Behavioral Health: Build In-House or Partner?
Building behavioral health in-house gives a practice full control but requires hiring clinical staff, standing up billing, and absorbing the ramp-up risk. Partnering with a vendor is faster to launch and shifts staffing and compliance work off the practice, at the cost of some control and a share of the revenue. The right choice depends on a group's size, capital, and risk tolerance.
Most primary care groups reach the same conclusion: their patients need behavioral health support, and referring out isn't working. The harder question is *how* to add it. Both paths — build and partner — can deliver the same evidence-based model. They differ mainly in who carries the staffing, billing, and startup risk.
Building means the practice owns the whole program. For a Collaborative Care Model (CoCM) program, that typically includes:
The upside is total control: the practice sets the workflow, owns the patient relationship end to end, and keeps all the reimbursement. The cost is that it's a real operational build. A single care manager can only hold so many active patients, so the economics don't work until the panel fills — which can take months.
Partnering means a vendor supplies some or all of the program's moving parts — commonly the care managers, the consulting psychiatrist, the registry software, and the billing infrastructure — and embeds them into the practice's existing workflow.
The practice still owns its patients and its primary care clinicians still prescribe. What changes is that the behavioral health staffing, the psychiatric coverage, and the compliance overhead sit with the partner. In most arrangements the partner takes a share of the collaborative care revenue or charges a fee, rather than the practice keeping 100%.
The upside is speed and reduced risk: someone else recruits the hard-to-hire roles, maintains the registry, and keeps up with billing rule changes. The trade-off is less direct control over hiring and workflow, and a revenue split.
Staffing is usually the deciding factor. Behavioral health care managers and consulting psychiatrists are in short supply, and a small practice recruiting one or two of them competes against health systems for the same candidates.
A practical middle question: does the group have enough eligible patients to keep a care manager's panel full? If not, a shared or partner-supplied care manager may be the only model that pencils out.
Speed. Partnering is generally faster to launch — often weeks to a few months — because the staffing and billing pieces already exist. Building takes longer: recruiting alone can run a quarter or more before the first patient is enrolled.
Financial risk. Building front-loads cost. The practice pays salaries before the panel is full and before billing revenue catches up, so it carries the ramp-up loss. Partnering usually converts that fixed cost into a variable one — the practice pays as revenue comes in — which lowers the downside if enrollment is slower than hoped.
Compliance risk. CoCM billing has specific documentation and time-tracking requirements, and the rules change. In-house, the practice owns audit exposure. With a partner, that burden is largely shared or shifted, though the practice is still the billing entity and should confirm how liability is allocated.
The billing codes are the same either way — collaborative care is a covered benefit under Medicare and, in many states including New York, under Medicaid. The difference is who does the work and who keeps the margin.
The honest framing is a volume question: at high patient volume, building tends to win on total margin; at low or uncertain volume, partnering de-risks the economics.
There's no universal answer, but a few questions usually settle it:
1. Scale. How many eligible patients can realistically be enrolled? Enough to keep a full-time care manager busy points toward building; a thin or uncertain pipeline points toward partnering. 2. Capital and risk tolerance. Can the group absorb months of salary before revenue catches up? If not, partnering spreads the risk. 3. Hiring reach. Can the practice actually recruit and retain a care manager and a consulting psychiatrist in its market? If those roles are hard to fill, a partner's staffing may be the deciding advantage. 4. Control. How much does the group want to own the workflow and patient experience directly versus hand pieces to a vendor? 5. Timeline. Is there pressure to launch this quarter, or room to build over a year?
Many groups also take a hybrid path — partner to launch and learn the model, then bring pieces in-house as volume grows and the economics shift. Build, partner, and hybrid are all legitimate; the fit depends on the group's specifics, not on which is "better" in the abstract.
It depends on volume. Building has higher fixed costs (salaries, billing setup) but keeps all reimbursement, so it's cheaper per patient at high volume. Partnering has little upfront cost but a revenue share, making it cheaper when volume is low or uncertain.
Partnering typically launches in weeks to a few months because staffing and billing already exist. Building usually takes longer — recruiting a care manager and consulting psychiatrist alone can run a quarter or more before the first patient enrolls.
Yes. In a collaborative care partnership, the practice retains its patients and its primary care clinicians continue to prescribe and manage treatment. The partner typically supplies the behavioral health care team and infrastructure that embed into the practice's existing workflow.
Many practices do exactly this. Partnering first lets a group launch quickly and learn the model, then bring staffing or billing in-house as patient volume grows and the economics favor ownership. Confirm any transition terms in the partner agreement upfront.
The codes are the same — collaborative care is billed through established Medicare and, in many states, Medicaid codes regardless of model. What differs is who does the documentation and billing work and how the reimbursement is split. Confirm coverage and code specifics with your payers.
