Abstract
Background: Care management in primary care can be effective in helping patients with chronic disease improve their health; however, primary care practices are often challenged to identify revenue to pay for it. This study explored the impact of direct reimbursement on the provision of care management in a primary care physician organization.
Methods: Using data on expenses and health plan reimbursement during the initial 16 months of care management implementation at 5 practices, we calculated the percentage of related costs that were covered by payments. Qualitative data from interviews with practice members were used to identify their perceived barriers to care management reimbursement and the impact of current reimbursement strategies on service delivery.
Results: Direct reimbursement for care management covered only 21% of the costs. Reimbursement varied by care manager background, patient diagnoses, insurer, and indication for the visit. Barriers to gaining reimbursement included patient resistance to copay, clinician hesitation to bill for care management visits (for fear the patient may receive a bill), differential reimbursement policies of insurers, and general lack of reimbursement for care management in many cases. Although practice-level quality improvement incentives were an alternative means of supporting care management, because these incentives were not directly tied to the service of care management, they were used for other activities ultimately supporting patient care.
Conclusions: This study highlights the need for sufficient reimbursement to initiate and maintain care management for patients in primary care as proposed for service reforms under the Affordable Care Act.
Many patients in primary care have chronic conditions such as diabetes,1⇓–3 and many others have health habits and risk factors such as tobacco use or dietary habits that contribute to the development of chronic disease.4,5 Patients offered self-management support in the form of guidance and coaching from a care manager (often called a case manager, disease manager, or health coach6,7) show improvement in their disease management and reductions in their risk for complications.8⇓–10 Despite evidence demonstrating the benefits of care managers embedded in primary care, most practices do not have care managers available as part of their patient care teams.11⇓–13 A key barrier to broader use of care managers is the lack of reimbursement to the practice. If there are monetary savings at the practice level resulting from patients' participation in care management, these typically do not accrue back to the practice in a fee-for-service payment structure.14⇓–16
In Michigan, some insurers have taken steps to reimburse primary care practices for care management.17 In this study, we examined whether these reimbursement mechanisms were sufficient to support the costs of care management services and explored issues related to perspectives of providers and staffs related to billing and payment for care management. Specifically, the current study describes the ability of a physician organization (PO) to obtain direct reimbursement sufficient to cover the cost of care management and the perspective of the primary care practices regarding the financial barriers to care management delivery. These perspectives are important because primary care operational and clinical leaders are the key decision-makers determining whether care management is offered. Even if care management is cost-effective from the perspective of the payer or society in general, if it does not maintain budget neutrality or demonstrate appreciable revenue to the practice or practice organization, it will be challenging to implement.
Methods
The current study was part of a larger evaluation of processes, outcomes, and sustainability of chronic illness care associated with the implementation of the chronic care model (CCM)18 for adults with diabetes and diabetes risk factors in primary care. The CCM is a model for organizing primary care such that the model elements (self-management support, decision support, delivery system design, and clinical information systems) work within community contexts to produce prepared proactive practice teams and informed, activated patients who can effectively work together to manage a patient's chronic condition. A PO in Southeast Michigan partnered with the researchers, and 10 adult primary care practices were randomly assigned to the CCM intervention or usual care. For the 5 practices randomized to the intervention, the PO implemented CCM with a focus on practice-based care management and electronic clinical information systems supporting care management as the key elements to implementing CCM. Three larger practices were assigned either a full-time care manager or a combination of care managers equaling full-time, and 2 smaller practices were assigned a part-time care manager. Care managers had training in nursing (LPN and RN), social work (MSW) and dietetics (RD). The payer mix for the 5 practices was 24% Medicare, 5% Medicaid or local primary care low-income insurance, 38% commercial insurance (non-managed care), 32% managed care commercial insurance, and 1% no insurance/patient self-pay. Six practices were family medicine and four general internal medicine. Institutional review board approval was sought and received at Michigan State University and the University of Michigan.
Identifying and Tracking Expenses
The CCM intervention costs were placed into a cost center so that all expenses and revenues could be tracked, accounted for, and analyzed. Although grant funds supported some initial implementation costs, the actual costs and revenues to the PO were tracked as if there were no grant funding. Only incremental costs that were directly attributable to the care manager intervention were included. Salary costs for individuals with significant time contributions to care management (including care manager time, care manager supervisors, and IT staff) were included proportional to their time involved in care management. Physician and medical assistant patient volume did not change in the period after the intervention; therefore, their costs were not included.
Identifying and Tracking Revenue
The process of identifying opportunities for reimbursing care management was initiated by the PO clinical director and insurance specialists. On the basis of their review of policies of the insurers that held contracts with the PO, 2 types of opportunities for reimbursement or support of care managers were identified: 1) reimbursement for activities in which the care manager was directly involved (direct reimbursement), and 2) organizational reimbursement for achievement of performance benchmarks for quality care that could be influenced by the provision of care management. Direct reimbursement for care management was tracked in detail. Table 1 describes the direct reimbursement options.
Direct reimbursement opportunities often were specific to the characteristics of the service provider and were reviewed according to the discipline of each care manager. For example, billing was available for nutrition counseling through Medicare for dietitians, but that same billing was not available for other educational backgrounds. Diagrams describing reimbursement opportunities were distributed to care managers who were responsible for charge entry and receptionists who had a verification role. Diagrams and reimbursement options in each practice were discussed with physicians and clinical teams.
All care manager patient contacts were tracked via the electronic medical record and practice management system. Care managers and other office staff were trained to standardize billing processes for every encounter (telephone or in person). Checkout options were put into the care manager's work flow so that encounter time and event type were accounted for, and both care managers and reception staff could select the appropriate billing code. After every care manager visit, an option for billing was selected and submitted. If an insurer did not reimburse for care management services generally or did not cover a specific service, a dummy code was used to capture a charge for the encounter and a bill was not initiated to the patient. Each direct billing opportunity was tracked separately by care manager, practice, and payer.
Qualitative Data Collection and Analysis
In addition to the revenue and expense data being collected and reviewed, qualitative data were collected and analyzed regarding the overall implementation of care management, from which financial aspects of the care management implementation emerged. These data were collected via individual interviews with key practice members representing roles involved with care management (per practice: 1–2 physicians, 1–2 medical assistants, the practice manager, a reception clerk, and the clinical supervisor) at baseline and 2 time points during the intervention, as well as monthly interviews with each of the care managers during the initial 6 months of the intervention. In addition, field notes were taken during care manager implementation meetings held at the PO, which included all the care managers and the director of clinical quality improvement. Interviews and observations were conducted primarily by the Principal Investigator (JSH), 1 of 2 qualitatively trained research assistants, and occasionally another qualitative coinvestigator. These data were transcribed and coded with a series of a priori and emergent codes that identified key factors related to the care management implementation using the qualitative analytic software Atlas.ti (www.atlasti.com). Three coders coded the data and produced quotation outputs for the key themes; and after the 3 separate analysts completed their individual coding and theme development, they compared their findings and interpretations to derive a consensus. Financial reimbursement issues emerged as a key factor and were further analyzed during regular team meetings of the qualitative analysis team.19,20 Member checking was used with key practice and organizational members to determine the appropriateness of the themes identified.
Results
Costs and Sources of Reimbursement for Care Management
The study period for this article was the 16 months in which the intervention was active from March 2011 through June 2012. Expenses included care manager salary and fringe, supervisor (10%) and IT support (5%), salary and fringes, and supplies and equipment. The total was $427,026 or $85,405 per practice or $604.85 per enrolled patient. The expenses included the intervention preparation time during which the practices and care managers were attending training, determining work flows, and developing resources. Thus, $44,475 of the total cost, which accounts for the first 2 months of salary and fringes, was attributed to this start-up period. Revenue results are reported in several ways: Table 2 reports the reimbursement based on allowable billing codes of each of the options for supporting care management; Table 3 reports the reimbursement by care manager type.
The results indicate that care management was insufficiently supported by direct reimbursement, with actual revenue received covering only 21% of the cost of the care management intervention. Just more than 68% of billed services were actually collected with the highest reimbursement associated with t-codes at 93% collected and the worst being the care manager evaluation and management (E & M) at 58%. Therefore, the larger the practice, the greater were the gaps in reimbursement. Services provided by the registered dietitian were more likely to be reimbursed than services provided by care managers with other professional backgrounds. Services provided to patients with diagnoses of diabetes were the most likely to be reimbursed, while services for patients with diabetes risk factors such as obesity but no diabetes diagnosis were frequently not reimbursable.
Themes Identified from Qualitative Interviews
Patient Copayment, Coinsurance, and Deductible Issue
Interview participants frequently noted that they often did not know at the time of the care manager visit whether the patient would need to pay a copayment or how much that copayment would be. More generally, the complexity of insurance reimbursement rules and the variations among payers on coding created a confusing system of verification of covered services for office administrators and care managers, and this confusion in some cases limited access to the service. Copayment amounts varied from zero to US $30 per visit, with the average being US $20. The variable amounts were related to the type of patient insurer, the level of coverage, and overall purchased plan design. A few providers refused to have their patients billed directly for care management services because they believed that such billing would negatively impact patient satisfaction with the practice. Patient copayments clearly were a barrier in 2 practices serving a higher proportion of low-income patients, and care managers reported that some patients either refused services or discontinued care management due to the associated out-of-pocket costs. Care managers speculated that patients may perceive care manager visits differently than visits with the physician and therefore may not be willing to pay for services seen as less important than a physician visit. Table 4 lists salient quotations from interviews with practice members illustrating these points.
Lack of Reimbursement for Primary Prevention
Reimbursement opportunities were available in some cases for patients with established chronic disease (such as diabetes or kidney disease) but were lacking for patients without established disease (such as obesity). At the time of the study, there were only limited codes for treatment of obesity as a primary diagnosis; generally, the only codes that actually paid were those provided by the RD and only for patients who actually had diabetes or chronic kidney disease. Since that time, there are payable codes such as Medicare Intensive Behavioral Therapy for Obesity.21
Care Manager Staffing
As noted above, care managers with different educational backgrounds were eligible for different reimbursement opportunities. Licensed social workers are viewed by insurers as independent providers who could bill separately and at a higher level than registered nurses. They could bill separate E & M codes as providers when meeting the “incident to” requirements. For dietitians, there are dedicated medical nutrition therapy codes for the services they perform. Interviewees noted that care must be taken in properly scheduling patient visits to meet billing requirements. For example, if the patient had Medicare, they could see the care manager and their primary provider the same day, but if the patient had another type of insurance, they had to receive services on 2 separate days to qualify for reimbursement of both encounters. Respondents noted that nurses and social workers were only able to bill under the t-code or nursing code mechanisms at a substantially lower reimbursement rate than what would be received if the service were delivered by a physician. An advantage of t-code or an E & M code was that these codes could be used the same day as a physician visit, allowing the care managers to see patients right before or after a physician visit, providing convenience for the patient and facilitating team care.
Implementation Success and Reimbursement Capture
Implementation of care management can be affected by many issues including organizational changes such as mergers, leadership, and staffing changes; electronic medical record upgrades; and organizational factors such as provider support for care management. Care managers who had a relationship with the practice before changing job functions generally were able to launch a caseload in a shorter time than those new to the practice. The delay in filling a schedule also had an impact on the ability to realize full revenue if slots were left open. The patient response to care management also greatly impacts how well the care manager is used. This patient response then has an impact on the funding of the service because of the number of patients that use it. In this study, interview data highlighted these issues and the resultant differential uptake of the care manager intervention by practice, which affected the care manager panel size and ultimately impacted reimbursement.
Organizational Reimbursement Opportunities
Although this study focused on direct reimbursement opportunities, mechanisms for primary care practices to capture revenue from quality improvements that are the result of the presence of a care manager may be available. The PO may decide to attribute some or all these funds to covering the costs of care managers. Generally, these fall into 3 categories: uplifts to all payments based on meeting designation as a patient-centered medical home, base payments for achieving quality metrics, and additional payments for achieving benchmarks related to service delivery and clinical values. In this study, the PO received uplift in E & M coding of 20% for achieving a patient-centered medical home designation through 1 insurer. They also received incentive payments in incremental levels for having a provider's panel of patients achieve service benchmarks such as the percentage of patients with diabetes receiving an A1c test or reaching a certain percentage of the patients with diabetes with an A1c under 8.0—all benchmarks that are typically more successful among patients receiving care management.22⇓⇓⇓–26 However, the PO participating in the current study used these funds in other ways (ie, supporting IT development) rather than to cover the cost of care managers. Because some of these incentives are likely to improve with the use of care managers, but do not require a care manager, many POs choose not to hire them. This is not to diminish the importance of these quality-improvement payments or the ultimate benefit they provide to improving patient care. By unloading this financial hit from practices, the POs are able to invest in other resources. Examples of organizational reimbursement opportunities are outlined in Table 5.
Discussion
The results of this study reveal that direct billing fell far short of care manager costs during the initial 16 months of implementation of this intervention. Payment for care management to primary care commensurate with its benefits to practice efficiency and patients are a must if care management is to continue.27 Primary care physicians and their organizations already earn less relative to their specialty partners, thus further burden is untenable.28 An extensive literature search identified little information regarding mechanisms for funding care managers, although the need to do so is recognized.29⇓–31 Initiatives such as the Michigan MiPCT Demonstration Project (www.mipct.org)32 and the Comprehensive Primary Care Initiative (http://innovation.cms.gov/initiatives/comprehensive-primary-care-initiative/) are exploring avenues for payment, but established mechanisms as yet generally do not exist within the fee-for-service payment system.
Additional research is needed in other settings with other insurance situations as well as with other disease conditions and using different care management models. Vertically integrated health care systems have the potential to reap the benefits of savings “down the line,” benefitting from patient's improved disease management; however, in a fee-for-service environment where primary care practices receive reimbursement for the services they provide and not a financial benefit for savings to other parts of the health care system, there must be other mechanisms to provide coverage for the costs incurred for care management at the primary care level. The findings from this study also suggest that direct reimbursement for care management visits rather than simply quality improvement incentives to the organization are needed. Since the completion of this study, health insurers are testing out per member per month payment for care management, which may prove a more productive vehicle for payment to primary care for this aspect of population management.33 The new Centers for Medicare and Medicaid Services billing code for care management to be released in 2015 may be an avenue to reimbursement for primary care.34,35
Another area that may deserve exploration is related to productivity. For example, shifting some of the work of the physician to the care manager so that the physician has the time to see (and bill for) more patients.16 This option has existed for some time and has not taken hold, suggesting that it is not a reasonable solution; however, it may be particularly useful for primary prevention issues such as weight loss and smoking cessation. We were disappointed to find few options for direct reimbursement for patients with a primary diagnosis of obesity in the absence of other chronic conditions. Yet, patients are highly desirous of weight loss and exploration of the primary care practice as an avenue for weight loss is something for which some patients may be interested in paying out of pocket.
Limitations
The current study was completed in 1 PO in 1 state, and the payment policies for the insurers described here may not represent those of other geographic areas. Although this PO had a variety of insurers and insurance options, it was largely a fee-for-service system with limited managed care. Another limitation of the current study is that only direct billing options were tracked. Although organizational revenues were available for quality improvement, the PO for this study did not consider organizational incentive reimbursement as directly attributable to the care manager intervention, and we suspect that this PO is typical in that more general PO incentives for quality improvement are likely insufficient to incentivize practices to use sufficient numbers of care managers. Because the current study included only 5 practices and 5 care managers, results may be different in other environments, even if market constraints are similar. However, the detailed costing and qualitative interview data from the current study make an important contribution to how best to promote care management under health care reform. We believe this to be an example of an effective intervention, given patient results indicating clinical improvement.
Conclusions
Our results supported the widely held perception that care management reimbursement is inadequate to fully cover the costs of implementation at the practice level in a fee-for-service payment system. Careful thought and likely policy changes are needed so that this important aspect of patient care can be implemented and sustained in primary care practices.
Acknowledgments
Thank you to the care managers and staff at the participating practices for their participation in this study.
Notes
This article was externally peer reviewed.
Funding: This study was funded by the National Institutes of Health, National Institute for Diabetes, Digestive and Kidney Diseases, grant R18 DK082377-01A2.
Conflict of interest: The authors declare no conflicting or competing interests.
- Received for publication July 17, 2014.
- Revision received October 12, 2014.
- Accepted for publication October 20, 2014.
