The sensitivity of mental health care use and cost estimates to methods effects.

The authors determined the sensitivity of estimates of the use and cost of outpatient mental health care to two methods effects: the definition of a mental health visit and strategies for allocating mental health care costs. They use data from the Rand Health Insurance Study, which has a random sample of the nonaged noninstitutionalized civilian population in six United States sites. Estimates of the use of mental health specialists are insensitive to alternative methods. However, estimates of the use and cost of the mental health care delivered by nonpsychiatrist physicians (e.g., internists) are quite sensitive to methods effects. Nevertheless, the cost of care from nonpsychiatrist physicians is so low that the total cost of outpatient mental health care is not meaningfully affected by methods effects.