Validity of Medicaid household respondent reporting of ambulatory visits for mental disorders.

Comparison with administrative records or "best estimate file" enables an evaluation of the accuracy of household reports of mental health use in the four-State Medicaid Household Survey conducted as part of the National Medical Care Utilization and Expenditure Survey. Underreporting of probability of ambulatory mental health use ranged from 14 to 24% compared to 5 to 7% for ambulatory health visits; household estimates of number of mental health visits seemed to be more accurate than administrative records. Household reporting of provider type seemed to be very accurate for psychiatrist visits, but there seemed to be a tendency to report psychologist visits as psychiatrist visits.