An evaluation of statewide emergency department utilization following Tennessee medicaid disenrollment

Objectives: During a series of reforms to the Tennessee Medicaid expansion program (TennCare) in 2005, approximately 171,000 adults were disenrolled from Medicaid. The objective of this study was to examine the statewide effect of such a disenrollment on Tennessee emergency department (ED) utilization. Methods: Administrative data on all ED visits in Tennessee from 2004 through 2006 were obtained from the State Emergency Department Database and State Inpatient Database under the Healthcare Utilization Project. Population statistics and uninsured estimates were obtained from the U.S. Census Bureau, and TennCare enrollment data were obtained from the State of Tennessee Department of TennCare. The proportion and rate of ED visits, assessed separately by payer type, were compared across a predisenrollment period from January 1, 2004, through July 31, 2005, and across a postdisenrollment period from August 1, 2005, through December 31, 2006. The proportion and rate of ED visits resulting in hospital admission, again assessed separately by payer type, were compared across the same disenrollment periods in a similar way. We fitted a series of linear models for the total number of ED visits and each proportion and rate, with various degrees of adjustment for seasonality and time trend. Results: The mean number of ED visits was 45,662 per week during the predisenrollment period and 44,463 per week during the postdisenrollment period (mean difference = -1,199; 95% confidence interval [CI] -1,722 to -676). By payer category, there was a decrease of 3,119 visits per week by TennCare beneficiaries and an increase of 2,203 per week by the uninsured. After disenrollment, the absolute proportion of ED visits by TennCare beneficiaries significantly decreased by 6.2% (95% CI = -6.6% to -5.8%), and the absolute proportion of uninsured ED visits increased by 5.3% (95% CI = 4.9% to 5.7%). The rate of ED visits by TennCare beneficiaries decreased by -0.091 ED visits/person/year (95% CI = -0.136 to -0.046) in the disenrollment period when controlling for time and seasonality. The rate of ED visits among the uninsured increased by 0.038 ED visits/person/year (95% CI = 0.011 to 0.065) in the postdisenrollment period when controlling for cubic time trend. The proportion of all TennCare ED visits that resulted in hospital admission did not change significantly between the two periods after adjusting for time trend and seasonality. The proportion of uninsured ED visits resulting in hospital admission, however, significantly increased after disenrollment by 2.0% (95% CI = 1.8% to 2.2%) and by 0.6% (95% CI = 0.0% to 1.2%) after adjusting for time and seasonality. Conclusions: The TennCare disenrollment of 2005 was associated with a modest decrease in the number of total ED visits in Tennessee. However, the payer mix among the Tennessee ED population shifted abruptly. The increased rate of ED visits by Tennessee's uninsured and the increased proportion of uninsured ED visits leading to hospital admission suggest an increased burden of illness in this highly vulnerable population.