Evaluating Implementation and Outcomes of the Achieving Depression and Anxiety Patient-Centered Treatment Collaborative Care Program in a Large, Integrated Health Care System: A Mixed Methods Observational Study Protocol

Abstract

Background

Collaborative care is an evidence-based multidisciplinary model shown to improve patient depression and anxiety outcomes. Although there is robust literature showing the effectiveness of collaborative care on depression and anxiety symptoms, there is little published on outcomes of collaborative care implementation or the efficacy of collaborative care compared with psychiatric referrals. Reported here is a study protocol examining a novel depression and anxiety collaborative care program in a large, integrated health care system.

Methods

This is a mixed methods study of the Achieving Depression and Anxiety Patient-Centered Treatment (ADAPT) program as compared to outpatient psychiatric care at Kaiser Permanente Northern California, a large, integrated health care delivery system. The ADAPT program was designed using collaborative care principles, including measurement-based care, accurate diagnosis, and population management. Eligible participants will be ≥ 18 years old with mild to moderate-severe depressive symptoms as measured by the Patient Health Questionnaire-9. Exclusion criteria include acute suicide risk and serious mental health comorbidities. Implementation is examined using the reach, effectiveness, adoption, implementation, and maintenance (RE-AIM) framework and interviews with program stakeholders.

Results Pending.

Conclusion

Study data will help inform future collaborative care efforts while expanding the literature base. The Achieving Depression and Anxiety Patient-Centered Treatment program may improve patient outcomes and access to quality depression and/or anxiety care.

Background

Depression and anxiety are two of the most common psychiatric disorders, impacting 16.6% and 28.8% of adults in the United States during their lifetimes. ¹ Depression is the leading cause of disability worldwide, ² and depressive and anxiety disorders are the top two causes of disability-related burden in the Americas. ³ Delays in depression and anxiety care can lead to increased disorder severity and poorer patient outcomes, including suicide, ^4,5 and depression is a common risk factor of this cause of death in the United States. ⁶ However, depression and anxiety care are often delayed due to long waits for specialty mental health care referrals. ^7–9 Delays in mental health care contribute to the 43% to 52% higher ambulatory costs ^10–13 and the over $210 billion in lost earnings yearly ¹⁴ associated with depressive and anxiety disorders. Thus, timely, evidence-based treatment of depression and anxiety is critical to addressing the disease burden of depression and anxiety.

Primary care is an important setting to implement evidence-based treatments for depression and anxiety as it is the most common first point of contact for patients, and more than half of all depression and anxiety care occurs in primary care settings. ^9,15–19 Up to 90% of patients report preferring to receive mental health care in primary care as part of routine medical care. ²⁰ However, the majority of patients receive specialty psychiatric care referrals. ²¹ Multiple facets limit patient connection with specialty psychiatric care, including a shortage of mental health practitioners, ²² long referral wait times, ^7–9 physical separation of primary care and specialty care, ²¹ and stigma. ⁸ Few patients schedule appointments after referral, and much fewer follow-up with specialty practitioners. ²⁰ Given the challenges in accessing specialty psychiatric care, demands for primary care management of depression and anxiety will likely grow, ^23,24 creating a need to implement effective solutions during routine medical care.

Collaborative psychiatric care is an evidence-based model for delivering early intervention for depression and anxiety in primary care ^25,26 and is associated with reduced symptoms, improved quality of life, and sustained remission. ^25,27 Collaborative care provides patient-centered and evidence-based interventions in a population management framework administered by a multidisciplinary team. ²⁸ Frequent and measurement-based care and accurate diagnosis guide the treatment plan. ²⁸ The model incorporates accountable care principles, tying reimbursement to quality and outcomes. ²⁸ Despite the robust evidence of collaborative care effectiveness for depression and anxiety, several barriers limit broad clinical implementation. These include problems with model reimbursement (including funding a care manager), practitioner discomfort with managing depression or anxiety, and patient engagement. ²⁹

Although there is much published regarding qualitative barriers and facilitators to collaborative care implementation, ²⁹ there are few reports of quantitative outcomes. For example, a statewide collaborative care model for depression was successfully implemented by providing funding for a dedicated care team and tying reimbursement to ongoing collaborative care implementation training. ³⁰ While collaborative care implementation was successfully adopted in this study, it was not associated with improved depression outcomes. ³¹ Another study of collaborative care implementation outlined the general program design and reported the percentage of patients with depression response or remission ³² but did not compare outcomes to specialty psychiatric care or trial novel model elements to address known collaborative care implementation barriers. ³³ Thus, further research is needed to understand if collaborative care can be successfully implemented to improve patient outcomes and how to translate the benefits of collaborative care into routine practice. Further, there is a paucity of data examining critical elements of collaborative care implementation within the existing clinical structures of diverse, integrated health care systems, which may help overcome organizational barriers to collaborative care implementation. ³³

In this paper the authors describe a novel collaborative care program and protocol to study the Achieving Depression and Anxiety Patient-Centered Treatment (ADAPT) program as compared to outpatient psychiatric care for depression. The ADAPT program was designed using stakeholder feedback and evidence regarding barriers and facilitators to collaborative care implementation. This study aims to examine outcomes of this collaborative care model compared to outpatient psychiatric care at Kaiser Permanente Northern California, implementation measures using the reach, effectiveness, adoption, implementation, and maintenance (RE-AIM) framework, and qualitative feedback regarding the program.

Methods

Setting

Kaiser Permanente Northern California is a large, integrated health care delivery system in the United States with 769 medical centers and hospitals serving more than 4.5 million patients (47% of the regional population) who are insured through commercial (eg, employer-based) insurance plans, Medicare, Medicaid, and health insurance exchange plans. Patients at Kaiser Permanente Northern California are highly representative of the ethnic, health, and socioeconomic diversity of the surrounding and statewide population. ³⁴ Approximately 12% of the Kaiser Permanente Northern California population is insured under Medicaid and is similar to the underlying population except with respect to income, where members underrepresent the very poor and the very wealthy. ^35,36

Study Design

The planned observational, mixed methods study evaluates quantitative outcomes related to implementation of the ADAPT program based on the RE-AIM framework. ³⁷ Qualitative interviews will be conducted with key program stakeholders to understand barriers and facilitators of program implementation given the novel design components to the model. The study protocol was registered with ClinicalTrials.org (NCT04839718) on April 9, 2021; the study procedures and protocol outlined here and as registered are the first version and are ongoing. This study follows SPIRIT guidelines for clinical trial reporting. The study was approved by the Kaiser Permanente Northern California Institutional Review Board as exempt from informed consent as it does not involve interactions with research participants. The research is conducted in accordance with the principles of the Declaration of Helsinki.

Study Participants

Eligible participants for ADAPT and the non-ADAPT comparator group will be ≥ 18 years old screening positive for depression symptoms (defined by a Patient Health Questionnaire [PHQ]-9 score of ≥ 5). ³⁸ ADAPT is primarily a depression treatment program, which is why patients are included based on the PHQ-9. Anxiety symptoms are measured by the Generalized Anxiety Disorder 7-item scale (GAD-7) to understand patient comorbidities and inform treatment planning. Patients participating in ADAPT are not limited in their ability to seek or receive concomitant care or interventions. Exclusion criteria will be acute suicide risk; diagnoses of bipolar, psychotic, dementia, or active substance use disorders; current hospice or home-based palliative care; residing in a skilled nursing or assisted living facility; not a Kaiser Permanente Northern California member; or already in established psychiatric care. Additional exclusion criteria for the non-ADAPT comparator group include a history of care in ADAPT. Patients receiving standard depression and anxiety care in the health care system but not participating in ADAPT will serve as the comparator group.

Measurements

Archival data regarding patient service utilization (contact type [phone, video, or in-person contact], provider and clinic type, and patient-level characteristics [visit-associated psychiatric diagnosis, patient history of previous psychiatric diagnosis, symptom severity documentation, age, race/ethnicity, and reported gender]) will be extracted from electronic health record data. Program efficacy will be examined using continuous PHQ-9 scores at 6 months as the primary outcome. Secondary outcomes include a 50% reduction in PHQ-9 symptoms (consistent with the Health Care Effectiveness Data and Information Set measure), depression remission (PHQ-9 score < 5), and GAD-7 scores compared to baseline. Service reach will be examined by summarizing the number of patients enrolling in the ADAPT program compared to the number of patients per site meeting eligibility criteria but not enrolled. Adoption will be measured as patients’ kept appointments, signifying engagement with the program. Implementation and maintenance will be examined by analyzing team member fidelity to the collaborative care model over time, including utilization of measurement-based care.

Depressive Symptoms

Depressive symptoms will be measured using PHQ-9. It consists of 9 questions reflecting the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) criteria as “0” (not at all) to “3” (nearly every day), such that scores range from 0 to 27. The PHQ-9 has been validated as a depressive symptom screening and follow-up instrument with high sensitivity ( < 88%) and specificity ( < 88%) for major depression, with PHQ-9 scores of 5, 10, 15, and 20 representing mild, moderate, moderately severe, and severe depressive symptoms. ³⁸

Anxiety Symptoms

Anxiety symptoms (secondary outcome) will be screened with the Generalized Anxiety Disorder 2-item (GAD-2) scale and, if positive, measured using the GAD-7 scale. The GAD-2 scale uses the first two questions of the GAD-7 and represents the core anxiety symptoms. ³⁹ It has been validated in multiple studies and shown to retain the psychometric properties of the GAD-7, with high sensitivity (76%) and specificity (81%). ³⁹ Due to its discriminant capability, it has been proposed as an essential first step for screening generalized anxiety disorder. ³⁹ Items are scored from 0 to 3; a GAD-2 score of 3 or more is considered positive and warranting follow-up with the GAD-7. ³⁹ The GAD-7 has been validated as a screening and measurement instrument for anxiety disorders with high sensitivity (61.3%) and specificity (72.7%). ⁴⁰ Each question is scored from 0 to 3, and total scores range from 0 to 21. ⁴¹ Cut-off scores for mild, moderate, and severe anxiety symptoms are 5, 10, and 15, respectively. ⁴¹

Qualitative Interviews

Key stakeholders in the ADAPT program, including pharmacists, therapists, primary care physicians, and patients, will be interviewed every 6 to 12 months to determine barriers and facilitators of program structure and implementation.

Analytic Plan

Clinical and demographic characteristics will be summarized using means and confidence intervals. Bivariate analyses of demographic and clinical characteristics, along with service reach and adoption, will be used to compare ADAPT participants to ADAPT nonparticipants per site on demographic and clinical characteristics, including age, gender, insurance type, socioeconomic status, and diagnoses at baseline and follow-up using Pearson’s χ² test for categorical and the independent samples t-test for continuous variables. Repeated-measure mixed effects regression modeling will be used to evaluate program efficacy, implementation, and maintenance compared to usual care. Important covariates including clinic site, provider utilization (including type, familiarity with, and frequency), and medication and treatment compliance (including number of therapy sessions and other types of appointments kept) will be incorporated into the models evaluating implementation.

Sample Size Calculation

For the primary outcome of program efficacy using continuous PHQ-9 scores at 6 months, the authors assume a significance level (type I error rate) of 0.05 for all tests. Approximately 640,000 members are estimated to have depressive or anxiety disorder diagnoses over the study time period. Given this population, the authors will have the chance of < 80% to detect a noninferiority between the two groups, with our estimated sample, based on extant literature. ^25,26 The study will remain open until adequate sample sizes have been obtained.

Intervention Concepts

In designing an implementable population management program, critical elements of collaborative care, evidence-based depression and anxiety care, and population care management were combined as described later. ^42,43 These included symptom detection with validated screening tools, engaging patients with a care team, using evidence-based treatments aligned with accountable care and population management, systematic follow-up, communication and coordination between the care team and patient, and case and program oversight. ^42,43 The authors integrated novel components into the program design to address previously reported implementation barriers in collaborative care from the literature, ²⁹ including care manager reimbursement and medication management support (Table 1).

Table 1:

Description of collaborative care principles for depression and anxiety and team roles in traditional collaborative care

Collaborative care principle	Components	Barriers	How addressed in ADAPT
Patient-centered care	Care that is respectful of, and responsive to, individual patient preferences, needs and values	Patient time, symptom burden	Care delivery via telehealth to improve patient access
Population management	Patient registry and symptom tracking	Reimbursement for staff/case manager to manage	Collaborative care visits booked into a unique department to create a patient registry
Measurement-based treatment	Measuring symptoms regularly during treatment	Reimbursement for staff/case manager to complete	Integrate symptom tracking into every visit
Evidence-based care	Therapy and medication management	Lack of knowledge or consensus regarding evidence-based care	Utilization of clinical pharmacists, prescribing workflows, titration, and augmentation protocols
Accountable care	Rational resource use in an integrated health system	Lack of leadership support, perceived fear of loss of autonomy by practitioner	Weekly meetings to review caseload and patient progress
Care manager	Unclear reimbursement	Reimbursement for staff/case manager	Practitioner workflows including stepped care processes

Barriers reported previously regarding these principles or dimensions of the model are listed along with the proposed plan to address in this study.

ADAPT, Achieving Depression and Anxiety Patient-Centered Treatment .

Collaborative Care

Collaborative care for depression and/or anxiety is an evidence-based intervention shown to improve depression and anxiety outcomes. ⁴⁴ Collaborative psychiatric care models emphasize patient-centered care, population-based care, measurement-based care, implementation of evidence-based treatments including medication and psychotherapy, and accountable care. ⁹ Evidence-based pharmacological and psychotherapeutic treatments and care managers have been identified as critical to depression remission. ²⁵ The ADAPT model included these principles as outlined next.

Patient-Centered Care

Patient-centered care is defined as “providing care that is respectful of, and responsive to, individual patient preferences, needs and values, and ensuring that patient values guide all clinical decisions.” ⁴⁵ ADAPT was designed to leverage the existing telehealth structure at Kaiser Permanente Northern California ⁴⁶ as a primary care modality for ADAPT to facilitate patient needs elicited in qualitative feedback (see Stakeholder section).

Population-Based Care

Principles of population-based collaborative care include 1) identification of appropriate patients, 2) measurement-based care, and 3) patient outreach depending on symptom progression. ²⁸

Measurement-Based Care

As outlined, systematic symptom tracking will be employed starting at ADAPT enrollment and continued at each subsequent program visit. The primary goal of care is depression symptom reduction. ⁴⁷

Evidence-Based Care

Time-Limited Psychotherapy

Time-limited psychotherapy describes solution-focused, short-term therapeutic interventions. Such interventions are evidence-based, ^48,49 result in faster symptom and functional improvement, ^48,50,51 and align with value-based^{48, 50} and population-based care. ⁵² Time-limited psychotherapy has efficacy in depression and/or anxiety symptom reduction. ^48,50,51,53 Practitioners are encouraged to utilize this modality as deemed appropriate in their clinical judgment.

Pharmacist-Assisted Medication Management

The psychiatrist shortage is well-documented and projected to grow over time. ⁵⁴ Other care providers as part of a comprehensive medication management plan could bridge this gap. Clinical pharmacists are a workforce with the training and expertise to fill this role, and a majority of states allow pharmacist medication prescribing. ⁵⁵ To develop this program, psychiatric medication prescribing and titration protocols were developed to guide the pharmacists regarding first-line treatments, augmentation strategies, and switch or changes in therapy by psychiatrists. ⁵⁶

Restructuring the Collaborative Care Manager

Previous work identifies paying for the care manager role a barrier to collaborative care implementation. ²⁵ This team member performs population management and patient symptom tracking and identifies patients needing care interventions. ²⁸ In the ADAPT model, the care manager role is integrated into each care experience by structuring symptom tracking into the visit and providing provider guidelines for stepped care depending on patient symptoms.

Accountable Care

Accountable care seeks to deliver the appropriate care at the right time and measure care quality for the population. This principle will be incorporated through weekly team meetings to review caseloads, patient progress, and patient care transitions. Depressive symptom tracking will be integrated with the system-wide National Committee for Quality Assurance Health Care Effectiveness Data and Information Set depression measures. ⁵⁷

Stakeholder Input Into the ADAPT Model

Stakeholders included primary care providers, psychiatry department medical directors and managers, patients, pharmacists, and therapists. Primary care physicians identified easy and clear referral workflows as priorities. Internal referral websites or secure staff-to-staff messages were noted as preferred modalities. Adult psychiatry department medical directors and managers shared that patient access to high-quality mental health services was a priority, with psychiatry serving a consulting role if patients did not respond to initial interventions. Patients identified easy and timely access to care as priorities, preferring care that “came to them” and that did not take much coordination to obtain. Pharmacists identified consistent ways to reach patients, regular symptom collection, and protocols to follow regarding medication escalation and reasons to refer to psychiatry as priorities. Therapists prioritized personalizing interventions to patients and an organic approach to diagnosis.

Implementation Strategy

The ADAPT program will be implemented as a telehealth-first program at 3 medical centers. Patients may be referred from primary care, psychiatrists, or psychiatry triage. At each site, the treatment team consists of a psychiatrist physician lead, therapists, and clinical pharmacists. Weekly meetings will be held with this team to review program enrollment, adherence, and workflow needs. One site will have psychiatry and pharmacist technicians as additional team members.

The ADAPT Intervention

The ADAPT program workflow is based on the concepts previously described and stakeholder input (Figure 1). Briefly, the referring practitioner will be asked to administer the PHQ-9 and GAD-2 for symptom measurement. After referral, patients will be scheduled for an ADAPT program screening visit. If the patient meets eligibility criteria and is interested in the program, they are scheduled with an ADAPT therapist. Treatment will begin with a goal-setting session to define treatment and establish diagnosis. Patient progress and symptom tracking will be determined through self-report measures of depressive (PHQ-9) and anxiety (GAD-7) symptoms measured at baseline, at each subsequent appointment while in the program, and at 3 and 6 months post-baseline. Safety assessments and addressing safety issues will follow standard medical center procedures, including reviewing question 9 regarding thoughts of dying on the PHQ-9, structuring a safety plan, and obtaining emergency care if needed. All team members and referring practitioners will receive ongoing education on the model and workflows.

Figure 1:

Outline of patient assessment and triage workflow in the ADAPT model. Patients seeking care for depression or anxiety in primary care are given the Patient Health Questionnaire-9 (PHQ-9) to determine the appropriate care path for a patient. Patients with mild to moderate symptoms are referred to an ADAPT practitioner, who performs further assessment for inclusion and exclusion criteria. ADAPT = Achieving Depression and Anxiety Patient-Centered Treatment.

ADAPT Therapy

After the intake appointment, patients will be scheduled for follow-up every 1 to 2 weeks initially. Each appointment will consist of reviewing content from the last session, symptom inventory, and evidence-based therapy for symptoms. As symptom severity decreases, the therapist will review the patient with the care team and develop a transition plan for the patient. If symptom scores do not improve, additional resources including referrals for medication, treatment planning with the care team, group therapy, or escalating to specialty psychiatric care can be incorporated. In the virtual ADAPT program, ongoing therapist training in brief psychotherapy will be provided for model fidelity.

Medication Management

If the patient expresses interest in medication or is not responding to psychotherapy alone, they can be scheduled with the clinical pharmacist. These appointments could be booked at intake, or at any time during the ADAPT program by any team member. The pharmacist will have an initial evaluation and review symptoms and medication options based on the clinical diagnosis. Clinical pharmacists will choose from several evidence-based treatment options, including antidepressants, augmentation strategies, and antianxiety medications. At any time, the pharmacist can consult with a psychiatrist. Pharmacists will review patients in the program with the care team and decide whether to transfer medication management to the primary care provider or escalate/augment the patient’s care. The pharmacist will collaborate with the primary care physician for any medications or workup needed. Suggestions will then be discussed with the patient so they can make informed health care decisions in line with their values, needs, and priorities.

Results are pending study completion.

Discussion

Depression and anxiety are the two most common mental health conditions in the United States, ¹ yet many patients with depression and/or anxiety do not obtain evidence-based care, ⁵⁸ obtain Health Care Effectiveness Data and Information Set care quality goals in symptom reduction, ^57,59 and have long care wait times. ⁶⁰ Collaborative care has emerged as a powerful model to help patients obtain timely access to high-quality mental health care with improved outcomes. ²⁵ While collaborative care is often successful for patients, there have been few large health care organizations implementing this model operationally given barriers in care manager reimbursement and psychiatrist integration. ²⁹

To address this gap, the authors designed a novel collaborative care program for patients with mild to moderate depression and anxiety. ADAPT integrates several evidence-based concepts shown to improve patient outcomes in depression and/or anxiety treatment while incorporating stakeholder feedback. Novel components to this program include integration of the care manager position into each touch point, utilization of pharmacist-led prescribing, and time-limited psychotherapy. These model differences can be incorporated into generalizable findings for other nonintegrated health care settings depending on local restrictions regarding pharmacist-led prescribing. ⁵⁵ The authors believe this model would perform well both in value-based and full risk-based health systems, as the model is designed to improve patient outcomes through early detection and treatment. By prioritizing measurement-based care, accurate diagnosis, and easy referral and access to care, this program is designed to improve patient access and response to depression and/or anxiety treatments. Additionally, this study aims to implement novel care delivery strategies, including the use of pharmacists and meeting the care manager role through a teams-based approach.

This study has limitations. The program depends on accurate patient diagnoses and symptom tracking. Both variables will be collected at regularly scheduled clinical visits as part of routine medical care. While it is the standard of care to perform symptom measurement using validated self-report measures at each visit and provide the most accurate diagnosis, variations in clinical practice are common, and missing symptom data or imprecise diagnosis would impact patient retention in the program. Additionally, pharmacist-led prescribing is dependent on a clinical provider diagnosing a condition appropriate for medication. The authors will account for these limitations by making the symptom measurement tools accessible in both electronic and paper forms and providing regular reminders at care team meetings.

Conclusion

Limited access to quality and evidence-based mental health care is an important limitation of the current health care landscape. Although collaborative care models ensure recognition of patient symptoms and engagement with evidence-based treatment, they have several features that limit their broad implementation in health care systems. By utilizing value-based and patient-centered concepts, the ADAPT program seeks to expand collaborative care implementation for depression and anxiety. If successful, this model can be applied to the broader health care organization and to other specialty populations.

Supplementary Material

Supplemental Material