County-level estimates of need for mental health professionals in the United States.

Objective: The goal of this study was to develop the best current estimates of need for mental health professionals in the United States for workforce planning and to highlight major data gaps.

Methods: Need was estimated indirectly, on the basis of several steps. The 2001 National Comorbidity Survey Replication (NCS-R) (N=9,282) was used to model the probability of having serious mental illness, given demographic predictors. Synthetic estimation was then used to construct national and county-level prevalence estimates for adults in households. Provider time needed by these adults was estimated from NCS-R respondents with serious mental illness who used mental health services (N=356); provider time needed by adults without serious mental illness was estimated from respondents to the 2000 Medical Expenditure Panel Survey (MEPS) (N=16,418). National mental health professional workforce practice patterns were used to convert need estimates to full-time equivalents (FTEs).

Results: Adult service users with serious mental illness typically spend 10.5 hours per year with nonprescriber mental health professionals and 4.4 hours per year with prescriber mental health professionals or primary care physicians in mental health visits; adults without serious mental illness spend about 7.8 minutes with nonprescriber mental health professionals and 12.6 minutes with prescriber mental health professionals or primary care physicians in mental health visits per year. With adjustment for mental health services provided by primary care practitioners, the estimated 218,244,402 members of the U.S. adult civilian household population in 2006 required 56,462 FTE prescribing and 68,581 FTE nonprescribing mental health professionals.

Conclusions: Available data indicate that need across the United States varies by demography and geography. These estimates are limited by several issues; in particular, they are based on current provider treatment patterns and do not address how much care ideally should be provided and by whom. Improved estimates will require refined standards of care and more extensive epidemiological data.