|
MARKETWATCHCoverage Of Vaccines In Private Health Plans: What Does The Public Prefer?
Underinsurance for recommended vaccines in private health plans may affect 15 percent of children and more than 30 percent of adults. We conducted a nationally representative, Web-based study using health plan vignettes to determine adults’ willingness to bear marginally higher plan premiums, to assure plan coverage of new vaccines as they are recommended for children and adults. Our results indicate a broad willingness (more than 75 percent of respondents) to pay the higher premiums. Such willingness was associated strongly with perceptions of vaccines’ effectiveness and safety. Policymakers, physicians, and public health officials should examine health plan enrollees’ preferences as they consider remedies for vaccine underinsurance.
Immunization rates for the primary series of vaccines among children and for the widely recommended influenza and pneumococcus vaccines among adults fall short of targets set in Healthy People 2010.1 Lack of insurance and the resulting economic barriers have been blamed in part for these low immunization rates. In addition, immunization officials and policy experts have recently turned their attention to the problem of underinsurance for vaccines. The most recent estimates indicate that as many as 15 percent of children and more than 30 percent of all adults are enrolled in health plans that do not pay for recommended vaccines.2 State mandates. The principal remedy for underinsurance to this point has been government statutes that mandate coverage of recommended vaccines. More than half of the states have such mandates for children’s vaccines, but none of the states have mandates for adults’ vaccines.3 States’ legislative mandates are not enforceable for self-insured health plans under the federal Employee Retirement Income Security Act (ERISA) statute; although the number of children enrolled in self-insured plans is not known, as many as half of employees are enrolled in such plans nationally.4 Previous studies have indicated that enrollees in self-insured plans are more likely than those in other types of plans to be underinsured for childhood and adult vaccines.5 Other remedies. To address vaccine underinsurance in general, and ERISA exemptions for self-insured plans in particular, an Institute of Medicine (IOM) expert committee recently recommended a federal mandate for all recommended vaccines that would override ERISA, supplemented by a subsidy to employers to cover the costs of the mandate.6 Such a mandate, however, presents many logistical and political challenges, and the National Vaccine Advisory Committee (NVAC) has advised against a mandate- and subsidy-based overhaul of the current system.7 Lost in the debate regarding underinsurance are the perspectives of U.S. children and adults for whom vaccines are recommended but whose employer-sponsored plans leave them underinsured. Although we found in a previous study that preventive services are broadly favored by people who are given the opportunity to design their own hypothetical health plans, we are not aware of any studies that have specifically examined preferences for vaccine coverage in health plans.8
Hypotheses. Our principal hypothesis for this study was that employer-sponsored health plan enrollees would be willing to bear higher monthly health plan premiums in return for coverage for vaccines recommended by the Advisory Committee on Immunization Practices (ACIP) of the U.S. Centers for Disease Control and Prevention (CDC). We presented the central research question as willingness to bear a higher premium rather than willingness to pay out of pocket for vaccines, for three reasons: (1) The premise of underinsurance as a barrier is that insured people without coverage for certain services are not willing to pay for those same services out of pocket; (2) paying a higher premium for broader coverage is a familiar choice faced by employees; and (3) willingness-to-pay studies regarding vaccines are often limited by challenges in anchoring respondents’ answers because they are not familiar with vaccines’ retail costs.9 We also examined two secondary hypotheses: (1) that providing information about the potential health and economic benefits of vaccines would increase the proportion of enrollees who opted for more comprehensive coverage of vaccines; and (2) that providing additional information about the potential benefits of individual immunization to others (the "public good" benefit) would further increase the proportion of enrollees who opted for comprehensive coverage of vaccines. Study design and sample. We conducted a randomized study to investigate the effects on health plan choices of information about the potential health and economic benefits of newly recommended adult and childhood vaccines. The sample was drawn from an online research panel established and maintained by Knowledge Networks.10 The panel is representative of the entire U.S. population and is recruited using probability sampling techniques that are not limited to current Web users or computer owners, through provision of WebTV units and Internet service to homes without computers. All new panel members are profiled at the time of system installation, to allow for stratified sampling based on demographic characteristics of the individual and the household. To adjust for sources of bias such as lack of Internet services, lack of telephone in the household, nonresponse, and panel attrition, Knowledge Networks applies poststratification adjustment to Current Population Survey (CPS) demographic and geographic benchmarks, to create panel weights that permit generalization of study findings to the U.S. population. Panel studies have revealed that people in the Knowledge Networks sample respond to questions via the Web as they would to questions via telephone, and that the prevalence of health conditions in the sample is virtually indistinguishable from that in national health surveys.11 For the present study, people age eighteen and older in the Knowledge Networks panel were eligible. Separate samples were selected from households with a minor present (under age eighteen) and without a minor present, in equal proportions, to achieve a target of 500 completed cases in each stratum. Selection of health plans. Each respondent was instructed to consider a vignette in which she or he was employed and was offered a health plan as a benefit. Respondents were advised that these plans would be the only source of health insurance for them and their families. Based on concurrent national data about mean employees’ share of monthly premiums, respondents were informed that an individual plan premium would cost them $45 per month and that a family plan premium would cost them $179 per month.12 Respondents were then asked if they would opt for coverage (individual versus family) or for no coverage. Respondents who selected no coverage were asked basic demographic questions and a sequence of questions about vaccine attitudes and then exited the study. Respondents who opted for coverage were provided with basic information about vaccines, in which they were advised that several new vaccines (meningococcal conjugate, hepatitis C, H. pylori, and human papillomavirus) may be approved sometime in the next one to five years.13 These vaccines were selected because they are designed to prevent potentially serious or fatal illnesses and, if selected for recommendation, together will be recommended across a broad range of ages. Respondents were then asked to indicate their preference among health plans that varied only with respect to vaccine coverage. Those who opted for individual coverage were offered a choice of Plan A, which included coverage for existing childhood and adult vaccines ("basic" vaccine coverage) at $45 per month, versus Plan B, which also would include coverage for any newly recommended vaccines ("comprehensive" vaccine coverage) at $3 more per month.14 The additional premium would remain in effect for the plan year, regardless of the number of new vaccine recommendations (whether none or more than one), because premiums are not adjusted midyear. Respondents who opted for family coverage were offered a choice of Plan A, with basic vaccine coverage at $179 per month, versus Plan B, which would include coverage for any new childhood vaccines ("comprehensive child" vaccine coverage) at $3 more per month, versus Plan C, which would include coverage for any new childhood or adult vaccines ("comprehensive child plus adult" vaccine coverage) at $6 more per month. Sample randomization. After respondents had selected a plan, they were block-randomized (by stratum) to receive one of two stimuli consisting of additional information about the potential health and economic benefits of vaccines. Respondents randomized to the "private gain" stimulus were advised that new vaccines might cost $75 or more for each set of doses out of pocket without insurance coverage and that vaccines could reduce medical costs by reducing the need for antibiotics and physician office or emergency visits.15 Respondents also were informed that vaccines may cause some side effects and may or may not be recommended for the respondents and their child(ren), depending on the vaccine. Respondents randomized to the "public good" stimulus received all of the "private gain" information. They also were advised that vaccines might protect people who do not receive the vaccine themselves and that even if vaccines were not recommended for respondents or their child(ren), those vaccines were likely to benefit other members of their communities. Following these stimuli, respondents were again asked to indicate their choice of health plans. Vaccine attitudes. All respondents were asked to indicate to what extent they agreed with statements about the effectiveness of childhood and adult vaccines and the safety of vaccines in general. Respondents who opted for no plan coverage or for basic rather than comprehensive coverage were asked to indicate how much they would pay for a new vaccine recommended by their (or their child’s) physician. Respondent characteristics. In addition to standard sociodemographic characteristics, all respondents were asked to provide information about current and prior health insurance status, as well as their current self-reported health status and the health status of their youngest child. Data analysis. Chi-square tests were used for comparisons of categorical responses. All analyses were conducted using sampling weights provided by Knowledge Networks and are reported as such; sample sizes are reported unweighted. We used Stata version 8.0 SE (Stata Corp., College Station, Texas) for all analyses. Item nonresponse was less than 2 percent for all questions; item nonrespondents were excluded from analyses on an item-by-item basis.
Sample characteristics. The study response rate was 77 percent, for an analytic sample of 995 respondents. Respondents in households without minor children differed sociodemographically in some respects from those in households with minor children (Exhibit 1  ). Respondents without minor children in the household were significantly older, more likely to identify themselves as non-Hispanic white, and less likely to be married. Those with no minors in the household were also more likely to report being in fair or poor health.
Selection of coverage. Among all respondents, 6 percent chose no health insurance (8.7 percent among households with children versus 4 percent in households without children; p <.01). Across all households, respondents who reported annual household income of less than $20,000 were significantly more likely to forgo coverage (15.4 percent) than respondents in all other income groups (p < .0001). Uninsured respondents were more likely to forgo coverage in the vignette (14.5 percent) than respondents with either private coverage (2.6 percent) or public coverage (9.8 percent; p < .0001). Among respondents without minor children, 27 percent chose individual coverage and 69 percent chose family coverage (individual plus nonchild dependents). Among respondents with minors in the household, 11.3 percent chose individual coverage and 80 percent chose family coverage.
Preferences for comprehensive child vaccine coverage.
Among respondents with minors in the household who chose family coverage, the majority (77.5 percent) indicated that they would prefer plans with comprehensive coverage for newly recommended childhood vaccines, alone or in combination with comprehensive adult coverage (Exhibit 2
The preference for plans with comprehensive vaccine coverage strengthened slightly (78.9 percent) after study participants were given additional information about the potential health and economic benefits. Approximately one of every five respondents who initially showed no preference for comprehensive coverage opted for comprehensive child and adult vaccine coverage after receiving additional information, and about one of every three respondents who initially preferred just comprehensive child coverage opted for comprehensive adult coverage as well. Interestingly, whether respondents received the "private gain" versus "public good" information did not affect their preferences for vaccine coverage. Further, final selection of comprehensive versus basic vaccine coverage by respondents in households with minors was not associated with respondents’ sociodemographic characteristics or with the age or health status of their youngest child. However, respondents who chose comprehensive rather than basic coverage were significantly more likely to strongly agree with statements that child vaccines are effective (49.3 percent versus 29.3 percent), that adult vaccines are effective (38.0 percent versus 18.7 percent), and that vaccines are generally safe (24.7 percent versus 6.2 percent; all p < .005). Preferences for comprehensive adult vaccine coverage. Among respondents who chose individual coverage or lived in households without children and chose family coverage for nonchild dependents, 78.6 percent indicated a preference for comprehensive versus basic coverage for adult vaccines in their vignette health plan at baseline. This preference was not affected by the provision of information about potential health and economic effects of vaccines. Respondents’ final preferences for comprehensive versus basic adult vaccine coverage were significantly associated with only one sociodemographic factor: current insurance status. Respondents currently on public insurance were significantly more likely to opt for comprehensive coverage (86.5 percent) than were those who had private insurance (77.4 percent) or no insurance coverage (76.7 percent; p < .0005). In contrast to findings for childhood vaccines, preferences for comprehensive adult vaccine coverage were not associated with attitudes about vaccine effectiveness or safety.
Paying out of pocket for newly recommended vaccines.
Respondents who opted either for no insurance coverage or only basic vaccine coverage were asked how much they would be willing to pay out of pocket for a new vaccine series recommended by their (or their child’s) physician (Exhibit 3
Those who would not pay for newly recommended child vaccines were significantly less likely than those who would pay to agree or strongly agree that children’s vaccines are effective (63.8 percent versus 83.7 percent; p < .05), that adult vaccines are effective (48.2 percent versus 70.6 percent; p < .005), and that vaccines are generally safe (37.1 percent versus 63.7 percent; p < .05). No sociodemographic characteristics were associated with willingness to pay for respondents in this subsample. For adult vaccines, nearly half of the sub-sample said that they would not pay for a newly recommended vaccine, while approximately one in six said that they would pay more than $50. Those who were unwilling to pay for a new adult vaccine had significant differences in attitudes about vaccines’ effectiveness and safety compared with those who were willing to pay, similar to those reported above for new childhood vaccines (data not shown). In addition, those unwilling to pay for a new adult vaccine were significantly more likely to report annual income of less than $40,000 compared with those who were willing to pay (73.8 percent versus 50.6 percent; p < .005, data not shown).
Public’s preferences. This is the first study to indicate broad public preferences for coverage of all recommended childhood and adult vaccines in employer-sponsored health plans. Almost four of every five respondents indicated that they would be willing to bear a marginally higher premium cost each month in exchange for guaranteed coverage of newly recommended vaccines. This finding has implications for the current national debate about how best to address the problem of underinsurance for vaccines among children and adults. Until now, policymakers have considered legislative coverage mandates that are administratively challenging.16 Although public programs such as the federal Vaccines for Children (VFC) program, Section 317 grants, and state funds dedicated to immunization purchase include some provisions for underinsured people, their effects are limited by programmatic challenges related to provider participation and restrictions on where such publicly purchased vaccine may be administered.17 Our findings suggest that a more direct alternative may be to allow employees to express their vaccine coverage preferences to employers and insurers so that health plan benefits can be modified and employees’ premiums adjusted accordingly. Whether employers will share in this additional premium cost appears immaterial to employees’ preferences. We intentionally set the additional premium at a level that would represent no cost sharing by employers and found that the majority of employees would be willing to bear the additional premium in exchange for comprehensive vaccine coverage. The growing appeal of consumer-directed health plans and high-deductible health plans among employers as a means of curbing increases in health care costs raises some questions about how our findings might be applied in such plans.18 One of the key reasons that comprehensive vaccine coverage could be added to a lower-deductible plan at only a marginally higher premium is that the cost of the vaccine is borne by all of the plan enrollees, whereas only a subgroup is typically eligible to receive the vaccine (for reasons of age, sex, and so forth). If cost sharing is diminished through high-deductible plans, the preferences for comprehensive vaccine coverage may be reduced when employees are asked to pay full price for vaccines as part of the deductible. Our findings suggest two responses to this concern. First, we found that more than one-quarter of parents who opted for basic vaccine coverage said that they would pay more than $50 for a newly recommended vaccine series for their children; parents who would want comprehensive vaccine coverage might be willing to pay even more. Of note, out-of-pocket enthusiasm for newly recommended adult vaccines was more attenuated, consistent with lower vaccination rates among adults than children nationally. Second, some employers and insurers create incentives for the use of preventive services by exempting such services from the plan deductible——that is, providing "first-dollar" coverage for preventive care. Such exemptions preserve cost sharing for preventive benefits in employer-sponsored health plans and are strongly encouraged by national immunization authorities.19 Impact of consumer information. In this study we also explored the potential effects of information provided to employees about vaccines in their choice of health plan. We did so because we were skeptical about whether employees have such information in hand when they make health plan decisions. Nonetheless, providing such information had no significant effect on respondents’ plan choices. One possible explanation is that respondents have already formed their opinions about vaccines and were not swayed by additional information. This argument is supported by the finding that respondents’ general attitudes about the effectiveness and safety of vaccines were strongly associated with their choice of comprehensive vaccine coverage for their children. These general attitudes were likely formed prior to the study. Another possible explanation is that the information we provided about the potential health and economic benefits of newly recommended vaccines was not persuasive. We found it challenging to write concise information that would be accurate across a broad variety of vaccine products, partly because of differences in the products and their target populations but also because of a lack of information about the short-term benefits of new vaccines prior to their release. The latter information deficit is a potential problem for all new vaccines, which threatens their timely adoption as covered benefits in health plans following the release of new national recommendations. Interestingly, we found no effect of information about the "public good" aspect of vaccinations versus the "private gain" aspect. This may have occurred because, although vaccines are commonly regarded as the prototypical public good with positive effects even for members of the community who do not receive them, we examined preferences for coverage within private health plans. This underscores a tension in the U.S. approach to vaccine financing, in which private and public markets function side by side but not necessarily in concert with respect to public health goals such as preserving vaccine supply.20 Do we expect families to behave with the public interest in mind if they are being asked to purchase vaccines with their own money? If there is a compelling public interest in providing vaccines, should the government assume responsibility for all vaccine purchases? These issues remain to be resolved at the national level. Of note, for adult vaccine coverage in this study, people who had public health insurance were more likely to opt for comprehensive vaccine coverage than were those with private insurance. Perhaps publicly insured adults recognize the benefits of public provision of vaccine that privately insured adults do not. Study limitations. The chief limitation in this study is its simulation methodology, a limitation shared with willingness-to-pay methods used widely in analyses of economic behavior. We cannot know with certainty how study respondents would behave "in real life" if they were presented with the plan choices we designed. However, we are reassured by strong indications that their responses were internally consistent. For example, the rate of plan uptake was slightly higher than national averages, consistent with our stipulation to respondents that this would be their only source of health insurance.21 In addition, the strong association of childhood vaccine coverage preferences with respondents’ attitudes about vaccines’ effectiveness and safety suggests that the study allowed them to express preferences consistent with their beliefs. Similarly, among respondents who opted for plans with basic vaccine coverage, their willingness to pay for a newly recommended vaccine out of pocket was also consistent with their attitudes. Another limitation is that we set the incremental premium associated with comprehensive vaccine coverage at a fixed level but did not specify what respondents’ wages would be in the simulation. However, we found an income effect on choice of insurance coverage—that is, those at lower incomes opted out of coverage at higher rates. We believe that this finding suggests that respondents were answering with their own wage expectations in mind and indicates that this approach is appropriate for developing national estimates of preferences regarding vaccines. An additional potential limitation is that we asked respondents to consider themselves employees for the purposes of the study, but not all respondents were currently employed. However, we found no association of current work status with vaccine preferences. Our findings are also limited by the fact that we did not ascertain respondents’ underinsurance status for child or adult vaccines. Our experience is that many people are unaware that they are underinsured until a physician recommends a vaccine, and therefore we doubted the validity of responses to such a question. Therefore, instead we measured the appeal of comprehensive vaccine coverage regarding potential new vaccine recommendations, for which no respondents’ actual health plans yet included coverage. As policymakers, physicians, and public health officials consider remedies for adult and child vaccine underinsurance, they might not need to look further than plan enrollees. Enrollees’ preferences, if solicited, could support greater breadth in vaccination benefits offered in employer-sponsored plans; these need not be customization options at the point of purchase but, rather, features of different plans offered to employees. Furthermore, this study may shed new light on opportunities to structure plan benefits around consumers’ preferences, particularly regarding benefits such as vaccines that have broad health and economic implications for the population.
Matthew Davis (mattdav{at}med.umich.edu) is an assistant professor of pediatrics, internal medicine, and public Policy in the Child Health Evaluation and Research (CHEAR) Unit, Divisions of General Pediatrics and General Internal Medicine and the Gerald R. Ford School of Public Policy, at the University of Michigan in Ann Arbor. Kathryn Fant is a research associate in the CHEAR Unit, Division of General Pediatrics. The authors gratefully acknowledge the support of the Robert Wood Johnson Foundation Generalist Physician Faculty Scholars Program, Grant no. 045439, Matthew Davis, principal investigator. The authors also acknowledge data collection by Time-Sharing Experiments for the Social Sciences, NSF Grant no. 0094964, Diana Mutz and Arthur Lupia, principal investigators.
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Study Protocols
Study Results
65 Years," Morbidity and Mortality Weekly Report 51, no. 45 (2002): 1019–1024; and "Preventive-Care Practices among Persons with Diabetes—United States, 1995 and 2001," Morbidity and Mortality Weekly Report 51, no. 43 (2002): 965–969.
