Evidence note: sources include Doximity's 2026 Physician Compensation Report for the 44 percent leave-within-two-years figure, the 66 percent career-change-consideration figure among overworked physicians, and the jump from 34 to 46 percent in early-retirement consideration; and the Association of Staff Physician Recruiters' 2025 Benchmarking Report for median physician turnover (7.3 percent), time-to-fill (118 days) and offer-acceptance trends. A commonly cited claim of a mean physician clinical-exit age of 48.1 could not be traced to a verifiable primary source in this research pass and is therefore not used as a figure in this article; readers should treat that specific number, if encountered elsewhere, as unverified pending a confirmed citation. Per-departure replacement cost figures ($770,000 to $1.2 million) are drawn from recruitment-cost research cited elsewhere in this series and should be read as estimates rather than precise universal figures, since actual cost varies by specialty and market.
