We conducted a retrospective analysis of 1992 hospital discharge data to de
termine the incidence of mental illness hospitalizations among elementary-s
chool-aged children and to analyze differences in hospital use by selected
population characteristics. We analyzed population-based records of hospita
lizations of 6- to 12-year-olds (n = 4,460) with a principal diagnosis of m
ental illness and calculated relative risks (RRs) for hospitalization by se
x, race/ethnicity, and payment source. Mental illnesses accounted for 8.1%
of hospitalizations and 28.9% of hospital days for 6- to 12-year-olds. Hosp
ital charges totaled $85 million. Boys had a higher risk of mental illness
hospitalization than girls (RR 1.96; 95% confidence interval [CI] 1.84-2.08
). Latino children had a lower risk than whites (RR 0.22; 95% CI 0.20-0.24)
, as did children in the "Asian/other" group (RR 0.12, 95% CI 0.10-0.15). I
npatient hospitalizations for mental illness have a major impact on hospita
l morbidity for elementary-school-age children. Boys are overrepresented an
d Latinos and Asians/others are underrepresented among mental illness hospi
talizations. Clinical implications for these findings and barriers to the d
elivery of inpatient mental health care are discussed.