Care for mentally ill veterans is as good or better than in other health systems, study finds

The study found that while the proportion of veterans who received the care recommended for their mental illness varied widely, the overall quality of mental health care offered by the U.S. Department of Veterans Affairs was as good as or better than that reported by privately insured, Medicare or Medicaid populations. The findings are published online by the journal Health Affairs and will also will appear in its November edition.

"While the VA does a better job at providing mental health services than other health care systems, there is still substantial room for improvement," said Dr. Katherine Watkins, lead author of the study and a senior natural scientist with RAND, a nonprofit research organization. "With some changes, the VA could provide even better and more cost-effective care for the nation's veterans, as well as serve as a model for other health care systems."

Researchers say the VA already has undertaken several mental health-specific initiatives that may increase the proportion of veterans receiving evidence-based treatments.

The RAND study is the first comprehensive look at the full spectrum of clinical services provided to veterans with mental health issues, from assessment to treatment to chronic care management. It also examines variations in care for veterans across different regions of the United States and for different diagnoses.

Between fiscal years 2004 and 2008, the number of veterans diagnosed with mental illness and substance-use disorders increased 38 percent to 906,394, with the greatest increase seen among those with post-traumatic stress disorder. While the veterans in the RAND study represent only 15.4 percent of those who used VA services in fiscal year 2007, they accounted for 32.9 percent of the total health care costs because of higher use of both inpatient and outpatient physical and mental health care services.

In 2006, the VA commissioned RAND and the Altarum Institute to do a comprehensive evaluation of its mental and substance-use treatment system. The evaluation took place during the implementation of the VA's five- year Mental Health Strategic Plan, a large initiative to expand and improve care for mental and substance abuse disorders.

The study included 836,699 veterans who received services from the VA in fiscal year 2007 for treatment of at least one of five diagnoses: schizophrenia, bipolar I disorder, post-traumatic stress disorder, major depression and substance-use disorders. To be included, veterans had to have had at least one inpatient episode or two outpatient visits.

Veterans in the RAND study had four times as many acute inpatient discharges as non-study veterans and three times as many outpatient encounters. The average cost per non-study veteran was $4,579, while the average cost for a study veteran was $12,337, or nearly three times as much.

The kind of care the veterans received varied widely, although none of the individual facilities stood out as consistently performing above or below the network average. Watkins and her colleagues identified 23 performance indicators to track the kind of care veterans got, such as whether veterans with post-traumatic stress disorder were given cognitive behavioral therapy as recommended. Of those indicators, three had rates about 75 percent, and nine were below 25 percent.

For example, among the assessment indicators, 82 percent of the study veterans were assessed for suicidal thoughts, but only 23 percent were assessed for response to psychotherapy. Use of medication varied from a high of 60 percent for the acute treatment of depression to a low of 16 percent for drug therapy for alcohol dependence. Less than a third of the veterans diagnosed with schizophrenia or bipolar disorder received continuous maintenance treatment with antipsychotics or mood stabilizers.

The study data does not reflect instances when a service or medication was recommended by a provider, but refused by a patient, or prescriptions that were written by the doctor, but never filled by the patient.

In addition to VA's ongoing improvement efforts, researchers say efforts are needed to better understand what underlines the quality differences and strategies should be developed to address the differences in the future.

Coauthors of the study are Harold Pincus, Susan Paddock, Carrie Farmer, Melony Sorbero, Marcela Horvitz-Lennon, Kimberly Hepner, and Jacob Solomon, all of RAND; and Brad Smith, Abigail Woodroffe, and Cathy Call, of the Altarum Institute; and Thomas Mannle, Jr., with Pilot Consulting Services.

Report.