Six weeks into her interim role, a hospital CFO is sitting with a vendor contract that commits the system to another four years and a service line that was quietly wound down eighteen months before she arrived.
The board minutes record both decisions in a single sentence each: a vote count, a date, an approval. They do not record what else was on the table, why the losing option lost, or who on the executive team argued hardest against the path that was ultimately chosen.
She calls the departed CEO's old assistant, who does not know. She calls a board member who remembers "concerns about volume projections" but not the numbers behind them. She asks two vice presidents, who each give her a different, partial version of the story, neither of which quite explains why the contract runs the length it does.
Eventually she does what almost every incoming executive in her position does: she brings in a consultant, at real cost, whose entire mandate is to reconstruct a decision that someone in this building already understood perfectly well two years ago, before they left for a different system entirely.
The vote is on the record. The reasoning behind it left the building the day the person who made the call did, and nothing in hospital governance was built to stop that from happening.
Turnover is not the exception, it is the baseline
Start with how often this scenario actually occurs, because the honest number is higher than most people outside hospital governance would guess.
Hospital CEO turnover held at 16 percent in 2022, according to the American College of Healthcare Executives, the third consecutive year at that rate. That is not a spike. That is a sustained baseline: roughly one in six hospital CEOs leaves their post in any given year, year after year.
The trend has not eased since. Hospital CEO exits climbed to 111 in full-year 2025, up 7 percent from 104 the year before, per Becker's Hospital Review's ongoing executive-move tracking.
CEOs are not the whole picture. CFO tenure at the ten largest US health systems averages 4.7 years, per Crist Kolder Associates' 2025 Volatility Report. CIO tenure at those same large systems runs "just under three and a half years," per Becker's reporting.
Layer CEO, CFO, and CIO churn together, and add CNO and CMO turnover that runs on broadly similar cycles, and the plausible picture for a typical health system's full executive team is turnover somewhere north of once every five years, on average, across the whole C-suite. Every one of those departures carries the same uncaptured context out the door.
What actually leaves when an executive leaves
The distinction that matters here is not between what a hospital knows and does not know. It is between two very different things a hospital's governance apparatus keeps: the decision, and the reasoning behind it.
Board minutes, contracts, budgets, and org charts capture the decision. What was approved, when, and by what vote. This is not a small thing; it is a legally necessary record and hospital boards generally keep it well.
What board minutes almost never capture is decision provenance: the alternatives that were seriously considered and rejected, the internal disagreement that shaped the final version of the deal, the specific person who can explain, from memory, why a service line was closed this way rather than that way, and what would have to change for the original decision to no longer make sense.
That second category lives nowhere except in the heads of the people who were in the room. When they leave, it does not get transferred. It does not get summarized into a file. It simply stops being accessible to anyone still working at the institution.
Why this is a distinct problem, not a subset of poor documentation
It is tempting to read this as ordinary poor recordkeeping, fixable with better minutes. That undersells the actual mechanism.
Board deliberation is often, appropriately, candid in ways that are never meant to become a permanent public record: a frank assessment that a physician group's leadership was unreliable, a blunt read on a competitor's financial weakness, an executive's private doubt about a project they nonetheless voted to approve for political reasons. None of that belongs verbatim in board minutes, and no reasonable governance reform would put it there.
The realistic fix is not "write everything down." It is preserving reachability: making sure a successor can find and talk to the specific person who was there, under confidentiality terms that respect what should and should not be shared, rather than inheriting a sealed vote with no path back to the reasoning at all.
Why nobody owns it
Every institution that touches an executive transition has a job. None of those jobs is decision-provenance preservation.
Executive search firms (WittKieffer, Cejka Search, B.E. Smith, and peers) are paid to fill the seat. Their success metric is placing a qualified successor, not ensuring that successor inherits the reasoning behind the decisions they will now own. Nothing in their engagement structure rewards building a context-transfer product, and doing so would be a different business entirely.
Boards are volunteer or part-time bodies, meeting on a fixed cadence, with fiduciary duties that center on oversight and major approvals, not on maintaining a living decision log between meetings. Asking an already-stretched volunteer board to also become an institutional archivist is asking for a role most boards are not structured to perform.
Consultants monetize the reconstruction work directly. When a successor needs to understand "why did we structure the JV this way," a consulting engagement to rebuild that history is a billable project, not a preventable cost. There is no commercial incentive on that side of the table to make the problem smaller.
ACHE and CHIME, the major professional associations for hospital executives and health-system CIOs, are institution-membership networks focused on education, benchmarking, and professional development. Neither has built, or is structured to build, a person-keyed decision-provenance product.
Governance research itself confirms the gap is being felt, not just theorized. A Governance Institute and NRC Health report found that hospital board governance capability has not kept pace with system complexity, according to Becker's 2026 coverage. That is a description of exactly this failure mode from the inside: governance structures straining to keep up with decisions whose complexity has outgrown the recordkeeping built to track them.
The cost of re-diagnosing decisions blind
Here is a way to size what this actually costs, stated as an illustrative estimate rather than a measured industry figure, because no audited total exists for this specific cost category.
There are roughly 6,100 hospitals and health systems in the US. If even half experience one C-suite departure in a given year, a conservative read once CFO, CIO, CNO, and CMO churn are counted alongside CEO churn, and each successor spends a conservatively estimated 100 hours of executive and board time re-diagnosing "why did we do this" on major standing decisions (vendor contracts, facilities commitments, service lines) before acting, at a blended executive and board opportunity cost of roughly $300 an hour, the arithmetic works out to on the order of $90 million a year in re-diagnosis labor alone. That figure does not include the separate, likely larger cost of outright strategy reversals that a documented decision record might have prevented in the first place. Treat it as a plausible order of magnitude built from stacked assumptions, not an audited total.
Whatever the precise number, the direction is not in dispute. Every hospital executive who has lived through a transition on either side of the departure has a version of this story, and the pattern behind it, uncaptured rationale, blind successors, expensive reconstruction, repeats with metronomic regularity across an industry that turns over its top leadership at a documented, sustained rate.
Why now, specifically
Two forces are compounding at the same time, and they are pushing in the same direction.
Post-pandemic margin pressure and elevated distressed hospital M&A activity are accelerating executive churn precisely when strategic continuity matters most, forcing boards to make leadership changes under financial duress rather than on a planned cadence.
At the same time, boards facing that pressure increasingly install outside or interim leaders who are unfamiliar with the institution from day one, which widens the exact context gap this article describes rather than narrowing it. An external hire brings fresh perspective, which boards often explicitly want, and simultaneously has zero prior exposure to why the last five years of major decisions were made the way they were.
What would actually work
A decision-provenance record keyed to the individual, not the institution. The information that matters, who argued what and why, survives only if it is attached to a person who can be reached after they leave, under terms they control, rather than filed inside an institutional system that becomes inaccessible the moment they walk out the door.
Executives register their own reasoning, not employer-owned material. A departing CEO documenting her own thinking on a major decision, in her own words, redacted to what she is free to share, sidesteps the confidentiality and ownership problems that make a full board-record archive legally fraught.
No confidential board material and no material non-public information. Any workable version of this has to draw a hard line well inside what board confidentiality rules and securities-adjacent disclosure limits (for public or publicly-traded-adjacent systems) already require, so it augments governance rather than creating new exposure.
Routed to a successor or a relevant peer, not broadcast. The value is not publication. It is a successor's ability to reach the specific person who made an analogous call, either at their own institution if reachable, or at a peer institution if not, rather than starting from a blank page.
Built for the transition moment, not as ongoing surveillance. The record exists to answer "why did we do this" when a successor actually needs to know, not to monitor sitting executives' day-to-day decision-making, which would create exactly the kind of chilling effect that discourages candid deliberation in the first place.
Independent of any single search firm or consulting relationship. Because the existing incentive structure rewards reconstruction as billable work, the fix has to sit outside that relationship, as something an executive builds for their own portable professional record rather than something a hiring or search process controls.
Realistic about the population size. The relevant population, hospital and system C-suite and board members nationally, is bounded at roughly 30,000 to 50,000 people. Any serious version of this has to be designed for a professional network at that scale, not a mass-market consumer product.
What you can do now
If you are an incoming or interim executive
Ask for the losing options, not just the winning one, in every major briefing. When someone explains a standing decision, ask specifically what else was considered and why it was rejected. That question surfaces the reasoning that board minutes never captured, faster than any document search will.
Identify the two or three people who were actually in the room, and talk to them directly. Org charts tell you titles. They do not tell you who argued for the losing side of a 6-4 board vote, and that person often has the most useful context of anyone.
Build your own decision log from day one. Every major call you make, and the reasoning behind it, is exactly what your own successor will eventually be missing. Starting the habit early costs little and compounds in value with every year you stay.
If you sit on a hospital board
Ask the departing executive, explicitly, to document their reasoning on the two or three biggest live decisions before they leave. This is a low-cost, high-value request that almost never happens as a matter of routine practice, and a board that asks for it systematically is doing something most boards currently do not.
Distinguish decision records from deliberation records in your own minds. You do not need, and should not want, a verbatim transcript of board debate preserved forever. You do need a way for a successor to reach the people who were part of that debate, which is a narrower and more achievable goal.
Name the re-diagnosis cost out loud at your next transition. If your system estimates even a fraction of 100 hours of executive time spent reconstructing "why did we do this" after your next executive departs, put a number on it and treat it as a governance line item, not an invisible cost of doing business.
If you are a departing executive
Leave a reasoning memo behind, even an informal one. A short, candid account of why the two or three biggest live decisions on your desk were made the way they were, written for your own successor and shared under whatever confidentiality terms your role requires, is worth more to the institution than almost anything else you could do in your final weeks.
Stay reachable, on your own terms. You are not obligated to answer every future call from a successor at an institution you have left. Many departed executives are, however, willing to answer a specific, bounded question about a decision they made, if there is a legitimate, professional way for that question to reach them.
Frequently asked questions
How often do hospital CEOs turn over? Hospital CEO turnover held at 16 percent in 2022, the third consecutive year at that rate, according to the American College of Healthcare Executives, as reported by Becker's Hospital Review. CEO exits climbed further to 111 in full-year 2025, up 7 percent from 104 in 2024.
What happens to institutional knowledge when a hospital CEO leaves? Formal records like board minutes, contracts, and budgets stay with the institution, but the reasoning behind major decisions, the alternatives considered, and who can explain the tradeoffs, generally leaves with the departing executive because no standard governance mechanism captures it separately from the vote itself.
How long does the average hospital CFO stay in the role? CFO tenure at the ten largest US health systems averages 4.7 years, according to Crist Kolder Associates' 2025 Volatility Report. CIO tenure at the same size of system runs just under three and a half years, per Becker's Hospital Review.
Do hospital boards have written CEO succession plans? Many do, but succession planning as commonly practiced focuses on identifying and preparing candidate successors, not on preserving the rationale behind the outgoing executive's major decisions. A Governance Institute and NRC Health report found board governance capability generally has not kept pace with health-system complexity.
What is decision provenance in healthcare governance? It refers to the record of why a decision was made, not just that it was made: the alternatives considered, the internal disagreement, and the people who can explain the tradeoffs behind a standing strategic decision. Board minutes typically capture the vote but not the provenance, which is why successors often cannot reconstruct the reasoning behind decisions they now must implement or reverse.
Why can't executive search firms solve this problem? Because their engagement and revenue model is built around filling the vacant seat with a qualified candidate, not around transferring the departing executive's decision rationale to the successor. Consultants are sometimes hired separately, at additional cost, specifically to reconstruct that missing context after the fact.
The bottom line
The CFO staring at a four-year vendor contract and a quietly closed service line is not dealing with a documentation failure in the ordinary sense. The documentation exists. The vote is recorded. What is missing is the layer underneath the vote: why this contract, why this length, why this service line and not another, and who in the building could once explain all of it from memory.
Hospital CEO turnover has held at 16 percent for years running, with exits climbing to 111 in 2025 alone. CFOs last under five years on average at the largest systems. CIOs last under three and a half. Every one of those departures takes the same kind of uncaptured reasoning with it, and nothing in hospital governance, not board minutes, not succession planning, not the search firms paid to fill the vacancy, was built to stop that from happening.
The realistic fix was never a fuller board record. Board deliberation is candid precisely because it is not meant to become a permanent public transcript, and no governance reform should try to make it one. The realistic fix is reachability: a way for a successor to find the specific person who made a specific call, and ask them directly, under terms that respect what should and should not travel.
Until that exists, every transition repeats the same sequence. A vote on the record. A departure. A successor working from a document that tells her what happened and nothing about why. And, more often than the industry likes to admit, a consultant hired to spend real money reconstructing what someone down the hall, or one system over, already knew for free.
Part of a series on the missing professional infrastructure of healthcare. Previously: The Border Handoff
Evidence note: hospital CEO turnover (16 percent, 2022, third consecutive year) is from the American College of Healthcare Executives, as reported in Becker's Hospital Review, August 2023. The 2025 CEO exit count (111, up 7 percent from 104 in 2024) is from Becker's Hospital Review, February 2026. CFO tenure (4.7 years at the ten largest systems) is from Crist Kolder Associates' 2025 Volatility Report, cited in Becker's, October 2025. CIO tenure (just under three and a half years) is from Becker's Hospital Review, April 2025. The finding that board governance capability has not kept pace with system complexity is from a Governance Institute and NRC Health report, cited in Becker's, April 2026. The $90 million a year re-diagnosis-cost estimate in this article is this publication's own illustrative arithmetic, built from the cited turnover rates plus assumed figures (100 hours per transition, $300 blended hourly cost, half of roughly 6,100 US hospitals affected annually) that have not been independently measured or audited; it should be read as a plausible order of magnitude, not a verified industry total. Nothing in this article is governance, legal, or financial advice for any specific institution.
