Preference-based selection effects in elders with long-term care insurance (poster)

Hyun J, Garber AM, Calvillo T, Mahlow P, Nelson L, Goldstein MK. Preference-based selection effects in elders with long-term care insurance (poster). Paper presented at: AcademyHealth Annual Research Meeting; 2005 Jun 26; Boston, MA.

Research Objective: The asymmetric information modelpredicts a positive correlation between insurance and theoccurrence of risky events due to moral hazard or adverseselection. Adverse selection occurs because those most likelyto purchase insurance are most likely to file a claim; they maybe at a higher health risk of a health event or have strongercare preferences. Underwriting practices attempt to correctfor selection based on observable risk factors. Insurers,however, are not privy to information about preferences forcare and risk aversion. We examined elders in a Medicaremanaged care program comparing individuals with andwithout long-term care insurance (LTCI) on health behaviorand differences in their preferences for nursing home care.Study Design: The study uses cross-sectional data from alongitudinal study investigating the impact of ADLdependency on health utility ratings.Population Studied: This study used a convenience sample(n = 292) of Medicare managed care enrollees 65-years old andolder, of whom 82 had LTCI. Those with and without LTCI didnot differ significantly in terms of age, number of livingchildren, ADL and IADL limitations, and total householdincome. Those with LTCI were more likely to be White andhad more years of education than those without LTCI.Principal Findings: Elders with LTCI spent less timedepressed or sad and were more likely to engage in preventivehealth behaviors, such as getting a flu shot, within the last 12months. Ratings of general health, health satisfaction, ADLand IADL limitations, and chronic disease did not differsignificantly between LTCI and non-LTCI groups. Elders withLTCI had higher utility ratings for nursing home care than forcare in their own homes given hypothetical ADLdependencies, controlling for current health utility scores. Onaverage, those with LTCI rated their preference for nursinghome care 0.12 points (on a 0 to 1 scale) higher than thosewithout LTCI.Conclusions: Elders with LTCI spent less time depressed,engaged in more preventive behaviors, and showed a trendtoward rating their health better than their peers. This studysuggests that individuals with LTCI may have more awarenessof future health events--a concern that is manifested inpreventive health behaviors. Elders with LTCI are much lessaverse to nursing home care than those without LTCI, apreference that may increase the likelihood of nursing homeutilization. The study design did not allow us to determinewhether the preference differences pre-dated the decision tobuy LTCI.Implications for Policy, Delivery, or Practice: Findingssuggest that those who seek LTCI have risk-aversioncharacteristics, manifested in health-seeking behaviors, whichbalance preference-based selection effects for nursing homecare. Although these findings are consistent with asymmetricinformation, they do not suggest an increased probability of ahealth event. LTCI providers should assess the impact ofmanaged care programs to minimize moral hazard problemspost-insurance purchase and should be more attuned topotential advantageous selection characteristics.Primary Funding Source: NIA, Veterans Affairs HSRandD