Letter

To the Editor: Youths placed in out-of-home settings often receive mental health services from several sectors of care, but the extent to which they use hospital emergency services is uncertain ( 1 ). Better information on the use of emergency services in this population merits attention because such services are costly and disruptive to families.

We used Medicaid Analytic Extract (MAX) claims from North Carolina to examine the use of hospital emergency departments among a statewide population of youths under age 22 who received out-of-home care in congregate residential treatment and family-type residential treatment (often referred to as therapeutic foster care) during calendar year 2003 (N=2,937). MAX claims contain the age, race or ethnicity, gender, Medicaid eligibility category, and a primary ICD-9-CM diagnosis. If the youth had more than one out-of-home stay with different primary diagnoses, we selected the most frequent diagnosis. We used North Carolina claims because these data are of high quality and because youths with a psychiatric diagnosis were not enrolled in capitated managed care, as they are in some states ( 2 ). The data did not permit us to examine whether emergency visits occurred before, during, or after out-of-home stays. With SAS software version 9.1, we used chi square tests to examine whether the demographic and diagnostic characteristics of youths who used emergency services (N=1,616) differed from those who did not (N=1,321). We adhered to confidentiality requirements in a data use agreement with the Centers for Medicare and Medicaid Services.

Fifty-five percent of youths (N=1,616) in out-of-home care received emergency services during the year (mean±SD=2.0±2.7 visits). Forty-nine percent (N=1,453) visited the emergency department for a physical health problem, and 19% (N=558) visited the emergency department for a mental health problem.

Twenty-six percent of girls (N=218) received emergency services for a mental health problem, compared with 16% of boys (N=340) (p<.001). Thirty-one percent (N=107) of youths diagnosed as having major depression or affective disorders while in out-of-home care visited the emergency department for a mental health problem, compared with 18% (N=250) of those diagnosed as having conduct disorder or emotional disturbances (p<.001). Of youths who visited the emergency department for a mental health problem, the proportion who were enrolled in Medicaid because of low income was smaller than the proportions enrolled because of foster care or disability (14%, 22%, and 25%, respectively; p<.001). [A table with data on demographic and clinical characteristics of the total sample and by type of emergency visit is included in an online supplement to this letter at ps.psychiatryonline.org .]

We report these preliminary findings to stimulate research on the use of emergency services among youths in out-of-home care and to suggest that systems of care designed for this population should account for the role of emergency departments. Visits for physical problems may have been common, in part, because youths with emotional and behavioral problems have an increased risk of injuries ( 3 ). Although no diagnostic or demographic characteristics were associated with having an emergency visit for a physical health problem, emergency mental health visits were more common among girls, youths with major depression or affective disorders, and youths who were disabled or in foster care. Further research is needed to examine why these groups may be more likely to use emergency departments and how the use of emergency departments is related to the availability of other services ( 4 ).

This research was supported by contract 280-03-1501 between Mathematica Policy Research and the Substance Abuse and Mental Health Services Administration (SAMHSA). The views expressed are those of the authors and not necessarily those of SAMHSA or the Centers for Medicare and Medicaid Services. The authors are grateful for the support of Jeffrey Buck, Ph.D., and Judith Teich, M.S.W.; for the expertise provided by Allison Barrett, M.A., Deo Bencio, B.S., Ann Cherlow, A.B., and James Verdier, J.D., in guiding the authors' use of Medicaid data; and for feedback provided by David Baugh, M.A.

The authors report no competing interests.

The authors are affiliated with Mathematica Policy Research, Washington, D.C.