INTRODUCTION

Collaborative care (CoCM) is an evidence-based, integrated model for the treatment of common mental health conditions in the medical setting1, 2. In CoCM, the primary medical provider works closely with a behavioral health care manager who provides assessments, brief psychosocial interventions, and medication management guidance, all with support from a designated psychiatric consultant2 (Table 1). Other key CoCM components include treatment to evidence-based targets and the systematic use of a patient registry. The efficacy of CoCM has been demonstrated by more than 80 randomized controlled trials (RCTs)3 and its core principles have been applied to the treatment of numerous mental health conditions4,5,6,7,8,9 in a variety of settings9, 10. However, its non-traditional, team-based workflow (including an emphasis on the time between face-to-face visits) has made reimbursement a major challenge2. In many cases, uncertainty about CoCM service costs and the lack of a clear pathway for reimbursement have made the model appear financially untenable11.

Table 1 Detailed CoCM Team Member Roles
Full size table

In response, the Centers for Medicare and Medicaid Services (CMS)12 introduced the opportunity for fee-for-service (FFS) reimbursement of CoCM for Medicare beneficiaries in 2017 with three temporary G-codes: G0502, G0503, and G050413. In 2018, these G-codes were transitioned to largely identical CPT codes 99492, 99493, and 994942, 14, 15 (Table 2). As with Medicare’s 2015-initiated, time-based Chronic Care Management (CCM) codes16, 17, CoCM codes are billed by a medical practitioner to reimburse for the cumulative time that the CoCM team spends managing patients over the course of a calendar month. Although the CoCM billing practitioner is typically the primary care provider, specialists (e.g., cardiologists, nephrologists) are also eligible. Practices billing the codes must inform patients of possible co-insurance, demonstrate their delivery of evidence-based CoCM, and meet all other requirements outlined by CMS (Table 3).

Table 2 Medicare CoCM Services Codes
Full size table
Table 3 Core Components of CoCM Billing
Full size table

Despite extensive efforts by health care leaders to raise awareness of CoCM codes11, 13, the authors know of few adopting organizations to date. In a 2017 qualitative study, O’Malley and colleagues found that only about 4.5% of eligible primary care providers (PCPs) billed Medicare for the CCM service during the first 15 months of reimbursement17. Given notable clinical and billing similarities with CCM, CoCM is likely to follow a comparable uptake pattern. At present, however, the lack of sufficient Medicare claims data for CoCM codes makes it challenging to effectively quantify their usage to date.

Using data obtained from qualitative, semi-structured interviews with health care organization representatives and payers, the authors present the first known study describing the experiences of early adopters and explorers of Medicare’s CoCM codes. It is hoped that these results will assist current and future practices with CoCM reimbursement and enable payers (including Medicare) to optimize reimbursement for behavioral health integration services.

METHODS

Sample Selection

Leveraging the professional contacts of the University of Washington’s AIMS Center18 and a nationwide network of CoCM stakeholders, the authors employed a purposeful sampling strategy to access a regionally diverse selection of U.S. health care organizations that were either considering or currently billing for CoCM services using Medicare’s billing codes. On the recommendations of our contacts, the study sample was expanded to include payers that were reimbursing for the codes.

Data Collection

From October 2017 to May 2018, the primary investigator (ADC) conducted 1-h individual or group interviews with individuals at organizations nationwide that were knowledgeable about CoCM billing. Telephone or video conference interviews were facilitated by a semi-structured interview guide adapted from O’Malley and colleagues’ previously mentioned study on CCM Medicare billing.17 The CoCM study interview guide addressed five core categories: health care organization characteristics, CoCM services (e.g., implementation, eligible diagnosis, staffing), patient consent, CoCM operational components (e.g., registry, time accrual, electronic health record (EHR)), and CoCM billing processes. All interviews were facilitated by a psychiatrist skilled in interviewing techniques and were audio-recorded and professionally transcribed. No honorarium was provided for study participation. This protocol was reviewed and granted exemption by the University of Washington Human Subjects Division (00003585).

Analysis

A content analysis approach19 was used to systematically summarize respondent responses. Concurrent with data collection, the analysis was conducted collaboratively by the four-author interdisciplinary team comprised of two psychiatrists with expertise in CoCM and billing, a physician-scientist experienced in health workforce research, and a qualitative medical sociologist. All team members reviewed the initial transcripts, meeting regularly to discuss data and identify relevant text within the documents. The selected text was labeled with codes using the interview guide’s core categories as the overarching coding scheme, with codes continually added or modified to capture emergent ideas.

Once major categories and codes were determined, matrix display tables20 were created in Microsoft Excel. This tabular format organizes data by code and case for comparative analysis within and across organizations, enabling observation of similarities, differences, and other key patterns. Two members (ADC, RLM) coded the remaining transcripts and entered data into the matrix tables20. Team members then summarized the range of CoCM billing experiences reported by respondents based on matrix data. All tables and summaries were reviewed in team consultation. Discrepancies were resolved through consensus.

RESULTS

Sample Description

This investigation included 15 in-depth interviews and a total of 25 respondents (Table 4). Two additional health care organizations declined from participating and a third did not respond to our interview request.

Table 4 Characteristics of Respondents from Health Care Organizations and Payers (n = 25)
Full size table

Respondents represented 12 health care organizations and 2 payers located in urban, geographically diverse areas. Of the 12 health care organizations, which were all large, multi-clinic systems, 8 were academically affiliated and the others were community-based. The number of clinics offering CoCM services within each health care organization varied considerably, as did the size of active patient panels and the number of staff members involved in providing the service. At the time of interview, 5 of the investigation’s 12 organizations were actively billing Medicare for CoCM. The remaining 7 were in various phases of billing initiation, ranging from anticipating launch within the year to deferring efforts for at least the near future. Four had prior experience with Medicare’s CCM codes. Detailed characteristics of the organizations can be found in Table 5.

Table 5 Characteristics of Health Care Organizations (n = 12)
Full size table

Health Care Organization Perspectives

The following sections highlight findings from interviews with the 12 participating health care organizations about contextual factors influencing billing implementation, core CoCM billing components, and financial sustainability. Data from organizations currently billing Medicare (referred to as “billers”) are contrasted with those not yet billing (referred to as “non-billers”).

Internal and External Contextual Factors

Buy-in from Key Stakeholders for CoCM Billing

Medicare CoCM codes require significant adjustments to established billing workflows and processes. Consequently, buy-in from stakeholders within (and occasionally outside) the organizations was of paramount importance. Respondents reported engaging key individuals early in the planning process and working collaboratively across units and departments. As one respondent noted:

It was a big, big lift from a lot of different people, from the [EHR name] folks, to the registry, the analytic folks; it was a big project. [BH Medical Director, Physician (Medicine/Psychiatry), Biller]

Among organization leadership, the need to cover costs of the non-reimbursable services associated with CoCM served as the impetus for CoCM billing. Goals of increasing mental and behavioral health capacity were additional billing promoters, as were mission and values (e.g., addressing underserved population needs).

Differences in primary motivators were observed between billers and non-billers. Billing organizations were more likely to have implemented CoCM after (or in response to) the release of Medicare codes. In contrast, most non-billing organizations implemented CoCM with the assistance of alternative funding prior to Medicare code release. Loss of this alternative funding could incentivize organizations to more actively pursue Medicare billing, as one respondent explained:

We have a … grant, that’s ending ... and actually, for the purposes of collaborative care staffing, it has ended … which is why ... I think people are much more interested now, in trying to figure out where we can tap into all the other payment sources… [BH Medical Director, Physician (Psychiatry), Non-biller]

Buy-in from compliance officers was also critical and, as one organization experienced, could bring CoCM billing implementation efforts to a halt if certain ethical and legal issues were left unacknowledged. Some such concerns included Medicare-specific cost sharing, variable CPT code adoption by other payers, procedural issues regarding proper coding, and sufficiency of CoCM clinical documentation. Buy-in discussions with billing and information technology (IT) staff were important as well, and concentrated on the technological infrastructure for time accrual, code entry, and medical record interoperability or integration.

Multiple organizations noted the significance of obtaining buy-in from PCPs and mental health clinicians. Respondents reported that PCPs were most concerned about being asked to obtain consent and discuss cost sharing with patients (despite sharing the service responsibilities with behavioral health staff). For behavioral health care managers, the focus of buy-in discussions was on the process of time accrual, clinical documentation (in both the EHR and registry), and billing code entry.

When discussing the commercial insurance space, respondents reported confusion about the impact of behavioral health benefit carve-outs on eligibility for CoCM reimbursement, given that the codes are entered in the PCP’s name, but the service is categorized as behavioral health. One organization felt that CoCM codes billed by primary care could circumvent network restrictions imposed by behavioral carve-outs and increase access to mental health services. Another had a different interpretation, reporting that siloed billing and credentialing resulting from carve-outs could deleteriously complicate reimbursement of integrated care models like CoCM.

Finally, several respondents reported efforts to secure buy-in from stakeholders beyond their health care organizations, most notably with non-Medicare payers, by promoting CoCM’s value and long-term cost savings in medical and behavioral health service lines. One health care organization even advocated at the state level using similar arguments.

Selection of Care Team in the Context of CoCM Billing

Under the Medicare rule, organizations billing CoCM codes must have a behavioral health care manager and psychiatric consultant working in collaboration with a treating practitioner, although specific profession types or licensure requirements for these roles have not been stipulated21. Several organizations, however, reported purposely reconfiguring their staff in preparation for CoCM billing to allow for the flexible use of other traditional FFS codes that have more specific provider-level restrictions. Most commonly, LICSWs were preferentially hired for their ability to bill psychotherapy codes that are unavailable to RNs and unlicensed social workers. As one organization explained:

…In the other clinics they’re all staffed by LICSWs because of the flexibility in being able to bill both collaborative care coding and CPT coding [FFS] psychotherapy. [Organization-Level Leadership, Physician (Psychiatry), Biller]

Conversely, one organization that was unable to hire an eligible billing clinician for traditional FFS codes reported that CoCM codes were especially helpful in moving toward financial sustainability.

Payer Mix and Availability of CoCM Reimbursement from Different Payers

Organization-level payer mix had a significant impact on the CoCM billing strategies of participating sites. While the opportunity for Medicare CoCM reimbursement was consistent nationwide, the Medicaid and commercial landscapes varied regionally. Depending on the specific patient’s payer and the credentials of the care manager, organizations had different approaches. Billing organizations either (1) used CoCM codes for both Medicare and commercial payers or (2) billed a mixture of CoCM codes for Medicare beneficiaries and traditional FFS codes for patients of other non-reimbursing payers. Though not billing Medicare CoCM codes, several non-billing clinics used traditional FFS codes for any billable services delivered as part of CoCM. In some cases, billing differentially by payer generated ethical and legal conundrums for organizations:

…it does feel a little unsettling to have some people have to pay a co-insurance, and then everybody else gets it for free because their insurance doesn’t recognize it as a service. [Organization-Level Leadership, Clinical Psychologist, Non-biller]

In response to this concern, one organization stated that individual clinics newer to the CoCM service were being prioritized for billing to minimize the number of long-standing individual patients that had to transition to cost sharing.

Despite having a variety of possible billing strategies, all organizations reported providing CoCM services to beneficiaries for whom they would not be reimbursed. Some elected to accept the financial loss on these patients (as a result of their strong belief in offering the service), while others referred these patients to other service lines (e.g., specialty mental health care). One health care organization described their acceptance of the financial loss:

…we’ve used general assessment codes and sort of eaten the costs of the other piece of collaborative care…we treated everybody the same way, except in terms of billing. Because people with Medicaid or no insurance or other insurances, there wasn’t a billing mechanism for that. So, you just bill what we currently could, and there is no reimbursement for the other pieces. [BH Medical Director, Physician (Psychiatry), Biller]

Primary reasons non-billers avoided the codes included access to funding from alternative sources (e.g., grants) and poor CoCM service code coverage, typically due to insufficient involvement of commercial payers in the region.

CoCM Billing Components

Attainment and Documentation of Patient Consent for CoCM

Medicare regulation stipulates that patients must be informed of possible cost sharing and provide consent for the team to speak with relevant specialists (including psychiatric consultants) prior to the commencement of CoCM services. This process may be verbal and must be clearly documented in the medical record. Some organizations reported that consent was performed by the PCP, while others mentioned that it was conducted by a different team member (although CMS has stipulated that it must be performed by the treating practitioner12). Consent was occasionally experienced as a point of contention for PCPs, mostly due to discomfort discussing financial terms or concern about how long it would take:

...we had to involve our providers because we’re required to have the providers actually gain their consent. And we dealt with a decent amount of pushback really, about providers having to consent patients and tell them that they might have to pay for the service. [BH Medical Director, Physician (Psychiatry), Non-biller]

Organizations also reported that, on occasion, patients had expressed confusion when contacted by the care manager to schedule an intake visit because they had not fully understood CoCM from the consent process. In response, several organizations (including billers and non-billers) used guidelines or a formal script to ensure that required information was conveyed to patients. Further, most billers and one non-biller reported that an EHR template or shortcut had been created (or was planned) to facilitate documentation of consent.

Tracking Time Accrual, EHR Integration with CoCM Registry, and Billing/Revenue Flow

The organizational transition to billing CoCM codes requires substantial workflow adjustments for billing staff, administrators, care managers, and clinicians. One such adjustment is that all team members must keep track of the time spent caring for a patient over the course of a calendar month, instead of billing at the time of service. This major shift necessitates the development of new strategies and infrastructure for patient-level data accrual and maintenance. Although three participating organizations had optimized their EHR and registry for this purpose, two successful CoCM code billers did not have EHR-registry integration at the time of interview. This demonstrates that a lack of IT integration or interoperability is not an insurmountable barrier to billing CoCM codes.

Of those health care organizations that had achieved EHR-registry integration, some reported a perceived increase in efficiency by allowing documentation (including time accrual) to occur in one place. Others highlighted their capacity to efficiently document consent, identify eligible patients, or run reports of aggregated data. At the time of interview, however, no organization had succeeded in fully automating the process of billing accrued encounter-based time directly from data entered into an integrated EHR-registry. Even with integration or interoperability, the time accrual and code entry tasks were still often left to the care manager and billing department, which occasionally became a source of contention. One respondent noted:

…we’re trying to work on that…the care managers complain about the registry a lot. It takes a lot of their time. It decreases the amount of time that they spend with patients… [Organization-Level Leadership, Physician (Psychiatry), Biller]

The in-house IT teams were vital for organizations with successful EHR-registry integration. Those that had not yet integrated IT systems, but sought to do so, were challenged by competing EHR modification priorities, a reluctance to disrupt ongoing systems, a lack of IT financial support, or general technical difficulties.

CoCM billing challenges also arose directly from the atypical structure of the service, which is delivered in the primary care setting, billed in the name of the primary care provider, and rendered partially by behavioral health providers. Mechanisms are frequently unavailable for reimbursement dollars paid to the primary care department (incorporating the work of the psychiatric consultant) to be directly used to compensate the psychiatric consultant or care managers. In this way, CoCM codes expose how siloed departmental revenue acquisition can pose additional difficulties for the financing of integrated care. According to one respondent:

I guess from the billing perspective…you can bill the G [CoCM] codes but then where does my salary come from…? How does that money trickle down to pay the people who need to be paid? [Organization-Level Leadership, Physician (Psychiatry), Non-biller]

One organization initially utilized ledger transfers to move reimbursement revenue across departments but ultimately found this to be burdensome. Instead, it developed a virtual integrated care clinic to solve the problem:

…this idea of building a virtual clinic in the computer system. It’s going to be called behavioral health integration, and it will be primary care clinic where primary care doctors can bill, but the revenue will come to the psychiatry department. This new virtual office, we have to build it in [EHR Name] … requires a huge [EHR Name] build. [Organization-Level Leadership, Physician (Psychiatry), Biller]

Despite the virtual clinic’s effectiveness in resolving ledger transfer obstacles, it simultaneously created credentialing and insurance paneling challenges for providers in both departments.

Finally, both billing and non-billing organizations noted apprehensions about loss of revenue from months in which accrued time did not meet stipulated thresholds and, more generally, what constituted billable time. Despite the clinical and billing similarities between CoCM and CCM, an organizational history of the latter did not appear to mitigate these or other billing-related concerns.

Impact of CoCM Codes on Financial Sustainability

Financial sustainability was a major factor for several billing organizations that aimed to expand mental health services. Both billing and non-billing organizations mentioned collecting CoCM cost data to assess the feasibility of maintenance or expansion. For some non-billing organizations, breaking even on salaries and generally covering program costs were considered indicators of success. A few health care organizations also noted that anticipating more payers to activate CoCM codes was a facilitator in moving toward billing implementation:

Again, we’re only looking at the Medicare codes…when we get to being able to do it for all payers, we might be able to measure…our…costs…versus what we get reimbursed, it might be worth it [BH Medical Director, Physician (Psychiatry), Biller]

Perspectives from Payers

The authors conducted additional interviews with two payers that were already reimbursing for CoCM. Both payers commented extensively on their confidence in the evidence-based CoCM service, appreciating that it improves access, increases behavioral health clinical support for PCPs, and is a high-value investment. Payers hoped that activating the codes would be an effective way to incentivize practices to implement the CoCM service. According to one respondent:

For the plan, it’s in our best interest [to implement CoCM], because the overall medical spend is cheaper…We can actually give people lesser increases of their premium…We have happier patients that are healthier…Whereas if you don’t keep your prices down, then they…go somewhere else. [BH Medical Director, Physician (Psychiatry), Payer]

Additionally, payers reported that their confidence in the CoCM service was highly dependent on clinics and providers closely adhering to the service’s gold standard guidelines. According to one respondent:

…essentially, in order to get the…reimbursement for collaborative care, we need to be confident that they are using the collaborative care model…and not just saying, ‘I’m doing collaborative care.’ [BH Medical Director, Public Health Practitioner (MPH), Payer]

As with the health care delivery organizations, payers experienced challenges with internal buy-in to activate CoCM code reimbursement within their own organizations. This was attributed to the composition of the insurance workforce, which often includes non-clinical staff members with little mental health experience. As one respondent explained:

…the political push to be able to…bring your stakeholders…along with your line of thinking. That takes a learning curve to break it down…for all of those different stakeholders, administrators, people that aren’t physicians, or medically oriented [BH Medical Director, Physician (Psychiatry), Payer]

Unique to their vantage point, payers noted challenges associated with informing practices that CoCM codes were activated and billable. They struggled to find effective means to deliver the message to clinicians and health care leaders beyond existing virtual forums, meetings, and email threads. This sentiment was generally consistent with that of health care organization representatives, many of whom endorsed difficulties in determining which local and regional payers were reimbursing for CoCM.

DISCUSSION

Although implementation of newly available CoCM billing codes poses numerous operational challenges, the results of this investigation demonstrate that five early adopters have used the Medicare CoCM codes successfully. The remaining seven participating health care organizations were in various phases of considering or initiating billing. Successful organizations secured institutional buy-in from key stakeholders, clearly outlined CoCM billing and service populations, defined workflows for time accrual, and instituted processes for code entry and consent. These attributes were often not easy to achieve, with even successfully billing organizations readily acknowledging them as challenges. Table 6 provides a summary of these key findings and the authors’ proposed strategies to help overcome common barriers.

Table 6 Implications of Findings and Proposed Strategies/Solutions for Health Care Organizations
Full size table

Data from this investigation had notable parallels to a recent qualitative study that assessed early experiences of providers, practices, and professional societies with Chronic Care Management (CCM) billing codes17. For both CCM and CoCM, results demonstrated the following: (1) a generally positive response to the clinical service; (2) challenges with consent and cost-sharing discussions (albeit not primary or insurmountable ones); (3) concerns about long-term financial sustainability; and (4) difficulties with billing workflow changes, IT interoperability, and time accrual. To date, the impact of CoCM codes on financial sustainability has yet to be assessed empirically, though a recent simulation study suggests that they may provide an opportunity for long-term viability22.

We acknowledge several limitations of this study. While reflecting regional diversity, our sample of large, urban, predominately academic organizations does not necessarily reflect the experiences of smaller, non-urban counterparts, nor those of independent primary care practices. Further, the novelty of the CoCM codes provided a relatively short exposure period for organizations and restricted our sample to early adopters (which may be substantially different from later initiators). A majority of the study participants were psychiatrists, which limited our capacity to understand the perspectives of other provider types, administrators, and patients. However, our study did include a number of dually trained (medicine–psychiatry) and internal medicine physicians, as well as diversely trained organizational leaders. Notable strengths of this study include the presentation of viewpoints from both health care organization representatives and payers across a variety of geographic settings, and the inclusion of data from billers and non-billers along the Medicare CoCM billing implementation continuum. Finally, this study provides an important glimpse into the experiences of early CoCM code adopters, reveals that “one size fits all” solutions are of limited utility, and describes a range of tailored billing strategies for health care organizations.

Additional research is needed to further assess the experiences of PCPs, behavioral health care managers, and patients with CoCM billing and to measure the service’s overall implementation and sustainability costs. Given the immense need for high-quality, evidence-based mental health services nationwide23, 24, future investigations should examine the broad impacts of these codes on real-world CoCM service delivery and behavioral health clinical outcomes.

CONCLUSIONS

Although CoCM is an evidence-based and effective model for behavioral health integration, its financing and reimbursement remain significant challenges. The results of this investigation suggest that with deliberate planning, persistence, and widespread organizational buy-in, successful utilization of newly available FFS CoCM billing codes is achievable by health care organizations.