A quality of earnings report tells you whether last year's EBITDA was real. It does not tell you whether the clinicians who produced it will still be there in month nine, or whether the payer mix behind it survives contact with a new owner.
Both statements can be true at once: the numbers are clean, and the deal is mispriced. In behavioral health that combination is common enough to be the default. Earnings in an outpatient psychiatry or therapy platform are produced by a few hundred licensed people making individual decisions about where to work, how many patients to see, and how much documentation burden they will tolerate. A financial workstream can verify what those decisions produced. It is not designed to test whether they will be made the same way under different ownership.
That is the gap. We call the category clinical integration risk: the risk that the clinical and operational engine behind the financial statements does not transfer with the asset. It is the risk financial diligence does not price.
What quality of earnings is built to do
Quality of earnings work is rigorous and necessary. It normalizes reported earnings, tests revenue recognition, isolates one-time items, examines working capital, and scrutinizes the add-backs a seller has proposed. Done well, it tells a buyer what the business actually earned and how much of that is repeatable on the same inputs.
The words doing the work in that sentence are on the same inputs. A quality of earnings analysis holds the operating model constant. It assumes the clinicians, the schedule, the payer contracts, the documentation workflow, and the referral pattern that generated last year's revenue continue to exist. In most industries that assumption is reasonable for the twelve months after close. In behavioral health it is the assumption most likely to break, because the asset is the workforce and the workforce is mobile, licensed, and in demand.
Aluria does not perform quality of earnings. We provide the operational overlay that sits on top of it, and the two read very differently.
Five things the clinical read surfaces
1. Revenue concentration in individual clinicians
A platform with 40 providers and no single-payer concentration can still have severe concentration risk if six clinicians generate a third of the contribution margin, three of them are within two years of retirement, and one of them is the reason the largest referral source sends patients. Financial diligence sees a diversified revenue base. Provider-level production data, compensation structure, and tenure curves show something else.
The test is not whether concentration exists. It is what happens to the model if the top decile of producers leaves in the first year, and whether the non-competes, compensation structure, and clinical leadership bench make that more or less likely than the base rate.
2. Earnings that depend on documentation and coding behavior
In behavioral health, revenue per encounter is a function of what the clinician documents. When a platform's collections per visit sit meaningfully above its peer set, there are two explanations: better operational discipline, or coding behavior that will not survive a new compliance regime. Only one of those is worth paying a multiple for.
A chart-level review against the billed code distribution answers the question in a way a revenue trend cannot. It also tells you whether the post-close compliance tightening a sponsor will inevitably apply carries a revenue cost nobody has put in the model.
3. Payer risk that lives in policy, not in the contract
Contracted rates are the easy part. The harder questions are whether the payers behind the revenue treat the service lines the same way across the states the platform operates in, how much revenue sits behind prior authorization, what the re-authorization approval rate looks like, and how many providers are credentialed individually versus delegated. A platform that has been operating on out-of-network rates, single-case agreements, or a favorable medical policy has an earnings base with a policy expiry date on it.
4. Capacity that does not exist
Growth cases in behavioral health are usually capacity cases: more visits per clinician, more clinicians per site, more sites. Each of those has a physical constraint — rooms, supervision ratios, credentialing timelines, clinician supply in a given market. Utilization data at the provider and site level tells you how much headroom there actually is. Frequently the answer is that the first year of the growth plan requires hiring the platform has never demonstrated it can do.
5. Integration cost that arrives before the savings
Four EHRs, three billing vendors, two credentialing processes, and no standard supervision model is an operating structure, and consolidating it has a price in dollars, clinician hours, and temporary revenue cycle disruption. That cost lands in months one through nine. The savings land later, if the clinical model holds. Sequencing the two correctly is the difference between a value creation plan and a wish.
How the two workstreams should fit together
They are complementary, and they should be run in parallel with a shared data room and explicit hand-offs. The financial workstream establishes what the business earned. The clinical and operational workstream establishes whether the engine that earned it transfers, at what cost, and on what timeline.
Where the two disagree, the disagreement is the finding. A clean quality of earnings paired with a fragile clinical base is not a reason to walk; it is a reason to reprice, restructure the earn-out, or fund the retention plan properly in the sources and uses. Those are all better outcomes than discovering the same facts in month six, which is when they usually surface on their own.
Deals rarely fail on the numbers. They fail on clinical integration, provider retention, and operations after close.
What a findings memo should contain
An operational diligence deliverable has to be usable inside an investment committee process, which means it has to be short, specific, and organized the way the committee thinks. Four sections do most of the work:
- Clinical integration risk. Where the care model depends on individuals rather than systems, and what it takes to change that.
- Provider retention exposure. Named risk concentrations, the compensation and non-compete picture, and a quantified downside case.
- Reimbursement and payer risk. Policy exposure by payer and state, authorization dependence, and revenue cycle performance against the reported result.
- Operational readiness. Management depth, systems, and the realistic first-year capacity of the platform to absorb both integration and growth.
Each finding should carry a cost or a timeline. A risk without a number attached does not change a decision, and diligence that does not change decisions is documentation rather than diligence.
About the authorEdisa Shirley, PhD, LMHC, is Founder and Managing Partner of Aluria Advisory. She has led more than 100 post-acquisition integrations across a behavioral health platform that scaled to 300+ centers in 37 states. Start a confidential conversation.