Impact of a Service Line Management Model on Behavioral Health Care in the Military Health System

The Army’s transition to clinical service lines began in 2011 when it shifted more authority for policy making, strategic planning, and program development to the headquarters level and away from individual hospitals. To operationalize the change, the service line model was selected to manage each major clinical area, beginning with behavioral health. The systems-level BHSL team, led by the Army director of psychological health, included administrators, analysts, resource managers, and clinicians.

The BHSL changed the management and delivery of care in several ways. First, to reorganize hospital- and clinic-level operations, the systems-level team implemented a policy that eliminated all departments organized around discipline (psychiatry, psychology, and social work) and formed a single Department of Behavioral Health at each Army hospital. The BHSL also integrated disparate clinical programs into a single system of care. In 2012, BHSL leaders identified more than 60 behavioral health programs that had been developed by Army hospitals to address the surge in behavioral health conditions attributed to the wars in Iraq and Afghanistan over the prior decade. Because the expansion had not been previously managed at the system level, wide variation in accessibility, quality, and efficiency had developed. The BHSL identified a group of exemplary clinical programs and worked to replicate them in all hospitals. BHSL leaders developed change management strategies, including clinical staff training programs and manuals; redesigned administrative processes, such as clinical personnel rosters and workload standards; disseminated clinical practice guidelines for high-incidence conditions, such as posttraumatic stress disorder (PTSD); and ensured that resources were available to hire staff and renovate facilities. The BHSL also reconfigured administrative data sources and analytic processes to display accurate and timely information about clinician performance. The BHSL used the newly created data to develop a suite of metrics and used those metrics to inform an oversight process that monitored program implementation and identified variance between hospitals.

The exemplary clinical programs identified as best practices for widespread dissemination were embedded behavioral health (EBH), intensive outpatient programs (IOPs), the child and family behavioral health system (CAFBHS), and a clinical outcome platform called the Behavioral Health Data Portal (BHDP).

EBH is a model for providing outpatient behavioral health care to soldiers assigned to combat units. EBH relocates behavioral health providers from centrally located hospitals to smaller clinics placed as close as possible to where soldiers live and work. Between 2012 and 2017, the Army implemented 61 EBH clinics across the Army. IOPs provide an intermediate level of care for patients who require more frequent visits than can be provided in an outpatient clinic but who do not require inpatient care. Between 2015 and 2017, the Army established 37 IOPs at 19 hospitals. CAFBHS clinics deliver outpatient care to family members of soldiers on active duty and provide consultation to primary care providers. CAFBHS includes a school-based behavioral health program. Providers in the school-based program deliver care in small clinics within the school and consult with teachers and counselors. Between 2014 and 2017, a total of 31 CAFBHS clinics and 15 school-based behavioral health clinics were created. BHDP is a Web-based application that collects patient-reported outcome measures by using standard clinical instruments. Implementation began in 2012 and grew steadily. By 2017, BHDP was used in 834,481 outpatient encounters, 70.3% of all individual encounters with adults in that year.

To determine the relationship between the implementation of the service line and performance at the hospital level, an observation period of 2013 to 2017 (fiscal years) was defined. It encompassed the establishment of service line management and the major structural and process changes made by BHSL leaders. This column focuses on five utilization metrics and three clinical outcome metrics. Outpatient market share is the portion of outpatient behavioral health care provided in Army hospitals and clinics divided by the total outpatient behavioral health care provided in Army hospitals and in the civilian TRICARE network. The measure “PTSD and depression treatment continuity” indicates the proportion of patients with a new diagnosis of PTSD or depression (based on ICD-9 or -10 codes in the first three diagnostic positions) who received a minimally adequate number of treatment sessions, defined as four or more encounters in a behavioral health clinic in the first 90 days (three visits in addition to the one in which the diagnosis was initially made). Inpatient bed-days is the total number of days spent by soldiers on active duty and their family members during an admission to any hospital for behavioral health and substance use disorder services. Clinical outcome metrics captured how frequently patients achieved a clinical response or remission for PTSD, depression, and generalized anxiety disorder, three of the most common disorders in the Army population, on the basis of symptom change on standardized clinical instruments.

The evaluation found that during the observation period, outpatient market share increased, indicating that the Army was able to meet more of the demand for care within its own facilities and referred fewer patients to outside providers through the TRICARE network (Table 1). A decrease was noted in the use of inpatient services, with the largest change between 2013 and 2015. The data also show a trend for more consistent attendance at outpatient follow-up visits among patients with a diagnosis of PTSD or depression. Clinical outcomes showed generally consistent trends in achievement of clinical response or remission, with some fluctuations from year to year. Approximately 35%−44% of patients with PTSD, depression, or generalized anxiety disorder showed clinically meaningful responses on standardized clinical scales.

This column describes the implementation of the service line management model in a large, multisite health care system. Through the service line structure, system-level leaders directly influence hospital-level decisions by issuing binding directives (orders) to hospitals, establishing and reviewing metrics, and shifting financial incentives to the highest clinical priorities.

Evaluation of the service line model showed a decrease in use of inpatient behavioral health care during the observation period, despite little change in outpatient utilization. This finding suggests that the decreased use of inpatient care was not attributable to a reduction in the overall demand for behavioral health care. Although the evaluation design did not permit any inferences regarding causal links, the temporal relationship between the decrease and the implementation of EBH as the standard structure for outpatient care suggests that it played a role. A similar temporal association between EBH and reduced inpatient utilization was found at an Army post in an evaluation of the EBH pilot program in 2010 (during the peak of war activities in Iraq and Afghanistan) (5). EBH may have focused greater treatment resources on patients with PTSD and depression. During the observation period, an increasing percentage of patients with those conditions received at least four outpatient visits during the 90 days after their diagnosis. The roughly 70% rates of achievement of minimally adequate care, as defined by a standard health services definition, is far higher than has been reported previously, either in the military or the U.S. Department of Veterans Affairs (6, 7).

By the end of the observation period, all Army hospitals routinely collected clinical outcome data, a critical step toward value-based care. As of 2017, between 33% and 40% of patients with PTSD, depression, or generalized anxiety disorder achieved at least a 10-point improvement or reduced their symptom scores to a level consistent with clinical remission in the first 6 months of outpatient treatment. Few health care systems have published similar outcome data, limiting the ability to make comparisons. However, one highly regarded health system found that 40% of its patients with depression who had scores ≥10 on the nine-item Patient Health Questionnaire (PHQ-9) achieved improvement between the initial visit and follow-up (an average of 131 days later), defined as a 5-point decrease in PHQ-9 scores (8). Although the Army cohort is different from the one in that study, the evidence indicates that the care provided by Army clinics under a service line model compares favorably with national benchmarks. The clinical improvement rates for PTSD documented in this evaluation study are also highly comparable to data from randomized clinical trials of the efficacy of PTSD treatments for veterans (9).

This study had important limitations. Because the data were collected in the midst of servicewide changes to the health care system, no comparison group was available, and the strength of the data are largely limited to temporal observations and consistency across outcomes. It is impossible to definitively establish causal relationships between the implementation of the service line model and health outcomes. In addition, the decrease in pace and scale of deployments to Iraq and Afghanistan and changes in recruitment policies in the latter years of the observation period could have influenced outcomes, although outpatient utilization remained consistent during this period.

This evaluation of the Army’s BHSL was intended to provide preliminary evidence concerning temporal and logical connections between the change to the service line model, the structural changes made by service line leaders (standardization of clinical programs), trends in service utilization (improved patient engagement in outpatient care), changes in clinical outcomes (reduced symptom burden of common conditions), and population health (decreased need for inpatient care). The Army’s experience indicates that the service line model may help health care systems reduce variation between geographically distinct care delivery locations and improve total system performance. More rigorous quasi-experimental or controlled studies are needed to examine the impact of service line models of care.

The authors acknowledge Jamie Moore, Psy.D., Ron Dean, Cheri Ruiz, Osceola Evans, Mike Faran, M.D., Kendra Lebel, and Ranilo Laygo, Ph.D., for invaluable contributions. The views expressed are those of the authors and do not necessarily reflect the official policy of the Department of Defense, Defense Health Agency, Department of the Army, U.S. Army Medical Department, or the U.S. government.