UC Berkeley Press Release
Researchers release first report to compare mental health indicators in California counties
BERKELEY – In the first attempt to compare measures of mental health and general well-being among California's general population on a county-level basis, health policy researchers at the University of California, Berkeley, found significant differences.
A number of counties, including Sacramento and San Diego, scored higher than expected -- with fewer reports of mental health problems -- based upon their population's socioeconomic and demographic characteristics. At the same time, Alameda and San Francisco counties were among those that scored lower than expected.
The report, "Measuring Mental Health in California's Counties: What Can We Learn?" was released today (Wednesday, Jan. 19) by the Nicholas C. Petris Center on Health Care Markets & Consumer Welfare at UC Berkeley's School of Public Health.
"We expected that factors such as gender, income level and amount of education would explain a certain amount of the variation in mental health indicators across counties," said Timothy Brown, associate director of the Petris Center and co-author of the report. "The fact that variations across the state were found even after we adjusted for socioeconomic and demographic factors suggests that other factors are at play."
The analysis includes data from 55,428 adults who participated in the 2001 California Health Interview Survey (CHIS), conducted by the UCLA Center for Health Policy Research. The researchers compared responses to standard questions used to indicate mental health and general well-being, such as limitations on general activity or work due to emotional problems, sadness, anxiety, lack of energy, binge drinking, and perceived need for mental health care.
"The questions were not meant to make a clinical diagnosis for the respondents, but were used to construct a comparative picture of mental health indicators in California counties," said Daniel Eisenberg, formerly a research associate at the Petris Center and lead author of the report.
The researchers combined 25 of the smaller-population counties, including Modoc, Colusa and Mariposa counties, into eight county groups to correspond with the CHIS sampling design. Based upon the responses, they created scores for the counties that ranged from a low of 0 to a high of 10. The report includes a color map of the state showing each county's score.
In addition to San Diego and Sacramento, other counties and county groups that scored a high 7 out of 10 include Butte, Nevada/Plumas and Riverside. Alameda and San Francisco both scored a 3 in the report, the lowest in the state. The researchers note that San Francisco's ranking was low despite the relatively large number of mental health providers in the county.
Among the many other factors the researchers compared was state funding for mental health services. They report that statewide in fiscal year 2000-2001, the per capita average for mental health funding was $66. The highest ratio of dollars per population was in Santa Cruz at $140 per capita. Sacramento, which ranked high on mental health measures, had the lowest ratio of mental health spending at $16 per capita.
"This report raises more questions then it answers, but the questions are really important," said Richard Scheffler, UC Berkeley professor of health economics and public policy, director of the Petris Center and co-author of the report. "Why do the self-reported indicators of mental health differ so widely among the California counties even when you adjust for socioeconomic differences? Do the counties that scored higher have better-organized mental health systems? How are they spending their money? Are their treatment programs different? Finding the answers to these questions requires us to dig down and understand the mental health systems -- both public and private -- available in these counties."
In calculating their results, the researchers used Los Angeles County as a baseline because it is near the middle of the distribution for most of the mental health measures used in the study.
"County by county comparisons are always difficult because California is a large and very diverse state -- geographically, economically and culturally. Even within counties there are significant amounts of diversity," said Dr. Neal Adams, director of Special Projects for the California Institute for Mental Health and one of the report's external reviewers. "That being said -- and recognizing that the size of Los Angeles County alone tends to drive the mean of any measure -- making comparisons to Los Angeles as the midpoint is a reasonable approach."
The researchers note that prior to the report's release, the only information available came from medical claims records of people who sought out mental health services.
"This report gives information on a broader denominator of the population," said Dr. Harold Alan Pincus, professor and executive vice chair of psychiatry at the University of Pittsburgh School of Medicine, director of the RAND-University of Pittsburgh Health Institute, and one of the report's external reviewers. "Counties will find the report useful because they can compare themselves with each other."
Adams noted that California voters recently passed Proposition 63 in the November election, giving new incentive for counties to expand their mental health programs. The state legislation places a 1 percent tax on the adjusted gross income of residents who make at least $1 million to be used specifically for mental health services.
"Policymakers and planners in California will be able to see variations across counties and consider what factors may be contributing to the differences in mental health scores," said Eisenberg, who began this research as a UC Berkeley post-doctoral student and is now an assistant professor of health management and policy at the University of Michigan. "This report is a starting point."
Nicole Bellows, a research associate at the Petris Center, also co-authored the report.
The full report can be found online at http://www.petris.org.