Guiding Development of Pilot Projects to Drive Quality Improvement

The Mental Health Quality Forum (MHQF), a 2-meeting multidisciplinary panel of 13 participants, was held in March and June 2013 to identify problems and opportunities in the management of serious mental illness (SMI). In the second meeting, participants reconvened to review issues identified at the first meeting, agree to principles that will guide solution development, and create a blueprint for solutions-oriented pilots across 3 critical areas. Again, they worked in 3 “ecosystem” groups to simulate the tension between real-world stakeholders in SMI care.

Considering The DIAMOND Collaborative Care Model

As a starting point for pilot design, the panelists discussed a recent initiative undertaken to improve collaboration between primary care and mental health providers: Depression Improvement Across Minnesota—Offering a New Direction (DIAMOND). A team-based collaborative care model was used in caring for patients with depression in 75 primary care clinics.1

Primary care clinics were trained on implementation of 4 processes and 2 roles. The processes included a consistent method for assessment and monitoring (the Patient Health Questionnaire-9 [PHQ-9]),2,3 a patient registry tracking system, a stepped-care treatment approach with treatment adjustments based on PHQ-9 scores and use of the Institute for Clinical Systems Improvement depression guideline,4 and a relapse prevention plan created by the care manager. The 2 roles were those of a care manager and a consulting psychiatrist (AV 1).

AV 1. Key Elements of the DIAMOND Study of Collaborative Care for Depression (01:39)


DIAMOND employed detailed quality measures, assessed at 6 and 12 months, that were based on the proportion of patients with PHQ-9 tests administered, PHQ-9 score increases or decreases, and patient suicide statistics.

DIAMOND also employed payment redesign: a single billing code for services was established and used only by certified DIAMOND clinics. This involved a single claim code for any of the following bundled services: patient tracking and registry updates by care manager, care manager contact with patients, care manager use and administration of PHQ-9, and weekly psychiatrist consultation and caseload review with care manager. Associated fees were paid to certified clinics for a set of services covered under the care management program. Payment was based on outcomes, rather than simply process of care. At the end of 2011, across all DIAMOND clinics, 53% of patients with depression achieved remission and 70% achieved response within 12 months of initial assessment.6

Panelists believed that DIAMOND was a useful case study and were particularly interested in the bundled reimbursement code. However, some limitations to implementing a DIAMOND-like pilot program for persons with other forms of SMI were noted. Panelists believed that financial incentives alone were not enough to change provider behavior or patient outcomes. Another limitation noted was that there is no analogous measure to the PHQ-9 for SMI, and current measurement tools are too long to use in clinical settings. Finally, DIAMOND was carried out in primary care settings. Panelists felt that the current health care system is harder for psychiatrists to navigate than PCPs.

Designing a Pilot Program for Change

The panelists next turned to their main objective: designing a blueprint for a pilot that could be implemented anywhere, demonstrate program success, and create momentum for system change. They focused on (1) care coordination, (2) measurement of results, and (3) payment change. Pilots could not be dependent on dramatic changes to infrastructure, should assume that the current fee-for-service (FFS) system driven by Medicare and Medicaid will remain intact except in limited cases, and should include measures that providers can assess and own.

Care Coordination
A list of elements to include in a care coordination pilot for SMI was created:

  1. Define coordination of care
  2. Define roles of stakeholders and patient care team
  3. Provide education and training to implement coordination tools
  4. Conduct risk stratification on an ongoing basis to facilitate care adjustment when needed—include unification of behavioral, social, and community factors
  5. Include a relapse screener and medication protocol
  6. Establish a unified care portal for all stakeholders
  7. Provide information technology (IT) to support care management
  8. Include patient input in goal setting

Care coordination may involve coordination between physical health and behavioral health, coordination within the behavioral health system, and coordination between behavioral health and social service agencies. Centralized care management was hypothesized to produce better outcomes than clinic-based management, as it might be easier to monitor the quality of care.

Care management was distinguished from case management, with care management encompassing the whole person’s well-being and case management focusing on the immediate needs of the patient’s behavioral health condition. The care manager was considered to be the chief coordinator and should be patient-focused and accountable:

“The care manager is attached to the patient. The role entails putting patients and their goals at the center and then figuring out which providers need to be connected in order to achieve those goals.”

Further, a new “chart manager” role was proposed to handle administrative tasks, in order to free up the care manager’s time for client care. Efficient use of information exchange systems was noted as an important component in improving care coordination.

Panelists believed that implementing care coordination will require changing the mental health care culture with regard to how patients are treated. For example, all practitioners participating in SMI patient care will need to be compensated:

“Whatever the practitioner adds to the coordination of care would have a dollar amount that applies to their effort.”

In addition, the forum believed that ownership of care (primary care vs mental health care) would be critical. As one panelist stated,

“The amount of money available in the system to pay for this coordination will depend on where the coordination lives. Is mental health being integrated into primary care, or is primary care being integrated into mental health? The answer will inform all the dollars.”

AV 2. Improved Care Coordination: Designing a Pilot Program for Change (1:16)

Quality Measures

Both positive and negative quality measures were proposed for SMI care coordination. The participants noted that quality measures include both those associated with good-quality care, such as appropriate follow-up care and patient and caregiver satisfaction, and those associated with poor-quality care, such as polypharmacy and emergency department overuse.

Participants asserted that quality measures in SMI should be practical, usable, and actionable and also be part of shared decision-making:

“In developing measures for people with severe mental illness, we need to think about shared decision-making. For patients with severe mental illness, sometimes people ask, ‘Is shared decision-making feasible?’ But patients are eager to be a part of the decision, as studies have shown.”

It was noted that quality measures should allow risk stratification and adjustment in terms of SMI severity. Measures should also leverage Centers for Medicare and Medicaid Services (CMS) quality reporting,7 since Medicare utilizes a wide variety of quality reporting measures that also may apply to the non-Medicare population. Care transitions were mentioned as a current problem area and therefore a particularly important target in quality measurement.

Enabling Financial Structures

Defining care coordination. The payment change breakout group defined care coordination as overall health care rather than mental health care, due to the very high expenditures associated with physical health care in the SMI patient population.

Reimbursement for care coordination. Reimbursement for care coordination in SMI is not widespread. Medicare and Medicaid pay for most SMI care via FFS. Because dual-eligible patients account for huge expenditures, panelists thought they would be a good target for an initial care coordination pilot, and early initiatives among these patients may be more likely to lead to payer acceptance.

Outcomes measures for payment increase/decrease. The care coordination breakout group identified outcomes measures for a payment decrease and for a payment increase. For a payment decrease, outcomes measures included hospitalizations, emergency room visits, and high cost; those for a payment increase included health outcomes (such as glycated hemoglobin levels), psychiatric measures, medication adherence, screening for chronic diseases, postdischarge visits, and improvements in quality of life. Panelists believed that such changes to payment may result in positive reactions from both payers and providers if the measurements were perceived as meaningful and impactful, and they asserted that accountability by all stakeholders would be crucial for success.

Successful incentivization. Three critical elements of successful SMI care coordination and incentivization were identified:

  1. Defining care coordination services
  2. Identifying payment and potential shared savings
  3. Using data and technology to support care coordination

Payment and potential shared savings should focus initially on short-term outcomes and include both mental and physical health outcomes and all potential providers. Effective use should be made of electronic health records, actuarial analysis, payer databases, and CMS data to support incentivization of care coordination processes.

Panelists also pointed to the need for a system of accountability to monitor appropriate incentives and comprehensive payment awareness that would increase per-member per-month payment to include care coordination in capitated arrangements and ensure appropriate case-mix adjustment for care coordination in FFS systems. Incentives for non–health care providers, such as judicial systems and child welfare services, must also be considered.

Other payment-related issues. Participants believed that the bundled coding scheme in DIAMOND was useful in allocating payment and suggested that a bundled code be developed for use in an SMI pilot. Importantly, the ideal pilot would need to be applicable to many states. Data would be required for measurement, but would also need to be sustainable and facilitate communication across provider types and payers. The pilot should focus on managed care plans, not state Medicaid FFS, due to state plan amendment issues and regulatory barriers.

  • Improved care coordination is critical to improved patient outcomes in SMI, while quality measures and changes to payment structures are also important. Improved physical health was identified as a suitable initial target because of its importance to payers on multiple levels and because it can be easily measured by specific indicators.
  • Other metrics suggested by participants for SMI pilots include healthy lifestyles, healthy eating habits, smoking cessation, and symptom reduction. SMI should be conceptualized as a chronic illness, like hypertension, and those involved in the SMI patient’s care should work together to treat physical and mental health as a team.
  • Case management must be remodeled to care management in order to facilitate improved care coordination.
Next Steps: Hand-off of the Findings

Participants could not single out a particular organization for hand-off of the forum’s findings. Organizations mentioned for possible involvement included the National Association of State Mental Health Program Directors; Association for Behavioral Health and Wellness; and primary care organizations, particularly those active in implementation of the ACA. Finally, the participants felt that hand-off may need to be approached on a state-by-state basis.

  1. Solberg LI, Crain AL, Jaeckels N, et al. The DIAMOND initiative: implementing collaborative care for depression in 75 primary care clinics.Implement Sci. 2013;8(1):135. doi:10.1186/1748-5908-8-135 PubMed
  2. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16(9):606–613.doi:10.1046/j.1525-1497.2001.016009606.x PubMed
  3. Löwe B, Kroenke K, Herzog W, et al. Measuring depression outcome with a brief self-report instrument: sensitivity to change of the Patient Health Questionnaire (PHQ-9). J Affect Disord. 2004;81(1):61–66.doi:10.1016/S0165-0327(03)00198-8 PubMed
  4. Institute for Clinical Systems Improvement. Depression, Adult in Primary Care. 2013. . Accessed March 21, 2014.
  5. A new direction in depression treatment in Minnesota: DIAMOND program, Institute for Clinical Systems Improvement, Bloomington, Minnesota.Psychiatr Serv. 2010;61(10):1042–1044. doi:10.1176/ps.2010.61.10.1042 PubMed
  6. Institute for Clinical Systems Improvement. The DIAMOND program: transforming depression care delivery and payment models. 2012. . Accessed March 21, 2014.
  7. Physician Quality Reporting System. Centers for Medicare and Medicaid Services website. . Updated June 9, 2014. Accessed August 5, 2014.

09US15EUP0050 April 2015

Mental Health Quality Forum Participants

Carol L. Alter, MD

Georgetown University

Jacob Balinky, PMHNP-BC, MN

LifeWorks NW

Rakesh K. Bansil, MD

Catholic Charities-Newark

Jeffrey A. Buck, PhD

Centers for Medicare and Medicaid Services

Douglas Del Paggio, PharmD, MPA

Alameda County Behavioral Health Care Services

Adrienne Hlavati, APRN

Mental Health Cooperative

Charles S. Ingoglia, MSW

National Council for Behavioral Health

Matthew M. Keats, MD, MMM


David L. Larsen, RN, MHA


Stuart H. Levine, MD, MHA

Accretive Health

Wayne Lindstrom, PhD

Mental Health America

Junqing Liu, PhD, MSW

National Committee for Quality Assurance

Nina Marshall, MSW

National Council for Behavioral Health

Clare Miller

American Psychiatric Foundation

Marcia A. Palmer, PharmD, MBA

UnitedHealth AmeriChoice

Mark J. Peterson, RPh

Genoa Healthcare

Andrew Sperling, JD

National Alliance on Mental Illness

Bradley D. Stein, MD, PhD

RAND Corporation

The meetings, manuscript development, and development of this brief report were supported by Otsuka America Pharmaceutical, Inc. All participants received a fee for service from Otsuka America Pharmaceutical, Inc. for participation in the meetings.

J Clin Psychiatry 2015;76:e00

​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​ The Journal of Clinical Psychiatry The Primary Care Companion for CNS Disorders The CME Institute Neurology Knowledge Terms of Use Privacy Policy