Using a collaborative care model, we are integrating mental health into primary care at the country’s largest public health care system, New York City Health + Hospitals. Our program provides screening and treatment in the primary care setting and has demonstrated improved depression symptoms in more than half of enrolled patients.
It is possible to screen for and treat depression in the primary care setting.
Collaborative care for depression can succeed in diverse settings with a range of staffing combinations, patient demographics, and physical layouts.
Centralized data support is essential to drive operational workflows and quality improvement across multiple sites.
Depression is pervasive (13% of Americans and nearly 20% of Medicaid recipients) and more common in people with chronic disease, affecting 17% of people with cardiovascular disease and 27% of those with diabetes. It is also costly — 65% more per month if a patient has a chronic disease that co-occurs with depression when compared to chronic disease alone.
In January 2016, the United States Preventive Services Task Force recommended that universal depression screening occur in primary care settings that have adequate systems for diagnosis, treatment, and follow up. This represents an opportunity to redesign care to better meet patients where they are — that is, where they already access health care, at their level of engagement. Indeed, only 10% of patients follow up with a referral to a mental health specialist if they are not co-located in primary care. Primary care physicians and their colleagues are well-positioned to treat their patients for mental health issues: they generally have an established and trusting relationship with the patient, and are aware of what their patients’ usual behavior is, which enables them to identify mental health problems; they are also accustomed to coordinating complex care plans.
Multiple studies have demonstrated the efficacy of the collaborative care model for depression. While mental health and primary care historically have been siloed, recent policy changes and evolving payment models are spurring more integrated models of care. This wave of innovation is particularly important in safety-net health systems, which serve a high proportion of uninsured and Medicaid patients — and where poverty, language barriers, and other social determinants of health may contribute to the complex physical and behavioral health needs of patients.
Our challenge at NYC Health + Hospitals was to redesign care to screen and treat patients for depression across our large, diverse system while allowing patients to remain within their primary care medical home.
Our goal was to deploy a population-based, stepped-care approach in the primary care setting to treat depression or depression co-occurring with a chronic illness. We sought to screen all primary care patients for symptoms of depression and demonstrate meaningful improvement in symptoms for patients enrolled in collaborative care.
In 2014, under the New York State Hospital-Medical Home Demonstration Program, NYC Health + Hospitals began universal depression screening for adults in primary care. Initially, adults were screened at least once per year using the Patient Health Questionnaire (PHQ). Challenges with standardizing this practice and ensuring that patients received at minimum an annual screen led to screening for depression at every patient visit as part of a vital signs assessment, just like taking a patient’s blood pressure. Universal screening has allowed us to identify patients with depression who otherwise may not have sought out mental health care due to stigma, financial constraints, lack of access, or challenges with child care or work leave.
Simultaneously, NYC Health + Hospitals implemented collaborative care for depression at 11 hospitals and 6 community health centers. The program is based on the IMPACT model, developed at the University of Washington. The model calls for a team-based approach (with the patient at its center) that includes a depression clinician, consulting psychiatrist, and primary care provider (PCP). The approach begins with an evidence-based, low intensity treatment in primary care, while monitoring progress and systematically adjusting treatment (stepping up) to a more intense level of care if the patient does not improve.
A patient who screens positive on the PHQ-9 and meets eligibility requirements can be referred to collaborative care. (Patients with active substance use, serious mental illness such as schizophrenia, or who already are being seen in behavioral health are excluded.) After an initial encounter with a depression clinician, the patient is enrolled in the program and entered into a system-wide collaborative care registry.
The registry is a key component of this program, generating data to support patient outreach and treatment workflows as well as actionable patient lists delivered to the point of care (for example, lists of patients with whom there has been no recent contact, prioritized by time elapse since their last visit).
The majority of patient outreach and counseling is performed by the depression clinician, typically a Registered Nurse or Licensed Master Social Worker with behavioral health experience. He or she coordinates the patient’s overall treatment, including weekly case reviews with the consulting psychiatrist. The primary care provider retains point responsibility for a given patient — e.g., making the clinical diagnosis and prescribing antidepressants when indicated — but works closely with the depression clinician.
The patient is the most critical team member. The depression clinician is armed with a toolbox of interventions to actively engage patients in their own care, including motivational interviewing, behavioral activation, and problem-solving treatment.
Intensity of engagement depends on the acuity of the patient’s symptoms. Encounters can be face-to-face or telephonic, which allows for more flexibility and decreases the patient’s travel burden. Early in treatment, patients may have 2–3 encounters per month, with decreasing frequency to once per month as the PHQ-9 improves. Encounters are typically 15–30 minutes and involve a PHQ-9 screen, symptom monitoring, medication management, goal setting, and brief counseling.
Specific workflows address diabetes and hypertension. Patients with depression and diabetes, for example, may have diminished self-care, such as lower levels of physical activity, unhealthy eating patterns, and poor blood glucose monitoring. Counseling interventions with the patient target both the depression and self-management of diabetes or hypertension.
Key Practices for Care Teams
A cohesive team is essential to the success of collaborative care. Over the past year, we tailored team workflows to better identify and prioritize patients who need outreach, case review by the consulting psychiatrist, or a potential change in treatment plan. Based on feedback from our teams, we increased the frequency of delivering triaged, actionable patient lists to the point of care, as often as weekly. Through these efforts, we have discovered a few core practices that drive patient outcomes:
- Targeted skill improvement, such as identifying staff who may benefit from coaching on accurate use of the PHQ depression screeners.
- A “warm hand-off” from the PCP to the depression clinician, which ensures early patient engagement in treatment. The warm hand-off provides an opportunity to introduce the program to the patient in person, to set a behavioral activation goal, and to demonstrate the collaboration between PCP and collaborative care team members.
- Using the depression registry to manage the stepped-care approach, such as psychiatric consultation or a change in clinical intervention for enrolled patients who are not improving.
Although the spirit and structure of the program is the same throughout NYC Health + Hospitals, differences in staffing, team structure, space, and patient populations result in variation across sites. In some cases, this variation helps drive innovation, as with one site that wished to expand collaborative care to address perinatal depression. In other cases, the variation is unwarranted, such as reluctance of some primary care physicians to coordinate with depression clinicians or even to prescribe antidepressants.
To address variation across sites, we have taken steps to create a community around collaborative care. We have added clinical coaches who rove from one site to another to facilitate standardization as well as cross-fertilization. A newsletter shares best practices among the collaborative care teams and conference calls with consulting psychiatrists also help stimulate dialogue across practices.
Another challenge has been coordination of clinical workflows with billing practices. Because the program is payor-agnostic — that is, we treat all patients regardless of ability to pay — we must develop clinical workflows for all patients but separate financial workflows depending on how the patient’s care is funded.
In 2015, NYC Health + Hospitals screened approximately 225,000 adult primary care patients (screening rate > 90%) and identified nearly 15,000 patients (6.7%) with a positive screen, defined as a score of 10 or greater on the PHQ-9.
For patients enrolled in collaborative care we have maintained a laser focus on our “bottom-line” outcome metric, the clinical improvement rate, which defines meaningful amelioration of depression symptoms. In Q2 of 2015, our clinical improvement rate across NYC Health + Hospitals was 17.7%, which rose to 44.7% for Q4 2015 and to 57.6% for Q1 of 2016. We attribute this steady rise to the key practices and dedicated quality improvement initiatives described above.
In April 2015, funding for collaborative care for depression services was made available through New York’s Medicaid program. The statewide initiative offers payment for care management based on a set of services provided and performance on key metrics. NYC Health + Hospitals is transitioning from grant-based funding for Collaborative Care to sustainable funding via this Medicaid reimbursement program.
In our experience, collaborative care is an essential model for reaching and treating patients with mental illness. We are exploring opportunities to expand the collaborative care model to address pressing patient needs in adolescent and maternal depression, anxiety, and substance use disorders.
We are grateful to Dr. Ross Wilson, Chief Medical Officer at NYC Health + Hospitals, for his leadership in developing and growing the collaborative care program.
This case study originally appeared in NEJM Catalyst on May 24, 2016.