Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
3 Public and Private Insurance Coverage Immunization is financed through private health insurance, public safety net programs, and patient out-of-pocket spending. The source of coverage varies substantially by age (see Table 3-1~. Private insurance covers 52 percent of children aged 0 to 5 years for immunizations, and public programs cover about 34 percent. The remaining 14 percent of chil- dren are underinsured; as noted in Chapter 1, this population is defined here as those who have insurance that does not cover immunizations (see Figure 3-1~. Children who have no insurance (public or private) are auto- matically covered by the Vaccines for Children (VFC) program; children who are underinsured can receive VFC-purchased vaccines only in feder- ally qualified health centers and in their doctors' offices in some states that have expanded the VFC program with state dollars. Americans aged 65 or older are almost universally covered for immu- nizations through Medicare. Adults aged 18-64 are covered less fre- quently for immunizations than either children or the elderly (see Figure 3-2~. Private insurance covers about 41 percent of this population for im- munizations, and public programs cover about 9 percent. The remaining 50 percent are either underinsured or uninsured. Unlike children, adults who are uninsured are not covered by any public programs until they become eligible for Medicare. While older adults have Medicare coverage, adults under age 65 have virtually no safety net coverage for immunization. Only 17 percent of iCoverage can refer to either the rate of immunization in a population or insurance enroll- ment. Throughout this report, the term is used exclusively to mean insurance enrollment. 63
64 FINANCING VACCINES IN THE 21ST CENTURY TABLE 3-1 Insurance Coverage for Immunization by Age Group, 2000 Category Number Covered % of Population (in thousands) Category U.S. Population 281,400 Children 0-5 Medicaid-enrolled 4,649 18.3 SCHIP Medicaid expansion 241 0.9 SCHIP Stand-alone 771 3.0 Native American VFC-eligible 261 1.0 Private insured immunization covered 13,143 51.8 Underinsured for immunization 3,494 13.8 Medicare-enrolled (disabled-ESRDa) 194 0.8 Uninsured 2,619 10.3 Subtotal 25,372 100.0 Children 0-17 Medicaid-enrolled 12,058 15.9 SCHIP Medicaid expansion 690 0.9 SCHIP Stand-alone 1,961 2.5 Native American VFC-eligible 261 0.3 Private insured immunization covered 42,113 54.6 Underinsured for immunization 11,195 14.5 Medicare-enrolled (disabled-ESRDa) 517 0.7 Uninsured 8,406 10.9 Subtotal 77,201 100.0 Adolescents and Adults 18-64 Medicaid-enrolled 10,582 6.0 Private insured immunization covered 72,050 40.9 Underinsured for immunization not high risk 38,270 21.7 Underinsured for immunization high risk 20,680 11.7 Medicare-enrolled (disabled-ESRDa) 4,778 2.7 Uninsured not high risk 21,805 12.3 Uninsured high risk 8,229 4.7 Subtotal 176,394 100.0 Adults Aged 65+ Medicaid-enrolled 3,293 5.9 Private insured 20,761 30.6 Medicare-enrolled 31,733 57.0 Uninsured 245 0.4 Subtotal 56,032 100.0 aEnd-stage renal disease. NOTE: Percentage totals may not add due to rounding. SOURCES: Coverage data from U.S. Census Bureau, 2002; underinsurance data based on an analysis by Wood, 2003; calculations by the committee.
PUBLIC AND PRIVATE INSURANCE COVERAGE Underinsured 1 3 f3°/n Private Insurance 51 .8% 65 Public Coverage (VFC/Other) 34.4% FIGURE 3-1 Insurance coverage of vaccination, children aged 0-5 (2000~. SOURCES: U.S. Census Bureau, 2002; calculations by the committee based on an analysis by Wood (2003~. Underinsured 33 4°/n Private Insurance 40.9% Public Coverage 8.7% Uninsured 17.0% FIGURE 3-2 Insurance coverage of vaccination, adults aged 18-64 (2000~. SOURCES: U.S. Census Bureau, 2002; calculations by the committee based on an analysis by Wood (2003~.
66 FINANCING VACCINES IN THE 21ST CENTURY adults aged 18-64 with chronic health conditions and therefore at high risk for disease had ever received the vaccine in 1999; only 31.9 percent of high-risk adults received an influenza vaccination (National Center for Health Statistics [NCHS], 2000~. State-funded county and city health departments are principally re- sponsible for adult immunizations, and they are subject to capacity limi- tations and funding uncertainties. Fortunately, influenza immunizations are relatively inexpensive, and it has become common to administer them in shopping malls and employer settings. But these may not be appropri- ate settings for those who have chronic health conditions such as lung or heart disease or diabetes, especially those with limited mobility; and the paucity of coverage may limit the availability of these vaccines within the office-based practice setting. The next two sections of the chapter provide a more detailed look at public and private coverage, respectively. The final section addresses the key barriers that currently constrain the ability of the immunization sys- tem to achieve the nation's immunization goals. PUBLIC INSURANCE COVERAGE Funding for both vaccine purchase and immunization infrastructure has historically been shared by the federal government and the states (see Chapter 2~. While expenditures on vaccine purchase are increasingly de- termined by entitlement programs such as VFC and Medicaid, expendi- tures on infrastructure are largely discretionary and vary considerably from state to state. Table 3-2 summarizes the various sources for public funding of immunization. TABLE 3-2 Public Immunization Funding, Fiscal Years 1999 and 2002 (in millions of dollars) 1999 2002 Funding Source Federal State Total Federal State Total Section 317 448 NK 448 628 NK 628 VFC 467 NA 467 990 NA 990 Medicaid 70 57 127 90 69 159 Medicare 115 NA 115 255 NA 255 Total 1,100 57 1,157 1,963 69 2,032 % Change 78% 21% 76 NOTE: N/A = not applicable; NK = not known. SOURCES: FY 1999: IOM, 2000a; FY 2002: CDC, 2002e.
PUBLIC AND PRIVATE INSURANCE COVERAGE 67 Public coverage for immunization includes both safety net programs designed to provide access for disadvantaged populations and non-safety net programs, such as health benefits for veterans, military personnel, and civilian government employees. These programs are described in depth in a recent IOM report (IOM, 2000a) and are briefly reviewed below. Vaccines for Children The Vaccines for Children (VFC) program was established by a 1993 amendment to the Social Security Act as an entitlement to provide federal funds for the purchase of vaccines for children under 18 years of age in four categories: Medicaid-eligible, uninsured, Native American/Alaska Native, and children who receive vaccines at federally qualified health centers (FQHCs). VFC funds are available to underinsured children only at FQHCs. VFC works as follows. After recommending that a vaccine be added to the schedule of recommended vaccines, the Advisory Committee on Immunization Practices (ACIP) takes a separate vote for inclusion of the new vaccine in the VFC program. If the ACIP recommendation is ap- proved by the administrator of CDC, the vaccine is automatically included in the VFC entitlement and must be provided free of charge to all eligible children. CDC then negotiates a discounted federal price with the manu- facturer. CDC allocates to each state a credit balance based on the esti- mated number of recipients, which the state can use to order vaccine sup- plies from the manufacturer at the discounted federal price. States purchase vaccines and either stockpile them for distribution to registered providers or make arrangements with the manufacturer to deliver the vac- cine directly to providers. By providing free vaccines to private providers for administration to VFC-eligible children in their medical homes, VFC has resulted in a large shift of public immunization from the public to the private delivery system (IOM, 2000a) (see Chapter 2~. The creation of VFC also transferred a significant financial burden for vaccine purchase from the states to the federal government. States were no longer obligated to purchase childhood vaccines from their state-funded Medicaid budgets. Section 317 Vaccines Section 317 of the Public Health Service Act was established in 1963 to provide states with discretionary grants for vaccine purchase and infra- structure support through two types of funding: (1) direct assistance (DA) funds, which make up the majority of Section 317 funds received by im- munization programs for vaccine purchase; and (2) financial assistance (FA) funds, which typically support program infrastructure but since 1999
68 FINANCING VACCINES IN THE 21ST CENTURY have also included support for the purchase of vaccines for which there is no federal contract, such as tetanus and diphtheria vaccines and adult vaccines for hepatitis B. influenza, and pneumococcus. Over the last decade, states have experienced a disruptive Section 317 funding cycle. Until the early 1990s, infrastructure had received minimal federal support. In response to the measles outbreak of 1989-1991, Section 317 funding for infrastructure development expanded dramatically. Be- cause states were unable to ramp their programs up rapidly to utilize this funding, they experienced high levels of "carryover" funds. In response, Congress reduced infrastructure funding sharply in 1997, precisely at the time when many states were beginning to establish programs (Freed et al., 2000~. States can use only minimal amounts of VFC funding for infra- structure; therefore, state health departments sought to replace the Sec- tion 317 cuts with new state revenue allocations, with generally limited success.2 A previous IOM committee recommended an increase in federal funding of $75 million per year for infrastructure. This level was subse- quently approved by Congress and included in the fiscal year 2001 and 2002 federal budgets.3 State and County Programs States and counties provide safety net coverage for immunization through public clinics and a variety of targeted outreach programs. These programs are funded by a variety of sources state general funds, federal maternal and child health block grants to the states, public health service block grants, federal programmatic grants, private foundations, and fed- eral Section 317 grants. Direct state funding to immunization programs for vaccine purchase is highly variable (Freed and Cowan, 2002~. Several states provide no such funding. In many states that do have state funding for vaccine purchase, the funds are earmarked for adult vaccines or special programs (e.g., to support a new law requiring immunization for hepatitis B for entry to school). State legislatures typically appropriate general revenue funds for vaccine purchase, but financial support may also be provided through an ongoing mechanism to generate funds through a specific tax or surcharge or insurer contribution. State Medicaid and State Children's Health Insur- 2Twenty-one states provide direct infrastructure funding; only four states provide state funding that exceeds 40 percent of their infrastructure budgets (Freed and Cowan, 2002~. 3The Department of Health and Human Services' fiscal year 2001 budget included a $42.5 million increase for fiscal year 2001, and the fiscal year 2002 appropriation included a $32 million increase (W. Orenstein, remarks to the IOM Committee on the Evaluation of Vaccine Purchase Financing in the United States, May 21, 2002~.
PUBLIC AND PRIVATE INSURANCE COVERAGE 69 ance Program (SCHIP) funds are also provided funds for vaccine pur- chase. Medicaid reimburses providers for the purchase of vaccines for those who are Medicaid-enrolled but not VFC-eligible (e.g., over 18 years of age). Vaccine purchases for children served under stand-alone SCHIP (i.e., non-Medicaid) programs must be made with state funds; 31 states have or are starting stand-alone SCHIP programs (Freed and Cowan, 2002~. States with universal purchase and enhanced-VFC programs expand eligibility for VFC vaccines by supplementing VFC vaccine purchases at federally discounted prices. They fund these efforts by providing a com- bination of Section 317 and state general revenue funds for the purchase of vaccines for non-VFC-eligible children. Enhanced-VFC states provide free vaccines to non-VFC-eligible and underinsured children who are seen in the public sector. In addition, state funds are used to purchase vaccines for underinsured children in the offices of private providers. There are currently 15 states with enhanced-VFC programs, while 14 states have universal purchase programs in which ACIP-recommended vaccines are made available to all children, regardless of insurance status (see Table 3-3~. Universal purchase states tend to use a larger proportion of state funds relative to Section 317 funds for vaccine purchase. Some states exclude the most expensive vaccines (typically pneumococcal conjugate and vari- cella) from both enhanced-VFC and universal purchase programs (Freed and Cowan, 2002~. Medicare Medicare covers virtually all Americans aged 65 and over about 32 million in all through Medicare Part B. as well as about 5,000 children and adults who are disabled or who have end-stage renal disease (Wood, 2003~. Medicare coverage for immunization, however, is limited to influ- enza and pneumococcal vaccines, although others can be administered when shown to be medically necessary (i.e., not preventive in nature). PRIVATE INSURANCE COVERAGE The nature and scope of coverage for immunization tend to vary by the type of insurance plan. For example, health maintenance organiza- tions (HMOs), which have traditionally emphasized preventive services, frequently provide immunization as a basic covered benefit. Preferred provider organizations (PPOs) and indemnity insurance, on the other hand, have more limited immunization benefits. The first four columns of Table 3-4 show coverage levels for different types of health plans accord- ing to recent surveys.
70 FINANCING VACCINES IN THE 21ST CENTURY TABLE 3-3 State Vaccine Purchase Financing Systems VFC Only Enhanced VFC Limited Universal Purchase Universal Purchase Alabama Arkansas California Colorado Delaware District of Columbia Hawaii Indiana Iowa Louisiana Mississippi Missouri Ohio Oregon Pennsylvania Tennessee Virginia West Virginia Wisconsin Floridaa Georgia IllinoisC Kentucky Maryland Michigana Minnesota Montanaa~c~d NebraskaC,e New York Oklahomaa~c South Carolina TexasC Utaha Wyoming Connecticutb Nevadab North Carolinab North Dakotab South Dakotab Vermontb Alaska Idaho Maine Massachusetts New Hampshire New Mexico Rhode Island Washington aMoved from a VFC-only to enhanced-VFC system in 1999. bMoved from a universal purchase system to a limited universal purchase system in 1999. CPCV-7 (and varicella in Illinois) available only for VFC-eligible children. dIn process of changing back to VFC-only system. ePrivate providers can choose to receive some state-purchased vaccines for non-VFC- eligible children. SOURCE: Freed and Cowan, 2002. In addition, the relative market share of different types of insurance plans affects the rate of private immunization coverage. According to the most recent Kaiser Family Foundation-Health Education and Research Trust (KFF-HRET, 2002) survey of employers, the growth of private HMOs plateaued in 1996 at 31 percent of covered employees and had declined to 26 percent by 2002. Point-of-service plans, many of which are similar in benefit design to HMOs, plateaued in 1999 at 25 percent of cov- ered employees and had declined to 18 percent by 2002. PPOs, on the other hand, grew steadily between 2000 and 2002, and now represent 52 percent of employer-based enrollment. The last three columns of Table 3-4 indicate enrollment trends for each type of insurance plan. Given the more limited range of immunization benefits within PPOs relative to HMOs and point-of-service plans, immunization rates are likely to de- cline should these trends continue, if only because of the change in the relative market share of insurance plan types.
PUBLIC AND PRIVATE INSURANCE COVERAGE 7 TABLE 3-4 Insurance Coverage for Immunization and Employer-Based Market Share by Type of Insurance Plan Rate of Coverage Childhood Adult Market Share Plan Type 1996 1998 2001a 2001a 2000 2001a 2002 Health maintenance 97% 99% 80% 66% 29% 37% 26% organization Point-of-service plan NA 98% 78% Preferred provider 82% 86% 79% organization Indemnity insurance 68% 79% 58% 22% 20% 18% 57% 41% 50% 52% NA NA 8% NA 5% abased on data from the National Survey of Employer-Sponsored Health Plans 2001, con- ducted by Mercer Human Resource Consulting, ~c. Survey excluded indemnity option. SOURCES: Wood, 2003; KFF-HRET, 2002. The rates of private insurance coverage indicated in Figures 3-1 and 3-2 somewhat overstate the level of private coverage as a source of pay- ment for immunization because they omit patient cost sharing in the form of deductibles and copayments. These costs are difficult to estimate, but indications are that they represent a significant portion of immunization payments. For example, a recent National Immunization Survey found that 79 percent of privately insured children had some sort of payment associated with their last immunization visit (CDC, 2002f), although the median amount was only $9.96. In a study by Lieu et al. (1994b), 66 per- cent of privately insured immunization patients at public clinics stated that they were using the public clinic rather than another source because of cost. Concerns have been raised about the potential drift toward reduced benefits on the part of all types of insurance plans. The committee has no specific evidence that insurance plans have dropped immunization cov- erage, but believes that immunization benefits are likely to follow trends in other benefits, especially as the total cost of vaccination rises relative to that of other benefits. In 2002, 17 percent of workers were in companies reporting a decline in benefits from the previous year (KFF-HRET, 2002~. Between 2000 and 2002, deductibles for PPO in-plan coverage grew by 48 percent. Employer health premiums increased by 12.7 percent between 2001 and 2002, the second year of double-digit increases. There is concern that the higher expected prices of new vaccines and the increasing num-
72 FINANCING VACCINES IN THE 21ST CENTURY her of recommended vaccines will lead to a further erosion of immuniza- tion benefits and continuing increases in cost sharing (Robinson, 2002~. Patient cost sharing is increasingly used by employers as a way to shift some of the cost burden of health care to employees. But the eco- nomic rationale that has propelled cost sharing into wide use among in- surers from the 1970s is its effect on patient demand. By placing some financial responsibility for the use of medical care on consumers, cost shar- ing encourages them to limit frivolous utilization and to shop for lower- priced services. The use of cost sharing in the case of immunization may indeed encourage price shopping (as is indicated by the high numbers of insured patients seeking immunizations at public clinics because of costs [Lieu et al., 1994b]~. On the other hand, cost sharing may discourage some people from receiving immunizations at all. Thus, while there are desir- able aspects of cost sharing, in the vaccine context it is potentially coun- terproductive in that it works directly against the policy of promoting the utilization of vaccines for the public good. Defined contribution plans, although not yet a significant market pres- ence, have the potential to grow rapidly (Gabel et al., 2002~; indeed, 6 per- cent of firms report that they are "very likely" to adopt such a plan within the next 5 years (KFF-HRET, 2002~. These plans allow consumers to allo- cate dollars from a medical savings account as they wish. The growth of these plans may have a negative effect on rates of immunization, as some consumers in these plans are likely to forego such preventive benefits.4 The ability of private insurers to drop coverage or increase cost shar- ing is constrained by federal and state insurance mandates. Federal main- tenance-of-effort laws were intended to prevent crowding out of private coverage with the passage of VFC. Immunization benefit mandates have also been enacted in 28 states. In 18 of these states, immunization benefits are exempt from deductibles, and 12 states prohibit insurance plans from charging patients deductibles and copayments (American Academy of Pediatrics [AAP], 2003~. State laws, however, apply only to state-regu- lated plans; self-funded employer health plans, which represent about half of enrollees in private insurance plans,5 are exempt from state regulation under the Employee Retirement and Income Security Act (ERISA). Fur- thermore, interpretation and enforcement of state mandates vary widely (Swartz, 2003~. 4Some plans fully fund preventive benefits to address this concern. 5In 2000, 48 percent of enrollees in private employer-sponsored health plans were in plans that were self-funded and therefore exempt from state insurance regulation under the fed- eral Employee Retirement Income Security Act (Bureau of Labor Statistics, 2003~.
PUBLIC AND PRIVATE INSURANCE COVERAGE 73 In universal purchase states, health insurance plans usually provide coverage for vaccines but do not bear their costs because the state pro- vides the vaccines free of charge to providers. These plans do, however, pay provider fees associated with vaccination. Several states have ex- cluded certain expensive vaccines from their universal purchase programs because of severe budget constraints, and one state has returned to a nonuniversal program. If this should become a trend, it is not clear that private health insurers would be willing and/or able to quickly resume funding for immunizations. Given the potential for increases in vaccine costs (as discussed in Chapter 1), no one can predict with certainty how insurers will address immunization benefits in the future. However, both health plans and employers have expressed alarm about the rate of increase in vaccine expenditures and the financial pressures that could ultimately lead to the further reduction or elimination of immunization benefits (Swartz, 2003~. There are two principal concerns regarding private coverage of im- munization. First, erosion of coverage shifts the financial burden to both individuals and public payers. For some individuals, this may mean that they do not have access to immunization services. For states, this erosion adds to budgetary stresses, delays in funding, and increased fragmenta- tion of financing (issues discussed below). Second, the difficulty of accu- rately measuring the shift from private insurance coverage to the public sector makes it difficult to estimate vaccine budget requirements to assure adequate funding. BARRIERS TO A WELL-FUNCTIONING IMMUNIZATION FINANCE SYSTEM As noted in Chapter 1, the combination of public and private insur- ance coverage and vaccine provider arrangements has resulted in suc- cessfully immunizing children in remarkably high numbers. Four key fi- nance-related barriers, however, constrain the ability of the system to achieve the nation's immunization goals. These barriers gaps in cover- age and patient cost sharing, funding delays, fragmentation, and a crowd- ing out of private insurance are discussed below. Other barriers include socioeconomic status, education, public awareness, and administrative barriers (Santoli et al., 1998; Szilagyi and Rodewald, 1996~. Such factors are also reviewed, followed by an assessment of the relative importance of finance-related and other barriers and a discussion of the important issue of pockets of need.
74 FINANCING VACCINES IN THE 21ST CENTURY Gaps in Coverage and Patient Cost Sharing While the existence of gaps in insurance coverage has been estab- lished, the effect of these gaps and of patient cost sharing on rates of im- munization has yet to be demonstrated. Until recently, vaccines were rela- tively inexpensive and within the means of the poor. It is therefore reasonable to ask how much of a difference coverage and cost sharing really make in immunization rates today. If they have a significant effect, public policies to increase coverage or reduce out-of-pocket costs would clearly be warranted; otherwise they would not. A handful of recent studies address these issues (see Table 3-5), but the evidence is not conclusive. The studies fall into five categories: provider-reported data on immunization by insurance status, patient- or parent-reported data on reasons for lack of immunization, comparison group studies between free-vaccine and non-free-vaccine states, before- and-after studies in states initiating free-vaccine programs, and random- ized trials based on different cost-sharing levels. These studies are fur- ther grouped into either childhood or adult immunization. Childhood Immunization · Lurie et al. (1987) examined data from the RAND Health Insurance Experiment (HIE) and found that children and adults in plans with free preventive care (including vaccines) had 20 percent and 70 percent higher rates of immunization, respectively, than those in plans with copayment. The usefulness of this study for present purposes, however, is limited by the fact that the key dependent variable the degree of cost sharing- applies not just to immunization, but also to other types of care received during the same visit. Thus the costs associated with a provider visit may be important, but this is not necessarily the case for the costs associated with the immunization itself. · Rodewald et al. (1997) evaluated a statewide program in New York that enabled children to receive free vaccines from providers throughout the state. Controlling for age, he found that the program increased immu- nization rates by 30 percent for those with no previous insurance and by 32 percent for those with Medicaid coverage. (The study took place before the VFC program was introduced.) · In an interview study conducted by Nace et al. (1999) in Tennessee, 19 percent of physicians cited inadequate insurance as the principal rea- son for the lack of immunization among patients. A lack of insurance was cited by 16 percent of parents; 11 percent specifically cited underinsur- ance. · Taylor et al. (1997) evaluated private physicians among a geo-
PUBLIC AND PRIVATE INSURANCE COVERAGE 75 graphically dispersed sample of participants in the American Academy of Pediatrics (AAP) Pediatric Research in Office Setting network. Using a regression that controlled for education, size of metropolitan area, and child's age, he found no significant differences in immunization attribut- able to free vaccines. However, the practices evaluated, by virtue of self- selection into the AAP research network, may not be representative of practices across the United States. · Freed et al. (1999) conducted a survey to determine the impact of universal purchase in North Carolina and found that immunization had increased in all insurance categories. The largest increases were in the Medicaid (4.1-6.3 percent) and uninsured (8.6-9.2 percent) categories. The underinsured experienced increases of 4.1-6.3 percent. Freed, however, did not control for other variables that might have affected immunization rates over this period. · Recent data from CDC's National Immunization Survey suggest that, at least for expensive vaccines such as varicella, there is a substantial (12 percentage point) differential in immunization rates between insured and underinsured children (CDC, 2002g). This analysis, however, did not control for confounding variables, such as income and education, and therefore may be misleading. Adult Immunization · Using data from the 1998 National Health Interview Study, Pleis and Gentleman (2002) showed that high-risk adults aged 18 to 64 were twice as likely to get immunized for influenza if they had insurance. The regression analysis controlled for age, sex, race, poverty status, health sta- tus, region, and smoking. · A well-designed study by Ohmit et al. (1995) found that a free- vaccine intervention strongly increased rates of immunization for influ- enza relative to controls in three of four test groups, after controlling for age, gender, smoking, and high health-risk conditions. Unfortunately, the study included multiple interventions and did not distinguish the incre- mental impact of free vaccines from the effect of intensive outreach ef- forts. · Randomized trials conducted by Ives et al. (1994) in Pennsylvania and Satterthwaite (1997) in New Zealand found extremely high impacts of free vaccines: rates of immunization were 18-28 percent higher than in the control group in the former study and 66 percent higher in the latter. However, because both studies included multiple interventions, the in- cremental effects of free vaccines cannot be isolated. · Nexoe et al. (1997) conducted a randomized controlled trial among 13 general practices in Denmark. It was found that free vaccines increased
76 V) to o be 5- cry o u 5- H o H o V) cry LO ¢ EM ~ .N cn cn ·_1 CD Co 5- Lr) ~ O ~ 00 0 CO .s ~ ~ ~ > O ~ ~ ~ ·- ~ ~ ~ ~ ~ ~ _ · _,, ~ ~ _ ~ . ~1 a) 00 ON ON ON ~.0 ~ I ~ ~ 5- ~ ON U ON ~ ~ ON ON cn ~ ~ ~ ~ ~ ~ 1 ~ 'en ~ ~ ~ , ~ O ~ · ~ ~ ~ ~ ~ ~ 2 - ~ 1 ° ~ :~ C ¢ ¢ ¢ ~ ~ ~ ~ ~ LOU ~ ~ ~ ~ ~ ~ ~ ~= c _ ON oo ~ o =o ¢
77 :§ . ' o ~ ~ Do En' Cal Lr) Do ~ , ~ ~ ~= CD r :^ ~ ~ 5- ~ o ~ ~ o ~ . ~ I ~ ~ . _, ~ ~ ~ ~ ' . 2 _ ~ ~ =- ~' ·U be ~ cU cU ~ ~ U 'm ~ ~ ~ ,~ ~ ~ m ~ o I ~ ~ U U ~ ~ ~ ~ ~ ~ ~ o ~ ~ ~ ~ ~ ~ ~ · · ~ ~ ~ ~ - _ ,~ o ~ ~ o o ~ ,3N :^ En
78 En ¢ Em ~ .N En En · _1 :~ o ~ 5- En ~ 5- ~ ~ ~ ~ ~ CO :^ o ,= . ~ ~ ` $; ~ = 7~ ~ arm-,= ON I Lr) Lr) oo oo ON ON ON ON ~ N~ , it, ~~ _ ON ~ _ ¢ ¢ ~ ~ ~ ~ V
79 Do to ON 00 ~ - u At) A ·- I-d <1\ O e ~ ~ ~ . ~ ~ ~ ~ ~ 5- ON O ON I ON O he ~ O in O ~ O c, ~ D ~ ~ U ~ ~ ~ ~ g =m ~ ·N ·_1 ~ ~ ~ ~ ~ ~ U ~ ~ ~ ~ ~ ~ ~ an' ~ ~ ~ ~ ~ · - ·~ Lr O ~ Cl) :^ . O CD o . - o u ~ CD ~ .=
80 FINANCING VACCINES IN THE 21ST CENTURY immunization rates by 47 percent, but the cost of immunization for the control group was excessively high at US$40-60. Many of the above studies have shortcomings with respect to the questions relevant to research on vaccine finance. For example, the im- pact of underinsurance, a key focus of this report, is addressed only indi- rectly in most of these studies, which therefore rely upon the impact of provider out-of-pocket costs or indirect evidence. The literature also does not examine directly the question of why insurance and cost sharing mat- ter, given the relatively low price of most vaccines. However, research on private physicians' referral of patients to the public sector for immuniza- tion, which is discussed in the next chapter, may offer more insight. This literature suggests that the relationship between immunization and in- surance/cost sharing may have less to do with demand than with physi- cians' decisions to provide vaccines to broad classes of patients on the basis of their insurance coverage and physician-perceived ability to pay. Finally, few of these studies control for confounding factors, such as non- financial barriers (e.g., access to providers), discussed later in this chap- ter, that may affect immunization rates. While these studies are unsatisfying in many respects, taken together they suggest that insurance and cost factors do influence immunization rates. This is the case especially for lower-income children without other insurance and for adults who lack compulsory immunization through state school entry requirements. Funding Delays Bottlenecks in the current immunization system can result in delays in coverage for vaccines. One such bottleneck results in delays between Food and Drug Administration (FDA) approval, addition to the recom- mended schedule, and negotiation of a contract between CDC and the manufacturer. Such delays create hardships for budget-strapped states and health plans. A second bottleneck often occurs with federal and state discretionary appropriations that are required to fund vaccines for non- VFC-eligible children and adults. The result can be a two-tier system in which providers can immunize VFC children but must turn away chil- dren who do not qualify for VFC or bear the costs themselves. These bottlenecks can discourage providers from immunizing children and can impede adult immunization efforts as states shift Section 317 and state funds away from adult programs to address urgent childhood vaccine shortfalls.
PUBLIC AND PRIVATE INSURANCE COVERAGE 8 Fragmentation The complex immunization financing system results in fragmenta- tion of coverage and several associated problems. Multiple funding sources make it difficult for providers to determine the eligibility of pa- tients. This is particularly true for underinsured patients. If private pro- viders are uncertain whether a patient is covered, they are more likely to refer that patient to a public clinic. Even public clinics must worry about eligibility. For example, VFC vaccines can be used for underinsured chil- dren, but only in FQHCs. Moreover, many patients are likely to shift in and out of eligibility with significant frequency (IOM, 2000a, 2002d). Households that depend on seasonal employment or are cut from Medicaid rolls during periods of fiscal austerity are examples of the turnover that can occur in safety net programs. As an example, 40 percent of children in California lose Medic- aid coverage each year (IOM, 2000a). This situation creates financial risk for providers, who may not receive payment if eligibility is determined incorrectly. As noted above, providers may choose to refer patients rather than deal with the complexity of the system, thereby creating fragmenta- tion of care for the patient, additional burdens for the patient or parent, and possibly delayed or missed immunizations. Providers are also placed in the awkward position of providing immunizations for some but not others on the basis of insurance coverage. Another form of fragmentation relates to fees paid to providers for administering vaccines. Under VFC, vaccine purchases and administra- tion fees have two separate funding sources. The federal government pro- vides an entitlement for vaccine purchases, while administration fees are reimbursed through state-supported Medicaid payments. Depending on the status of federal negotiations and state appropriations, providers may receive vaccine reimbursement but no administration fee, fee but no vaccine reimbursement, and other possible combinations. Furthermore, provider administration fees vary widely across states. In many states, provider payments barely cover vaccine costs, resulting in increases in referrals to the public sector (Fairbrother and Haidery, 2002~. This situation in turn makes it more difficult to estimate state discretionary funding needs. Crowd-Out of Private Insurance Historically, attempts to address gaps in the immunization system have led to a crowd-out of private-sector insurance (AHSRHP, 2001; see also Chapter 2~. Crowd-out occurs when public programs displace pri- vate markets. For example, public housing partially displaces private rentals and ownership when individuals who would otherwise have
82 FINANCING VACCINES IN THE 21ST CENTURY rented or purchased in the private market obtain free or subsidized housing. Similarly, public financing of immunization can result in private providers referring patients to public immunization clinics (as discussed above) and insurance companies dropping immunization benefits. Be- cause the public safety net system is there to immunize those without access to private providers, many of these individuals will receive appro- priate immunizations at the right times. But many gaps exist in the public safety net coverage for immunization. Also, referrals from private providers increase the chances of missed opportunities, leading to longer periods of vulnerability to vaccine-preventable diseases. Other Barriers While the above analysis argues for the importance of finance-related barriers to the achievement of national immunization goals, noneconomic factors are also significant (see Calling the Shots [IOM, 2000a] for detailed discussion of these factors). Evidence for the importance of noneconomic factors includes the observation that populations with high socioeconomic status sometimes have low immunization rates, (IOM, 2000a; Orenstein et al., 1999~. For example, a study of privately insured children of parents working in a large corporation revealed that only 65 percent of the chil- dren were up to date with the 4:3:1 series at age 2 (Fielding et al., 1994~. Noneconomic factors that influence immunization rates include both per- sonal and systemic variables (Bates and Wolinsky, 1998~. Personal Variables Personal characteristics, including socioeconomic status and educa- tional attainment, have been linked to immunization rates. In a small multivariate analysis of children, the majority of whom lacked private health insurance, Bates and Wolinsky (1998) found that underimmuniza- tion at age 2 was associated with mothers who were unmarried, had mul- tiple children, did not reside with a grandparent, did not receive adequate prenatal care, and lived in poverty. Other studies have demonstrated posi- tive relationships among parents' education level, family income, and immunization rates (Ortega et al., 2000; Hughart et al., l999~. On the other hand, when other socioeconomic variables are controlled for, race is not usually an important factor (Bates and Wolinsky, 1998; Marks et al., 1979~. Patient beliefs are also important correlates of immunization rates. One study found that more than 75 percent of parents had delayed bring- ing a child in for immunization at some time because of the child's minor illness, even though the vaccination was not contraindicated (Abbotts and Osborn, 1993~. Bates and Wolinsky (1998) found a significant positive as-
PUBLIC AND PRIVATE INSURANCE COVERAGE 83 sociation between immunization at age 2 and perceived "control and re- sponsibility over life events" and an unexpected negative association with the perception of the benefits of immunization. Others have looked at the impact of incentives for immunization, such as giving toys, money, or discount coupons. However, there is little evi- dence indicating the effectiveness of such techniques (IOM, 2000a). System Variables A wide range of variables relating to the immunization delivery sys- tem has been assessed, including the immunization site, vaccine availabil- ity, and provider variables. As noted earlier, the provision of vaccinations in the medical home is a hallmark of the VFC program. The implementation of VFC, as well as similar state-level universal purchase programs, coincided with substan- tial increases in immunization rates, suggesting that the medical home may make an important difference in the rates achieved (Freed et al., 1999; Nace et al., 1999~. Research findings are mixed, however, on the contribu- tion of a routine source of care to these increases. Data from the 1988 Na- tional Health Interview Survey (NHIS) indicated that before VFC was enacted, 90 percent of children already had a routine source of care, al- though only 77 percent had been fully immunized by age 2 (St. Peter et al., 1992~. These results were affirmed by the 1993 NHIS, which revealed that 90 percent of children who were not up to date had a medical home (Tatande et al., 1996~. Likewise, a recent study found no significant asso- ciation between immunization rates and immunization within the medi- cal home (Ortega et al., 2000~. On the other hand, Bates and Wolinsky (1998) found that children are more likely to be up to date at age 2 if they have a medical home and if their provider is a private physician. States have also promoted immunizations at medical sites not tradi- tionally used for the purpose, such as hospitals, pharmacies, and nursing homes (Briss et al., 2000; IOM, 2000a). Some have promoted nonmedical sites for immunization, such as Women, Infants, and Children (WIC) pro- gram sites, schools, child care centers, stores, malls, and patients' homes. The effectiveness of these interventions has not been established through research, however. One concern regarding such strategies is that the ap- propriate medical record for immunization may not be present at the point of service. States have also attempted to improve access at existing sites by extending hours, adding staff, and providing express services at im- . . . . mumzahon c .mlcs. Provider-level variables affect immunization rates as well. Clinicians defer immunizing patients with surprising regularity because of a child's illness, a large number of shots being administered during a single visit,
84 FINANCING VACCINES IN THE 21ST CENTURY or parental concerns. As noted above, they may also refer patients to pub- lic clinics because immunization is not financially rewarding; high rates of such provider referrals, even among insured patients, are well docu- mented (Zimmerman et al., 1997) (see Chapter 4 for a full discussion of this issue). In addition, immunization involves many clinical and admin- istrative tasks, including purchasing vaccines, managing inventories, de- termining immunization status and eligibility, counseling parents, admin- istering vaccines, recording and reporting immunization, and conducting reminders/recalls. Almost all of these tasks have expanded in recent years as a result of the growing number of vaccines and the complexity of the schedule, the rising cost of vaccines, and the recent shortages of both child- hood and adult vaccines. Many providers consider the financial rewards for immunization in the form of current fees to be inadequate (AAP, 2001a). Clinicians may also lack sufficient vaccine supplies at the time of ser- vice. They may voluntarily elect to limit supplies because of the increas- ing cost of purchasing vaccines and the uncertainty of reimbursement (Freed et al., 2002a), or they may be unable to purchase adequate supplies because of supply disruptions. While there is no direct evidence of re- duced immunization rates or increased disease incidence as a result of recent vaccine shortages, there is some indirect evidence that a supply disruption may adversely affect provider immunization practices (Oram et al., 2001~. For example, CDC has reported that 52 percent of states sus- pended school immunization requirements as a result of the tetanus vac- cine shortage (Orenstein, 2002c). Providers may also lack efficient systems for vaccine administration, such as reminder/recall systems, assessment and feedback processes, standing orders (which allow nonphysicians to administer vaccines with- out direct physician supervision), and even simple checklists (Briss et al., 2000; IOM, 2000a). Evidence supporting the benefits of such systems is limited, however, and their adoption by providers has not been wide- spread (Darden et al., 1999; IOM, 2000a). Unavailability of a child's complete immunization record to the clini- cian is a critical factor in underimmunization (Stokely et al., 2001~. This situation often occurs when families move within or across states or switch providers, and it is exacerbated by the increasing complexity of the im- munization schedule and the fragmentation within the immunization fi- nancing system discussed above. There is growing evidence that the use of electronic vaccine registries can significantly improve the accuracy of immunization records (Boyd et al., 2002; Davidson et al., 2003; Stille and Christison-Lagay, 2000~; the impact of such registries on immunization rates, however, has not been clearly demonstrated. An alternative strat- egy is to have patients or parents retain "handheld" immunization records; here too, however, there is little evidence that this approach is
PUBLIC AND PRIVATE INSURANCE COVERAGE 85 effective, and it may in fact cause administrative problems (Dickey and Petitti, 1992; IOM, 2000a). The increasing availability of combination vaccines may improve im- munization rates. The current vaccine schedule requires as many as five injections during a single visit. Two or more of these are often deferred because of parental concern about the fear and pain involved in multiple injections (Glode, 2001~. There is a clear parental preference for combina- tion vaccines. A survey by Meyerhoff et al. (2001) found that it is worth a median of $8.00 to parents to avoid having their child receive an addi- tional injection. Another study found that reducing the number of shots from four to three is worth a median of $25, from three to two is worth $25, and from two to one is worth $50 (Lieu et al., 2000b). Also, reducing the required number of injections can significantly reduce administrative time (Pellissier et al., 2000~. There is, however, no direct evidence on the impact of combination vaccines on immunization rates. There are also potential drawbacks, including the presence of competing combinations with various overlapping antigen menus and subtle immunologic differ- ences that may create confusion and/or administrative burden for busy practitioners (Le. 2001~. Mandatory immunization, such as that required for school entry, ap- pears to be effective in increasing immunization rates. All states adopted such school laws during the 1970s and 1980s (IOM, 2000a), and immuni- zation rates increased dramatically. But the relative impact of school en- try laws compared with other factors has not been determined. Also, the impact of school entry requirements occurs well after most vaccinations are typically administered: 20 of the 23 recommended childhood vaccina- tions are normally completed by 18 months of age. The increasing use of day care requirements addresses this shortcoming, but such requirements have not been as widely implemented as those for school entry. Other mandates are now being widely applied for nursing homes. There is how- ever, insufficient evidence of the effectiveness of any of these requirements (IOM, 2000a). Relative Importance of Finance-Related and Other Barriers Several studies directly compare the importance of finance-related and other factors. One study found that cost was the most important rea- son given by parents for having their children immunized at a public health clinic. Of those interviewed, 63 percent had come to the public clinic because of its lower cost, and 79 percent of families interviewed rated this as an important factor in choosing a public health facility (Lieu et al., 1994b). Other factors cited include convenient location, no appointment needed, recent relocation, and other access advantages. In another study,
86 FINANCING VACCINES IN THE 21ST CENTURY lack of insurance coverage was deemed the third most important reason for incomplete immunization by age 2 (cited by 16 percent of parents), after waiting time (33 percent) and child ill at the time of appointment (21 percent) (Nace et al., 1999~. Developing a full understanding of the relative importance in immu- nization of the finance-related and personal and systemic variables dis- cussed above will require substantial additional research. No clear evi- dence has emerged that can be used to rank the relative importance of these factors. It is clear, however, that both sets of factors are important and require attention if full immunization is to be achieved. If finance- related barriers the primary focus of this report are not addressed, strategies that address noneconomic barriers will not be fully effective. Pockets of Need Uninsured children fare well under the current system: they are cat- egorically covered under VFC and can receive free vaccines in their medical home or from any VFC provider or clinic. Ironically, the biggest coverage gap for children is among those who have private insurance coverage- 21 percent of insured children aged 0-5 lack coverage for vaccines (Wood, 2003~. VFC statutorily excludes these children because they are "insured." Fourteen states address this gap by providing state-funded coverage for the underinsured. In these states, children can receive free vaccines through their own provider, regardless of their insurance coverage. But parents of children in the other 36 states must either pay out of pocket or take their children to a public health clinic to obtain free vaccine. The same problem is experienced by children who are enrolled in stand-alone SCHIP programs. While Medicaid-eligible children are cat- egorically qualified to receive free VFC vaccines, children in stand-alone SCHIP states, such as California, are technically "insured," and the Cen- ters for Medicare and Medicaid Services (CMS) has ruled that they are excluded from VFC coverage. These children still receive vaccines from their SCHIP provider, but without VFC funding. Estimating the Number of Underinsured Estimates of the number of underinsured children vary considerably and are not deemed highly reliable. CDC is considering ways to improve estimates of underinsurance. The committee considered several estimates, which were summarized earlier in Table 3-4. A recent Partnership for Prevention study used by Wood (2003) esti- mates that 21 percent of privately insured children aged 0-5 have private insurance that excludes immunization. This suggests that 3.5 million chil-
PUBLIC AND PRIVATE INSURANCE COVERAGE 87 dren aged 0-5 are underinsured 13.8 percent of this age group. Wood points out that some of these underinsured are covered through state uni- versal purchase programs in 14 states. On the other hand, this estimate excludes the children who are covered by insurance for immunizations but face prohibitive copayments and deductibles. Preliminary first-quarter results from the 2002 National Immuniza- tion Survey indicate that 7.3 percent of children (aged 0-17) have some kind of insurance (either public or private) that excludes immunization (CDC, 2002d). Meyer and Waldman (2002), using data from KPMG and Health Research and Educational Trust (HRET) surveys, estimates that about 8 percent of children are underinsured for immunizations. The IOM (2000a) study cited earlier also uses KPMG data to derive an estimate of 5 percent for the underinsured population aged 0-17. Studies examining the insurance status of individuals using public clinics support the above estimates of the incidence of underinsurance. Lieu et al. (1994b) found that about 16 percent of immunization patients at public clinics were underinsured. A more recent analysis of data from the 2002 National Immunization Survey found that 44-63 percent of children vaccinated in public health clinics in South Carolina and California were underinsured (CDC, 2002i). Immunization Insurance Benefits for Adults The coverage picture for adults is far less positive than that for chil- dren. Like children, adults face the problem of underinsurance. Accord- ing to the Partnership for Prevention survey cited above, 59 million adults aged 18-64 have private insurance that does not include immunization benefits (Wood, 2003~. In addition, more than 30 million adults under age 65 are uninsured. Unlike uninsured children, who are categorically cov- ered by VFC, uninsured adults have no safety net immunization program. Thus, a total of 89 million adults under 65 50 percent of this age group- lack coverage for immunizations. Older persons generally require fewer routine vaccinations than chil- dren; but adults at high risk for vaccine-preventable diseases may need to be immunized against pneumonia, influenza, meningitis, and hepatitis. (Table 3-6 shows the population at high risk for severe influenza or pneu- mococcal disease within various age groups.) More than 65,000 deaths from influenza and pneumonia occur annually, most among older adults (CDC, 2002b).6 In contrast, about 300 deaths occur each year as a result of 6Note, however that the majority of these deaths are among those aged 65 and older. Some portion of these deaths are likely due to family decisions not to immunize because of extreme age or frail health.
88 FINANCING VACCINES IN THE 21ST CENTURY TABLE 3-6 Adults Considered to Be at High Risk for Influenza or Pneumococcal Disease Annual Number of Population High Risk Percent Population at High in Need of of Age Population at High Risk and Publicly Age Group at at High Risk and Under- Purchased group Population High Risk Risk Uninsured insured Vaccine 18~9 136 million 11% 15 million 2.7 million 5.0 million 7.7 million 50-64 43 million 1oo%a Total 180 million 32% adults 18-64 43 million 5.5 million 58 million 8.2 million 16.0 million 21.5 million 21.0 million 29.2 million al995 recommendation is for 100% of this group to receive influenza annually and pneumococcal vaccine one bme. SOURCE: Adapted from Wood, 2003. all vaccine-preventable diseases among children. In addition, 48,000 pneu- monia and influenza hospitalizations occur annually among adults over age 65 (CDC, 2001a). The monetary burden of adult vaccine-preventable diseases is estimated to be greater than $10 billion per year (CDC, 2002c). Although Medicare plans provide coverage for older adults for pneu- mococcal immunization (since 1981) and for influenza immunization (since 1993), immunization levels among older adults remain low. In 1997, 66 percent of this population received an influenza immunization, and a cumulative 50 percent had received a pneumococcal immunization still well below the national goal of 90 percent established for each in Healthy People 2010 (U.S. Department of Health and Human Services, 2000~. For- tunately, many adults have access to influenza immunization sponsored by employers, retail stores, and shopping malls. This may partly explain the significant gain in immunization rates among this population in the last several years. ACIP recommends pneumococcal vaccination for all adults over age 50, adults between age 18 and 64 at high risk e.g., those with chronic diseases affecting the lungs, heart, immune system, and selected other organ systems and all adults aged 65 and older (CDC, 2003f). Annual influenza vaccination is recommended for high-risk adults (aged 18-49) and for all adults over age 50. According to Wood, 58 million Americans,
PUBLIC AND PRIVATE INSURANCE COVERAGE 89 or 32 percent of adults under age 65, are considered to be at high risk for vaccine-preventable diseases (Wood, 2003~. Half of these, about 29 mil- lion adults, have no coverage for immunizations. FINDINGS · An estimated 13.8 percent of children between birth and 5 years of age are underinsured (that is, have private insurance that does not in- clude immunization benefits). · Half of all adults aged 18-64 lack immunization coverage; 32 per- cent of this population (29 million adults) is considered to be at high risk. · The proportion of children and adults without immunization cov- erage may increase as a result of current trends in insurance benefits and the increasing cost of the recommended vaccines on the immunization schedule. · Insurance coverage and patient cost sharing are among the impor- tant factors influencing rates of immunization. · The current vaccine financing system is fragmented and prone to funding delays; the result is missed opportunities, institutional barriers to . . . mmun~zahon. · Public vaccine financing programs have led to some crowd-out of private immunization coverage, and attempts to limit crowd-out have met with mixed success. · Increasing vaccine costs, crowd-out of private-sector financing, and federal funding lags place significant stress on state financing mecha- nisms, prompting limits on state contributions to immunization programs.