| The Institute of Medicine has documented a variety of problems facing the nation's emergency services system. This dissertation draws on the health care performance measurement literature to propose three system measures of emergency department (ED) performance in the area of children's mental health. The three measures are: annual utilization rate; rate at which visits to emergency departments result in an inpatient admission, or the conversion rate; and, for youth discharged from the emergency department, the follow-up rate. Services data come from the Bureau of TennCare's encounter and enrollment data system for state fiscal years 1997 to 2002. Supplemental information comes from state and other surveys of hospitals, provider registries, and U.S. Census 2000. The measures are calculated for youth ages 4-17 that have used a service with an ICD-9 diagnosis in the range 290-315. The analysis uses multi-level regression models to adjust for the clustering of measures among youth, counties, and providers. Selection of covariates was guided by the Aday-Andersen behavioral model of service utilization, which organizes covariates around predisposing, enabling, need, community-environmental, and provider variables.;This dissertation demonstrated the feasibility of administrative data-based performance measures for children's mental health services in a Medicaid managed care program. The principle findings were that at the end of the study period odds of ED utilization were five times greater than at the start of the study period; the system decreased the rate at which youth were converted from the ED to inpatient; and, fewer than one in five children discharged from the ED received a specialty mental health outpatient treatment within 30 days of discharge. Additional findings were that many ED visits, including those resulting in inpatient admission, carry an adjustment disorder diagnosis; and, one-third of ED youth who use the ED in a given year are at the beginning of a treatment episode. |
