Use of North Carolina Medicaid Collaborative Care Billing Codes After Statewide Approval for Reimbursement

Access to behavioral health services remains limited and unevenly distributed in many parts of the United States (1). Furthermore, the COVID-19 pandemic has highlighted the need to improve access to effective behavioral health treatment across all age groups (2). Broad adoption of the collaborative care model (CoCM) has been proposed as one way to improve access to and quality of behavioral health care, promote physical health, and reduce overall health care costs (3, 4).

CoCM is an evidence-based paradigm that integrates behavioral health care with primary care and that is predicated on the following elements: patient-centered team care, population health (using a patient registry to ensure patient engagement and to track symptom improvement), measurement-based treatment to target behavioral health outcomes, evidence-based care, and accountability to the health system (quality assurance and documentation of patient improvement) (5).

To incentivize the utilization of CoCM in primary care settings, in January 2017 the Centers for Medicare and Medicaid Services approved reimbursement for Medicare CoCM billing codes (6). Soon after this decision, individual states began approving Medicaid CoCM billing codes, and North Carolina began reimbursing CoCM Medicaid billing codes in October 2018 (7).

This column examines the use of CoCM Medicaid billing codes in North Carolina in the 15 months after the codes were approved for reimbursement. The primary objective was to understand whether approval for reimbursement of these codes resulted in widespread use of CoCM.

The analytic data set was derived from North Carolina Medicaid 2018–2019 enrollment, institutional, and professional claims data obtained through the North Carolina Department of Health and Human Services (DHHS). The study cohort included all North Carolina Medicaid beneficiaries with at least one CoCM claim (Current Procedural Terminology [CPT] codes 99492, 99493, and 99494) for any behavioral health condition between October 1, 2018, and December 31, 2019. All analyses were performed with SAS, version 9.4. This study was approved by the Duke University Health System Institutional Review Board and by North Carolina DHHS.

From October 2018 through December 2019, a total of 915 patients among the approximately two million eligible beneficiaries of North Carolina Medicaid had at least one CoCM claim. Of the patients with CoCM claims, a majority were adult (73%), female (63%), and non-Hispanic White (53%) and lived in an urban county (76%). Pediatric patients were less likely to be non-Hispanic White (32%) and female (45%) compared with adult patients (61% and 70%, respectively). Among patients with a CoCM claim, the mean age of adult patients was 57.4 years and of pediatric patients was 11.0 years (Table 1).

The median number of CoCM claims for adult patients was two (first quartile [Q1]=1, third quartile [Q3]=5) and for pediatric patients was one (Q1=1, Q3=2). Across the 15 months of the study period, the median number of patients with at least one CoCM claim in a given month was 137 for adults and 26 for children. Overall, 90% of pediatric patients and 59% of adult patients had at least one initial-month claim (CPT code 99492), compared with 35% and 74%, respectively, who had at least one subsequent-month claim (CPT code 99493) (Table 1).

During October 2018, there were 112 total CoCM claims, all of which were for adult patients. In December 2019, there were 298 total CoCM claims, mostly for adult patients (see the online supplement to this column).

Claims data indicate that adoption of North Carolina Medicaid CoCM billing codes did not lead to high levels of utilization of these services in the first 15 months. Although it is possible that some North Carolina health systems provided CoCM to patients without billing for care, it is also possible that simply making CoCM billing codes eligible for reimbursement did not lead to early adoption of CoCM. Consequently, at least in North Carolina, approval of Medicaid CoCM billing codes alone may not lead to widespread adoption of CoCM.

Furthermore, among patients with CoCM claims, the median number of claims received for adult patients was two and for pediatric patients was one. In the Improving Mood–Promoting Access to Collaborative Treatment study, the largest CoCM randomized controlled trial to date, participants had access to CoCM for up to 12 months and had a mean of 9.2 in-person visits and 6.1 telephone contacts (8). If CoCM was being implemented with high fidelity among those with North Carolina Medicaid CoCM claims, then we would expect that many patients would have more than one or two patient visits. Although it is possible that these patients received integrated care services that involved a primary care provider, care manager, and psychiatric consultant, the level of fidelity to CoCM is unclear given the limited number of sessions.

This North Carolina case study on the approval of reimbursement for CoCM Medicaid billing codes has implications locally and for other states looking to increase uptake of CoCM. As such, understanding the barriers to adoption of and fidelity to CoCM is critical to mitigate access challenges.

Many barriers to widespread implementation of CoCM have been identified, such as discrepancies between Medicaid and Medicare requirements, a limited number of private insurers that reimburse for CoCM services, variability in reimbursement rates for CoCM, logistical challenges with implementing CoCM, and developing a workforce that can implement this evidence-based model (9). During the study period, North Carolina faced many of these obstacles.

Although many professionals have demonstrated efficacy as care managers (e.g., licensed clinical social workers, registered nurses, and licensed professional counselors) (10), during the study period North Carolina Medicaid required that behavioral health care managers who billed for CoCM be trained at the master’s or doctoral level. Consequently, programs in North Carolina that may have preferred to use other professionals as care managers to deliver CoCM, or that were unable to recruit master’s- or doctoral-level candidates into the care manager role, were not able to launch or bill for CoCM programs.

Also, even for systems in North Carolina that have appropriate staffing to launch CoCM programs, uptake and sustainability may be difficult when all private payers do not support CoCM (9). For example, Blue Cross–Blue Shield of North Carolina, the largest private insurer in the state, did not reimburse for CoCM billing codes during the study period.

In addition, the start-up and maintenance costs of CoCM can be substantial, making development of a CoCM program seem impractical for many practices and health systems (11). At the beginning of the study period, North Carolina Medicaid nonfacility reimbursement rates for CoCM billing codes 99492, 99493, and 99494 were $130.64, $104.54, and $54.08, respectively (7). During the study period, Medicare nonfacility rates for the same codes were $162.18, $129.38, and $67.03, respectively (12). It has been proposed that the adoption and sustainability of CoCM will likely only be achieved when all payers reimburse for CoCM at or above Medicare rates (9, 13).

Finally, ensuring fidelity to CoCM will be critical to achieve positive health outcomes and cost-effectiveness. During the study period, North Carolina did not have a statewide mechanism to promote CoCM and ensure evidence-based training and fidelity to its principles. The New York State Collaborative Care Initiative has attempted to address challenges related to CoCM implementation, utilization, and fidelity by partnering with the University of Washington Advancing Integrated Mental Health Solutions Center and assembling a technical assistance team that provides guidance across sites and program attestations, thereby ensuring fidelity (9, 14). As such, national and regional technical assistance programs that can train and support clinicians to deliver evidence-based CoCM while demonstrating fidelity may be important to ensure effective care delivery. This need has recently been acknowledged by Congress in the proposed Collaborate in an Orderly and Cohesive Manner Act (H.R. 5218), which would provide start-up costs for practices to adopt CoCM and develop technical assistance centers (15).

CoCM has the potential to generate robust access to behavioral health services in primary and specialty care practices. However, this case study from North Carolina demonstrates that simply reimbursing for Medicaid billing codes is not enough to ensure the ready adoption of this evidence-based model. Future translational work to ensure that CoCM is adopted and maintained should ensure that billing requirements for CoCM are consistent with evidence-based standards, metrics for billing are consistent across all payers, reimbursement is provided by all payers and is at least consistent with Medicare rates, and providers have access to CoCM expertise and oversight to ensure workforce development and high fidelity to the CoCM model.