An Outcomes-Focused Approach to Mental Health Care

Over the past decade, insurance providers and self-insured health systems in the United States have increasingly moved toward value-based-care models that tie a portion of reimbursement to patient outcomes. This has fueled advances across fields from primary care to management of diabetes and cardiac conditions. Mental health care has been slow to make this shift, sticking with fee-for-service schemes that reward utilization: The more sessions a patient has, the more the provider makes, regardless of whether the patient improves. This system creates a disincentive for therapists to pursue the most rapid and effective treatment for their patients.

It doesn’t have to be this way. While delivering value-based mental health care has unique challenges, there’s no bar to doing it. In this article, we describe the methods and performance metrics that are making value-based mental health care possible and, drawing on our experience, offer guidelines to help providers, employers, or insurers design customized contracts. While our delivery model is primarily virtual, given sufficient availability of therapists, the approaches we use should be adaptable to an in-person or hybrid (in-person and virtual) environment.

Progress in Developing Programs

More than a decade ago, Harvard Business School’s Michael Porter argued for the integration of mental and physical health care into value-based treatment models, and today Medicaid providers and others have taken steps towards developing such value-based reimbursement programs. San Clemente, California-based Crossover Health, for example, bundles mental health care with comprehensive care, including primary care and physical therapy, in pay-for-performance contracts. Highmark Blue Cross Blue Shield of Western New York has entered into a contract with Value Network, a behavioral-health care collaborative of more than 55 area providers, under which reimbursement is tied to performance on behavioral-health metrics such as adherence to antipsychotic medication and follow-up. Outside of the United States, the cost of treatment for mental health has been successfully bundled into the overall cost of care for acute conditions such as heart disease or cancer.

Our teletherapy model has similarities to these, but, to the best of our knowledge, is the first in the United States that directly ties payment to specific clinical outcomes, with financial rewards for good results and penalties for poor ones. Our contracts with hospitals and health systems pay us a bonus when employees improve on an array of standardized mental health-related measures and require us to relinquish a portion of our fee when they don’t. While we focus on providing therapy to health care professionals, everyone in these organizations has access. Thousands of employees have used these services, with adoption rates typically around 7% to 9% but reaching 30% within certain residency programs and other high-acuity groups.

There has been an explosion of telehealth offerings for mental health in recent years, and studies confirm the efficacy of many of them. Telehealth in some settings has been shown to be as effective as in-person therapy for treating depression, anxiety, PTSD, and obsessive-compulsive disorder. What’s been missing from the growing array of offerings is a performance guarantee.

Drawing on our experience and that of others, we offer five considerations that mental-health-care providers, insurers, or employers should bear in mind as they embark on developing pay-for-performance contracts.

1. Specify outcomes.

In any value-based arrangement, you’ll need to specify the outcomes you intend to deliver. Without clarity there, it won’t be possible to create the required performance metrics and design incentives. For example, will a contract guarantee improvement in employee depression, anxiety, or both? Will it promote a reduction in mental-health-related turnover? Will it enhance workplace engagement or reduce burnout? Or perhaps there are other therapeutic outcomes a purchaser wants that a provider can confidently deliver.

To be sure, the complexity of some mental health challenges complicates outcomes measurement, and we wouldn’t argue that every mental health concern is amenable to a value-based treatment approach. Nonetheless, there are effective therapies for many of the most common mental health concerns, and these are good candidates for value-based treatment.

Providers can use any of several standardized tools for tracking patients’ mental health, including the PHQ-9 inventory for depression, GAD-7 for anxiety, the Mini-Z and Maslach Burnout Inventory for burnout. We use each of these to measure aggregate trend data in the populations we cover and share the data in anonymized form with client organizations. (It should go without saying that any telehealth provider must provide secure, HIPAA-compliant, strictly confidential care.)

Parties may also want to explore whether improvements in mental health could contribute to non-clinical outcomes of interest to employers such as increased productivity or reduced absenteeism. For example, we collect data on custom measures such as employees’ workplace engagement, likelihood of turnover, and sense of psychological safety, and we are now studying whether improvements in clinician burnout following therapy can help reduce medical errors. If we can show a strong connection, error reduction could become one of the outcomes we track for employers.

2. Tie payments to outcomes.

In its simplest form, a value-based contract creates monetary incentives for delivering specified outcomes at the population level and provides penalties for failing to. Both the provider and purchaser have an interest in creating a fair contract in which all parties benefit, not least of all patients.

The hospitals and health systems we work with pay us an implementation fee to connect them with the platform and an annual subscription fee for a suite of services, analytics, and reporting. Most of our client organizations are self-insured employers, and we work with them, as well as outside insurers as needed, to secure coverage through insurance benefits for the individual sessions.

The overall cost is capitated; we charge a capped amount to provide services to a given population for a specific period, without limiting the individual number of sessions, which is determined according to the clinical judgment of the provider. This leads to a predictable fee for health system partners and aligns incentives so that we assume some risk if it fails to provide effective care.

While the specifics of individual contracts vary, a typical performance agreement looks like this: We guarantee that at least 70% of patients who are being treated for depression or anxiety who had initial scores on the PHQ-9 and GAD-7 at a specified level and who completed their initial treatment plans will have improved on these measures by the end of that treatment phase. If 70% to 79% of patients improve, our performance guarantee is met and we receive no additional fee. However, if 80% or more of these guaranteed patients improve, we get an incentive payment. Conversely, if fewer than 70% improve, we provide a credit against the following year’s subscription fee.

A similar calculation can be done for measures of voluntary employee turnover or patients’ overall satisfaction with and utilization of the service. Depending on the specific contract, hundreds of thousands of dollars can be put at risk if the provider underperforms — a powerful incentive to provide effective therapy.

3. Tailor the therapist network.

For any mental health provider to guarantee performance, it’s critical to assemble a network of therapists skilled in addressing the particular needs of the purchaser’s population and provide the tools and support the therapists require to do their best work. While generalist therapists might be satisfactory in some cases, in others, such as with our clinical population, specific expertise is important to rapidly build rapport and a strong therapeutic alliance between the clinician and patient.

In our case, we select therapists for their experience working with health care professionals and provide additional continuing-education certification and training, which focuses on this population’s unique challenges and specific therapeutic needs. This isn’t just good clinical practice; it’s the centerpiece of any value-based business model. The more experienced your therapists are with a particular patient population, the more likely treatment will be effective and the faster patients will improve.

In our contracts with hospitals and health systems, staff get access to a platform through which they can receive teletherapy with one of hundreds of psychiatrists, psychologists, or social workers. Whom they match with depends on the patient’s risk profile and care needs as well as preferences, including provider demographics (such as race or gender), their goals for therapy, their preferred language, and their availability. A good match between therapist and patient helps speed effective treatment.

4. Carefully screen patients.

Regarding teletherapy, while it has proven itself capable of dealing with an array of mental health problems, it’s not the right modality for every case. No one benefits if teletherapy can’t meet patients’ needs and, obviously, psychiatric emergencies need acute care. Therefore, providers need to do careful screening at intake, with eligibility criteria agreed upon with the employer, to assure that patients get the care they need and that the provider can deliver the outcomes it promises.

In our case, patients are initially screened by an intake coordinator to confirm that teletherapy is right for them. Presenting problems such as acute suicidality or addiction may require in-person or inpatient treatment, and we refer those patients to appropriate specialists and facilities. However, the vast majority of patients who sign up for our services — more than 95% — are deemed at intake to be good candidates and are then matched with a therapist.

One potential problem that employers and insurers need to guard against is a platform rejecting patients with more challenging problems so it can more easily achieve its performance targets, even if it would be capable of helping them. The contract should include language prohibiting such behavior. Incentives for the provider to achieve a certain level of utilization, such as a bonus if 10% of employees use the platform in a given year, should also be considered; this can help discourage inappropriate rejections. Finally, the employer or insurer should monitor the provider’s rejection rate. If the percentage of patients rejected per year is above, say, 10%, a review of selection criteria is warranted.

5. Optimize access.

Many employees who could benefit from mental health services are reluctant to seek them out because of the perceived stigma. Patients’ hesitance to get help is only compounded if the care is difficult to access. Providers and payers alike need to assure that patients are comfortable using these services — ensuring strict confidentiality is essential — and that therapists are readily available when they are needed.

Even when they might want to see a teletherapist, many employees can’t easily take the time out of their workday. This is certainly the case for health care providers. We have therapists on call 24/7 (not surprisingly, given doctors’ schedules, 80% of our sessions occur after standard business hours and on weekends). Obviously, providing around-the-clock access adds costs, but in a value-based model, it can be a smart investment. The faster patients get well, the better for them and for the business.

What to Expect of Value-based Care

It’s early days for this model. But value-based care has proved itself across specialties and the early indications are that mental health care will be no exception. Obviously, we have the greatest visibility into our own performance, so we provide it here as one data point in support of these models. Across the thousands of employees who have used our services, we have seen a combined average 38% improvement in individual burnout scores, with more than 80% demonstrating an improvement in depression or anxiety from the start of therapy to completion. These outcomes have led to financial rewards for our company as well as for client organizations who, among other benefits, have seen reductions in expensive turnover (the cost of replacing a single doctor can exceed $500,000).

The University of California Hospital System based in San Francisco provides a good snapshot of the model at work. With a total network of a half-dozen campuses across the San Francisco Bay Area, the UCSF Medical Center has 23,000 employees. Of the employees who have used our services, 92% reported high overall satisfaction, and 78% completed at least five teletherapy visits. Overall, we measured an average of approximately 38% improvement in burnout scores from the first to final visit (as measured by the Mini-Z burnout survey), and 83% of patients showed a decrease in depression as measured by the PHQ-9. Seventy-nine percent demonstrated an improvement in anxiety, as measured by the GAD-7.

• • •

Our experience shows that a value-based mental health care model can work. We hope to see other providers follow our lead, particularly as the organizations who pay for employees’ mental health care demand value-based services. Patients, providers, employers, and society at large all stand to win.