OCA work group wrestles with limits of measuring behavioral health delivered in primary care
The California Office of Healthcare Affordability (OCA) work group on Aug. 20 discussed how to identify behavioral health services provided in primary care settings and debated expanding the list of primary-care provider taxonomies to include behavioral health professionals — a move some members warned would risk overstating what counts as primary-care spending.
The debate matters because the subset of behavioral health spending OCA counts as occurring in primary care will affect future benchmarks, investment decisions and how policy makers track integration efforts.
OCA staff described a four-part test they plan to use to include a claim in the behavioral health–primary care module: a primary behavioral-health diagnosis (except for screening/assessment codes, which are included regardless of diagnosis), a provider taxonomy classified as primary care, a primary-care place-of-service, and a service code in OCA’s outpatient professional primary-care subcategory. Debbie Lindes, OCA’s health care delivery system group manager, said, “to maximize accuracy, a claim must meet 4 tests to be included in the module.” Non‑claim payments that explicitly support primary-care/behavioral-health integration and a portion of capitation payments would also be included in the module when encounters meet those criteria.
Work group members raised limits of that approach. A central example: using place‑of‑service code “office” plus an expanded taxonomy that includes psychologists could capture a 30‑minute psychotherapy visit that occurs in a psychologist’s private office rather than in an integrated primary‑care team. Lindes warned, “there’s a chance that this type of claim would be inappropriately included in the module.”
Several members urged caution. A longtime participant identified in the meeting as Kevin said the combination of taxonomy, place of service and service code “seems none…can distinguish the independent behavioral health professional providing services from that instance where it’s truly integrated in a collaborative model.” Dr. Soni and other members argued they would prefer to err on the side of excluding ambiguous claims rather than risk an inflated estimate of primary‑care spending that would misrepresent integration.
OCA staff emphasized this is a measurement‑level question that does not remove services from the overall behavioral‑health tally: expanding the taxonomy affects only the subset counted as occurring in primary care. Mary Jo noted the module is modular and can be reported both ways: with and without the expanded taxonomy to show differences empirically.
The group also clarified how OCA will treat mobile clinic and long‑term‑care categories. OCA proposed removing mobile clinic services and long‑term care from the public reporting subcategory list because the Healthcare Payments Database (HPD) shows negligible current spending. Margaret Brandt, assistant deputy director for health system performance, said the board supported removing those subcategories but “urged us to use HPD data moving forward to continue to monitor mobile clinic services and crisis care as care delivery patterns continue to change.” In response to a question from Carrie, Brandt added that spending for mobile clinics and long‑term care will still be collected and included in overall reporting, “just not distinctly called out in a subcategory.”
The work group did not take a binding vote on changing the taxonomy. Instead, OCA said it will continue supplemental HPD analyses, seek public comments on the proposed definition (the public comment period was announced to close Sept. 3), and return to the work group and the OCA board with findings and proposed next steps. Staff also reported ongoing regular meetings with the California Department of Health Care Services (DHCS) to explore ways to capture county behavioral‑health spending and said an update to the work group is expected “sometime this winter.” OCA plans to finalize the definition for its 2026 data submission guide by October and brief the full board later in the year.
What remains unresolved are coding and data limitations that make it difficult to distinguish integrated behavioral‑health visits from independent behavioral‑health care delivered in office settings; participants recommended empirical audits of samples of claims in future HPD analyses before any taxonomy expansion is finalized.