- ACG System
- About Us
- Update / SIgnout
MEDTAP 2001-2003 Research Project #1. Cincinnati, OH, USA: University of Cincinnati and Institute for Health Policy and Health Services Research.
University of Cincinnati and Institute for Health Policy and Halath Services Research, Cincinnati, OH, USA
The primary objectives of this study were 1) to describe the characteristics and utilization patterns of Medicaid patients with high health care costs and 2) to assess the risk factors associated with those high-cost recipients. “High-cost” recipients were defined as those whose average monthly Medicaid expense was at or above the 90th percentile of the Medicaid fee-for-service population. Comparison recipients were randomly selected from the 5th through 90th percentiles. Based on this selection process, 23,965 non-nursing home recipients and 14,421 nursing home recipients were studied through retrospective data analysis of Ohio Medicaid claims from 1/1/1999 to 12/31/2000. Adjusted Clinical Groups_ (ACG) software1 was used to categorize population-based clinical characteristics for high cost. Logistic regression analysis was conducted to assess the risk factors associated with high cost utilization. Multiple regression analysis was used to estimate the relationship between Medicaid total expense and its associated economic determinants.
For the sample of non-nursing home recipients, there were 12,222 high-cost recipients with a total service cost of $399 million (61% female, 26.5% black, average age 38.5, 53% disabled or blind, 6.5% deaths, and 45% with 10+ comorbidities) and 11,743 comparison recipients with a total service cost of $30 million (59% female, 26.5% black, average age 22.4, 12% disabled or blind, 0.7% deaths, and 10% with 10+ comorbidities). High cost recipients spent 13 times $cost/per recipient more than comparison recipients. Major cost components were hospitalization, outpatient care, prescription drugs and mental health. For both non-nursing home high-cost and comparison recipients, mental health costs, as a proportion of total cost, was very high (26%-34%) for school-age children ages 6– 17, followed by young adults ages 18-44 (12%), middle-age adults ages 45-64 (5%-10%), preschool children (4%) and elderly persons (2%-5%).
For the sample of nursing home recipients, there were 7,293 high-cost recipients with a total service cost of $1.23 billion (42% female, 15% black, average age 46, 90% disabled or blind, 8.7% deaths, and 54% with 10+ comorbidities) and 7,128 comparison recipients with a total cost of $406 million (72% female, 12% black, average age 81, 12% disabled or blind, 36% deaths, and 35% with 10+ comorbidities). Major cost components were long-term care, mental health services, and prescription drugs. Mental health illnesses were frequently diagnosed for over 65% of high cost recipients, and over 40% of comparison recipients.
Total Medicaid costs varied between non-nursing home recipients and nursing home recipients. The odds ratios of the risk for high cost also varied between both groups, i.e., 1.9- to 6.7-fold higher for disabled or blind recipients; 2- to 4.6-fold higher for recipients who died; 8%-38% higher per comorbidity; 80% higher for urban residents; and 35%-51% higher for males.
In conclusion, most high-cost Medicaid recipients were adults who had chronic diseases and many comorbidities. Controlling some comorbidities for high-risk recipients might control increased future health care costs. It is important to conduct evaluation programs (e.g., drug utilization review or disease management) for high-risk recipients who have severe chronic diseases and many comorbidities, as well as being disabled/blind eligible, short period of enrollment, being male, and/or live in urban areas. Due to high utilization of mental health services among school-age children and adolescents, it might be beneficial for Medicaid to cooperate with School-Based Health Centers that aim to prevent or control mental health illness. For nursing home residents, a major factor should be on the utilization of mental health services.
Please log in/register to access.