How the U.S. Army Redesigned Its Mental Health System

October 16, 2017 | By Jayakanth Srinivasan, Christopher G. Ivany, Dennis Sarmiento, Jonathan Woodson

Leaders in today’s complex health care systems need better processes and systems for aligning day-to-day, clinical-care activities with the strategic goals of their organizations. The U.S. Army has accrued valuable experience in this area over the last decade through the design and implementation of its behavioral health system of care, its term for all of the mental health and substance-use clinical care it provides. Accounting systems now uniformly capture administrative data across a huge, geographically dispersed system. Workload standards allow clinicians the time to engage with the key people who affect recovery, such as commanders and family members. And self-reported patient data is combined with other nonclinical measures to produce a more accurate assessment of quality. Understanding how the Army did it can help other health systems achieve the triple aim of improving the patient experience of care, improving population health, and reducing per capita cost.

From 2007 to 2012 the Army health care system grew rapidly — with little central coordination — to meet the surging care needs from the wars in Iraq and Afghanistan. The utilization of its outpatient, behavioral health care services more than doubled, and there was wide variation in the type and quality of those services in each location. Consequently, in 2012 the Army focused on reducing that variation and implemented a redesigned performance management system to address key cost and performance drivers. Here are the principles it employed that other health care systems could adopt:

Build standardized care on a foundation of uniform accounting codes. The initial, rapid, ad hoc expansion of services from 2007 to 2012 resulted in the creation of over 211 clinical programs, whose performance was captured through an unwieldy mix of 94 unique administrative accounting codes. Because those accounting codes were inconsistently used from one location to the next, there was no way to easily compare performance across the global network of 33 Army medical centers, community hospitals, and clinics.

When soldiers, family members, and leaders voiced concerns about the inconsistent quality of treatment, the Army rigorously redesigned the way it delivered care. The redesign culminated in the specification of a standardized system of care that consisted of 11 clinical “microsystems,” ranging from primary care to inpatient care. Now that the implementation is complete, a soldier accessing behavioral health services at Fort Bragg, North Carolina, can expect to receive the same care experience as a soldier in Seoul, South Korea. Like other health systems undergoing similar changes, the Army had to harmonize its legacy accounting infrastructure with the new care system so that administrative data accurately mirrors where the care was provided, who provided the care, and what care was provided.

Establish provider-performance standards that reflect both clinical and nonclinical work. The medical mission of the military encompasses both clinical care and occupational health (including leader consultation and education). But as of 2012, Army providers were only assessed on their ability to meet clinical-workload standards (for example, psychologists were required to provide 0.75 FTE, or six hours of clinical care per day). Providers therefore prioritized clinical care and often didn’t engage other nonclinical stakeholders who are essential for helping a patient recover, such as command teams and family members. For example, a provider might not inform a commander when a soldier could be around live weapons fire so that the leader could plan the soldier’s work accordingly and make sure other team members didn’t see that person as a slacker.

The Army addressed this disconnect in clinician incentives by developing new workload standards that reflect the type of care being provided. For example, a provider in a behavioral health clinic (which has greater nonclinical requirements) only has to satisfy a 0.65 FTE clinical-care workload. The rest of the workload is dedicated to shaping the occupational environment through activities such as leader education and being present in the workplace. This workload standard also freed up time for multidisciplinary treatment-planning meetings, where care teams from behavioral health and primary care could meet to discuss high-risk patients. For health systems transitioning to value-based care, it is critical to refine workload standards that account for the nonclinical workload with the same fidelity as clinical-workload standards.

Combine structure and process measures of system performance with patient self-reported data to gain a holistic understanding of care quality. Like most health care systems, the Army’s has historically relied on measures of structure (for example, staffing levels in Army clinics) and process (screenings after return from Iraq or Afghanistan) as proxies for performance. Now that the administrative data on the delivered care is trustworthy, episodes of care can be constructed that longitudinally track a soldier’s care and identify and measure potential safety risks.

A web-based Behavioral Health Data Portal has created new possibilities by systematically collecting and using patient-reported data on clinical outcomes and the patients’ care experiences. For instance, the Army can now examine the quality of the care for post-traumatic stress disorder from one location to another, going beyond proxy measures such as number of psychotherapy sessions in which an evidence-based treatment was used, to more rigorous measures such as significant reduction in distress recorded in the patient-reported data on outcomes.

Adopt a service-line-management process to systematically connect strategic goals to actual clinical care. The Army’s service-line-management strategy creates a virtuous learning cycle that allows senior leaders, clinic chiefs, and care providers to use the same underlying data to make decisions. Senior Army leaders can assess whether sufficient care capacity exists to meet known and projected population health needs, based on the staffing levels across the clinical microsystems. These leaders can now define and assess progress in achieving more-granular goals for improving quality of care for specific diseases.

For example, the Army has defined a strategic goal for improving quality of care for soldiers diagnosed with major depressive disorder. The standardized measure examines whether a soldier received at least four psychotherapy sessions within 90 days of diagnosis. Consistent tracking of this measure enables the care team to identify and follow up with patients who may be at risk for not receiving adequate care. Leaders at the system level compare performance of each hospital over time and in the context of similar facilities across the Army.

In addition, clinic chiefs can not only examine the productivity of individual providers within a department in their clinic but also compare individuals’ and departments’ productivity to that of other locations across the Army. These productivity reports are shared with all providers, promoting transparency across the health care system. At monthly review and analysis meetings, clinic chiefs and the Army service-line-management team discuss selected metrics and ways to improve performance. These meetings also provide opportunities for clinic chiefs to discuss errors in data and the limits of the underlying tools.

As health care systems transition from an encounter-based view of health care to a value-based focus on improved patient outcomes and satisfaction and reduced costs, they must map the system of care to identify the clinical microsystems and redesign the care experience. Legacy accounting systems must be updated to accurately reflect the care being provided and to create a common foundation for decision making by leaders, clinic chiefs, and care providers. And workload specifications must incorporate key nonclinical care activities.

The shift to patient-centered care also requires the routine collection and use of patient-reported data, the sharing of provider-productivity data to better plan capacity, and regular review of progress toward system goals. Only then can senior leaders, clinic chiefs, and care providers coordinate their decisions — even when they vary in scope and magnitude — to align day-to-day care with overarching strategic goals.

hbr.org