NASHVILLE, TENN. Proposals to increase cost sharing for Medicaid beneficiaries could reduce enrollment in the program, according to the preliminary results of a study presented at the annual conference of the National Academy for State Health Policy.
Genevieve Kenney, principal research associate at the Urban Institute, and her colleagues examined the impact of premium increases in the State Children's Health Insurance Program (SCHIP) as a way to inform policy changes under Medicaid.
More than 30 states have premiums for some children whose incomes are above the federal poverty level. No existing Medicaid program charges premiums for children below poverty, Ms. Kenney said.
However, proposals, such as one from the National Governors' Association, would permit states to charge up to $480 annually per child to low-income families.
The research, which was funded by the David and Lucile Packard Foundation, looked at enrollment and disenrollment patterns in three states that increased premiums in 2003Kansas, Kentucky, and New Hampshire.
SCHIP officials in Kansas increased premiums from $10 to $30 per family in February 2003 for families between 151% and 175% of the federal poverty level. The state then decreased the premiums to $20 in July 2003. For families between 176% and 200% of poverty, the premium was increased from $15 to $45 and then decreased to $30.
The total caseload growth rate 6 months before the premium increase in Kansas was 14.6%. Six months after the increase the growth rate had fallen to −4.2%, Ms. Kenney reported. Although there was an initial drop in enrollment, the caseload picked up over time, and there has been healthy growth, she said.
The results were similar in New Hampshire, where SCHIP officials increased the premiums from $20 to $25 per child in January 2003 for families between 185% and 249% of poverty. For families between 250% and 300% of poverty, the premium was increased from $40 to $45.
But in Kentucky, which instituted a premium for the first time, the decline in caseload was more dramatic. Officials there initiated a $20 premium for families between 151% and 200% of poverty in December 2003. Six months before the change, the total caseload growth rate was −0.2%. Six months after the new premium was instituted, the growth rate fell to −17.4%. The premium increases there also had a stronger disenrollment effect than in the other two states.
These findings add to a growing body of evidence that increased premiums appear to reduce enrollment and increase disenrollment, Ms. Kenney said, though the impact is different among subgroups.
The largest effects occur when new premiums are imposed, especially on lower-income beneficiaries. The availability and cost of employer-sponsored insurance and other public premium policies, such as sanction policies for nonpayment of premiums, may also play a role, she said.