Cost-Effectiveness Analysis of an Intranasal Influenza Vaccine for the Prevention of Influenza in Healthy Children
Objective. Intranasal influenza vaccine has proven clinical efficacy and may be better tolerated by young children and their families than an injectable vaccine. This study determined the potential cost-effectiveness (CE) of an intranasal influenza vaccine among healthy children.
Methods. We conducted a CE analysis of data collected between 1996 and 1998 during a prospective 2-year efficacy trial of intranasal influenza vaccine, supplemented with data from the literature. The CE analysis included both direct and indirect costs. We enrolled 1602 healthy children aged 15 to 71 months in year 1, 1358 of whom were enrolled in year 2. One or 2 doses of intranasal influenza vaccine or placebo were administered to measure the cost per febrile influenza-like illness (ILI) day avoided.
Results. During the 2-year study period, vaccinated children had an average of 1.2 fewer ILI fever days/child than unvaccinated children. In an individual-based vaccine delivery scenario with vaccine given twice in the first year and once each year thereafter at an assumed base case total cost of $20 for the vaccine and its administration (ie, per dose), CE was approximately $30/febrile ILI day avoided. CE ranged from $10 to $69/febrile ILI day avoided at $10 to $40/dose, respectively. In a group-based delivery scenario, vaccination was cost saving compared with placebo and remained so if vaccine cost was <$28 (the break-even price per dose). In the individual-based scenario, vaccination was cost saving if vaccine cost was <$5. In this scenario, nearly half of lost productivity in the vaccine group was attributable to vaccine visits, which overshadowed the relatively modest savings in ILI-associated costs averted.
Conclusions. Routine use of intranasal influenza vaccine among healthy children may be cost-effective and may be maximized by using group-based vaccination approaches. cost-effectiveness, influenza, vaccine, children.
Although influenza vaccine is recommended for children with high-risk chronic health conditions, influenza is a major cause of respiratory disease in healthy young children, with annual rates of infection of 35% to 50% per year. Approximately 10% of such children have lower respiratory tract disease; children who are younger than 5 years are at greatest risk.1–4 During annual influenza epidemics, 5% to 10% of children who are younger than 5 years visit an outpatient clinic with lower respiratory tract disease, accounting for up to 30% of the excess number of antibiotic prescriptions during winter seasons.5,,6 Population-based studies that have attempted to isolate influenza morbidity from other circulating winter pathogens note that rates of influenza-related hospitalization among healthy children who are younger than 2 to 3 years were approximately 4 to 20 times that of healthy children who are older than 5 years.6,,7
In adults, influenza causes substantial morbidity and mortality, with associated losses in work productivity and other indirect economic losses to society. Previous studies showed the potential cost-effectiveness of vaccination of working adult and elderly populations in managed care and other medical delivery settings.8–10 Economic losses also may be incurred when adults take time off from work or other activities to care for a child with influenza. Young children are thought to contribute significantly to transmission of influenza virus to one another and to adults, especially in school, child care, and household settings.11–13 Therefore, an influenza vaccination strategy that specifically targets young children may, in addition to preventing influenza disease, result in prevention of economic losses as suggested in recent studies of pediatric influenza vaccination.14,,15
Population-based studies have noted that use of the currently licensed influenza vaccine in pediatric and healthy young adult populations is low.4,,16 An investigational, live, attenuated, intranasal influenza vaccine may be an attractive alternative to currently available injectable influenza vaccines. We previously reported results from a large, randomized, clinical trial that demonstrated the safety and efficacy of this vaccine.17 The need for a cost-effectiveness evaluation of influenza vaccine campaigns targeted to healthy young children has been noted.18 The objective of the present study was to assess the potential cost-effectiveness (CE), including break-even costs, of intranasal influenza vaccine use in healthy children based on data from this large, prospective, clinical trial.17
Our study complied with published guidelines for minimizing bias in CE research that is funded wholly or partially by pharmaceutical companies.19 This study was approved by the appropriate institutional review board of each participating center. Informed consent for participation in the clinical trial was obtained from all participants.
Study Population and Vaccine Usage
Our economic analyses were based primarily on data collected during 2 influenza seasons (fall 1996 through spring 1998) during a multicenter, prospective, randomized, double-blind, placebo-controlled efficacy trial of a live, attenuated, trivalent, intranasal influenza vaccine17,,20 and thus reflect the clinical trial experience in these 2 seasons. This vaccine (Aviron, Mountain View, CA) contained 106.7–7.0 tissue culture infective dose50/dose of each of 3 attenuated strains that matched the antigens as recommended for the trivalent inactivated influenza vaccine by the Food and Drug Administration for the 1996 to 1998 seasons. Healthy children 15 to 71 months of age were randomized to receive either vaccine or placebo, and the vaccine group received either 1 or 2 doses of vaccine. Children were followed prospectively, and upper respiratory cultures were obtained and tested for influenza virus if influenza-like illness (ILI) occurred.
The main source of clinical, medical resource utilization and lost productivity data was the efficacy trial. CE analyses compared treatment with intranasal influenza vaccine with no vaccine. In the clinical trial, vaccine efficacy during the influenza season was measured by having children evaluated prospectively for the presence of influenza virus (by culture) based on a broad collection of clinical symptoms and signs to maximize detection of culture-positive children. Although culture-confirmed influenza was the primary outcome in the clinical trial, we believe febrile ILI to be a more meaningful outcome for clinicians, given day-to-day practice patterns and the realities of child care for caregivers. Thus, the outcome measure of effectiveness in the CE analysis was ILI fever days avoided. The cost measure included the direct costs and lost productivity associated with 1) provision of vaccine and its administration, 2) treating vaccine-related adverse events in the vaccine group, and 3) treating ILI in the vaccine and placebo groups.
Breakeven analyses were conducted to determine the vaccine/administration cost below which its use would be cost saving.
Definition of ILI
No accepted standardized definition of ILI exists for children; therefore, we developed a definition in this study. Any child who met any of the 5 criteria listed in the Direct Medical Resource Utilization section (below) and had a temperature of ≥101°F (oral, or its equivalent) was considered to have had ILI. Clinical visits that satisfied the criteria for ILI or in which influenza was culture-confirmed were used to estimate direct costs and lost productivity costs in the vaccine and control groups.
Definition of CE Analytic Perspective
The main CE analyses were based on 2 scenarios. In 1 scenario (referred to as individual-based vaccination), all caregivers were assumed to initiate a visit to a health care facility specifically for vaccination of the child. In the other scenario (group-based vaccination), the vaccination was performed in a group setting, such as a school or a child care facility, eliminating all caregivers' loss of time, productivity, and transportation costs associated with vaccination visits. The intranasal vaccine could be administered in the group-based vaccination setting by someone with very little training. In this scenario, we assume that vaccine cold chain requirements for transport and storage, proper screening methods, and authority and liability for vaccine administration are not at issue. Although the cost of vaccine plus administration would be different (likely less) in the group-based setting than in the individual-based setting, we assumed that the vaccine and administration cost was equal in each to minimize bias toward a positive vaccine effect.
The main analyses were conducted from the societal perspective.21,,22 The analyses included all direct costs and lost productivity costs that reasonably could be captured, estimated, or imputed. We also report results of an analysis that included direct medical costs only, as these results are particularly relevant for third-party payers. The CE is expressed as follows: CE = (Total cost V − Total costNV)/(Effectiveness V − Effectiveness NV), where V = vaccine group, NV = no vaccine group, Total cost = all costs associated with a given health outcome, and Effectiveness = given health outcome (ILI fever days). Thus, as compared with no influenza vaccine, the CE represents the marginal cost of treatment with vaccine divided by its marginal effectiveness.
The clinical trial evaluation period for which health outcomes and medical resource utilization was tabulated included the 10 days immediately following administration of each dose of vaccine or placebo and the 42 days during which vaccine-related serious adverse events were recorded. Costs and health outcomes were discounted at a rate of 3% in year 2 in the main analysis.21 Excel spreadsheet software was used to perform the analysis.
To a great extent, the resource utilization data were collected during the clinical trial. Data that were not collected directly were estimated from secondary sources. Table 1provides an overview of the types of costs and the resources included and their data sources.
Direct Medical Resource Utilization
Because the resource utilization behavior of patients who were enrolled in the clinical trial was expected to be much higher than that of the general population as a result of clinical trial protocol-driven visits, we defined clinical scenarios for which health care provider visits would be likely in a real-world population to avoid overestimation of medical resource utilization. We assumed that visits would be made by 1) any child who was younger than 36 months and had a temperature of ≥102°F (oral, or its equivalent); 2) any child who was younger than 36 months and had 2 or more systemic symptoms, including irritability, vomiting, decreased activity, muscle aches, chills, and headache; 3) any child who was younger than 36 months and had a temperature of ≥101°F (oral, or its equivalent) and a cough and/or a sore throat; 4) any child who was 36 months of age or older and had 3 or more systemic symptoms, including irritability, vomiting, decreased activity, muscle aches, chills, headache, or temperature of 102°F (oral, or its equivalent); or 5) any child of any age with wheezing, suspected otitis media, pulmonary congestion, or shortness of breath.
Medical resource use data collected during the clinical trial consisted of the number of doses of vaccine or placebo received, visits to a health care provider, days of prescription and over-the-counter medication use (no protocol restrictions were applied to medication use), other medical care as a result of study drug–associated adverse events, and lost productivity associated with ILI or culture-confirmed influenza in the child (Table 1). Because influenza vaccination reduced the occurrence of otitis media in the clinical trial, medical resource utilization associated with otitis media diagnosed concurrently with ILI was included in the analysis.
We used secondary data analysis and published data to account for resource utilization and lost productivity data not collected during the clinical trial, including hospitalization rates and length of stay for children with ILI, the proportion of the outpatient health care provider visits that likely would be with a physician versus a nurse, and the number and type of laboratory tests and procedures ordered (Table 2). The likelihood of hospitalization as a result of ILI during the influenza season was estimated as 0.00213, which assumes an ILI-related illness rate of hospitalization of 36/10 000 children who were younger than 5 years, 60% of which are attributable to influenza.23
We used data from the National Ambulatory Medical Care Survey24 and the National Hospital Ambulatory Medical Care Survey25 to estimate the percentage of visits as a result of ILI that would occur at a physician's office versus an emergency department and the types and percentages of procedures and laboratory tests used. These data were analyzed for similarly aged children with influenza defined by International Classification of Diseases, Ninth Revision Clinical Modification (ICD-9-CM) codes 487.0 to 487.8 for viral influenza, viral influenza with pneumonia, and viral influenza with and without other respiratory manifestations (Table 2).
The National Hospital Ambulatory Medical Care Survey is a national sample of hospital outpatient departments of noninstitutional general and short-stay hospitals, exclusive of federal, military, and Veterans Administration hospitals, located in the 50 states and the District of Columbia.25 In 1994, 29 095 patient records were obtained from 260 hospital outpatient departments, and 26 547 patient records were included from 418 emergency departments (affiliated with hospitals) that participated in this survey. The National Ambulatory Medical Care Survey data represent a national sample of individual-based nonfederal physicians excluding anesthesiology, pathology, and radiology.24 In 1995, 36 875 patient record forms were received from 1883 physicians who participated in this survey.
Direct Nonmedical Resource Utilization
Estimates were made for caregiver use of public transportation or vehicle mileage and parking associated with children's vaccine administration at a health care provider site, medical care for vaccine side effects, and medical treatment of influenza and related illnesses.
Clinical trial data included the primary caregiver's number of missed work days outside the home as a result of home care for a child with ILI. Information was not collected on the primary caregivers' missed usual activity (at home and in the workforce) associated with vaccination and missed usual activity for nonworkforce caregivers while caring for a child with ILI. Published literature and expert opinion were used to estimate the average caregiver hours associated with these activities.26–30 All analyses from the societal perspective included lost productivity as a result of both missed work and missed usual activity.
Direct Medical Costs
A base case total cost for the price of the vaccine plus the charge for vaccine administration (supplies, personnel, and other expenses) was set at $20. In previous studies of influenza vaccination, vaccine plus administration costs provided in an outpatient setting ranged from $10 per adult recipient of the intramuscular vaccine9 to $36 per child recipient of the intranasal vaccine (the latter representing the upper bound resulting in cost savings).15 Thus, a cost range of $10 to $40 was used in sensitivity analysis. The vaccine plus the charge for vaccine administration was set to 0 for the placebo group.
A fee schedule provided ranges of nationwide low and high private-sector charges for health care provider visits, procedures, and tests (Table 3).31 For all analyses, the midpoint of each range was determined and then weighted by the frequency of use by trial participants by the particular health care setting (office, emergency department) and health care provider (physician or nurse).
To estimate the cost of a hospitalization as a result of ILI, we used data from the Agency for Health Care Research and Quality, formerly the Agency for Health Care Policy and Research.32 Agency for Health Care Research and Quality's 1995 Nationwide Inpatient Sample (Release 4) comprises all-payer data on inpatient resource use and related charges. Designed to approximate a 20% sample of US hospitals (n = 900), the Nationwide Inpatient Sample contains records for all stays in the sampled hospitals, which are drawn from 19 states and selected to be representative of all US community hospitals. We determined the average inpatient charge and length of stay for ILI-related hospitalization for a similarly aged pediatric cohort with discharge diagnoses coded as ICD-9-CM 487.0 to 487.8 as a function of the mean national cost-to-charge ratio of 0.52,33 resulting in a mean cost per hospitalization of $2108 (Table 3); the mean length of stay was 3 days.
Direct Nonmedical Costs
The average caregiver transportation cost is noted in Table 3. We used clinical trial data collected on the number of children per family to weight the cost for vaccine administration, which assumed that all children were vaccinated at the same visit.
Lost Productivity Costs
We estimated the dollar value of missed work and other usual activity by multiplying the number of hours of missed activity by the average hourly US wage rate for full-time workers.34 The mean wage was weighted by gender only, as caregiver age was not available from the clinical trial data (Table 3).
When necessary, unit costs were inflated to 1998 to 1999 dollars using the medical component of the Consumer Price Index. These unit costs were applied to the volume of resources used by or estimated or imputed for clinical trial participants. Average costs per participant by resource component were summed for each arm to calculate the total direct medical cost per participant.
Other Assumptions for CE Analyses
The following additional assumptions, several of which were tested in sensitivity analyses, were made for the main economic analyses:
Children who were younger than 9 years received 2 doses of vaccine in year 1 and 1 dose per year thereafter. In the clinical trial, more than 80% received 2 doses in year 1. To simulate the expected usage of vaccine after licensure, in each year, 20% of children (the youngest of 5 hypothetical age cohorts per year) received 2 doses of vaccine in the vaccine group; the remaining 80% of children in the vaccine group received 1 dose per year.
The clinical trial included some cases for which no health care provider visit was made but for whom medication was prescribed. For 25% of such cases, we attributed a $4.46 cost to a telephone consult with a nurse.35
For any hospitalized child's caregiver, there were 3 8-hour days of lost usual activity and 1 round trip per day to the hospital, based on a mean 3-day stay in the hospital.32
A health care provider visit for vaccination (in the individual-based setting) or diagnosis of ILI required 2 hours of caregiver time.28,,30 This lost productivity was weighted by the number of children enrolled per family in the clinical trial at baseline, assuming that all children in a family are vaccinated at the same visit.
Employed caregivers worked an 8-hour day.
The proportion of caregivers in the workforce was 50%.34
Care at home for a child with ILI required 4 hours of usual activity lost per day for a caregiver who did not work outside the home.
The rate of transmission of influenza from an infected child to at least 1 family member was 18%.12
Based on the clinical trial data, the proportion of households with 2 adults was 85%.
An adult with influenza loses an estimated minimum of 1.5 days of productivity, based on reported lost work days.36
Other Sensitivity Analyses
We conducted several 1- and 2-way sensitivity analyses. The vaccine and vaccine administration cost was assumed to be $20 per dose in the main analyses for both scenarios. One-way sensitivity analyses included changes in the vaccine costs (50%–200%); caregiver hourly wage (50%–150%); daily hours of lost productivity for at-home caregivers (50%–200%); private-sector charges for physician visits, tests, and procedures; utilization of Medicaid charges for physician visits, tests, and procedures (combined with Medicaid recipients' average workforce participation rate and hourly wage)37; rate of hospitalization for ILI (50%–200%); rate of transmission of influenza to family members (44%–156%); and the percentage of children who received 1 or 2 doses of vaccine per year.
We performed 2-way sensitivity analyses in which we retained the same vaccine effectiveness as in the main scenario analyses but assumed the following: 1) 10% or 2) 0% of children per year were vaccinated twice in a given year with cost assessments as vaccine/administration cost varied from $10 to $40 per dose.
Clinical Trial Data
The clinical trial enrolled 1070 patients in year 1 and 917 patients in year 2 in the vaccine group and 532 patients in year 1 and 441 patients in year 2 in the placebo group. More than 97% of participants completed the study. No patients experienced adverse events that led to withdrawal, and no adverse events resulted in added resource utilization; therefore, no costs were attributed to vaccine-associated adverse events.
In each year, the baseline demographic characteristics of the patients in the vaccine and placebo groups were very similar; the mean age was 42 months at enrollment in year 1, approximately 85% of children were white, 52% were female, 35% were in child care 5 days/wk at enrollment, and 85% had 2 adults in the home. During the 2-year study period, a total of 1502 ILI visits were included in the analysis; vaccinated children had an average of 1.2 fewer ILI fever days/child than unvaccinated children (2.74 vs 4.95, respectively). For each group, approximately 20% of visits represented multiple visits in the same year.
CE Analyses for the Main Scenarios
Per-child CE was calculated separately for the individual-based and group-based vaccination scenarios from the societal perspective. CE averaged over the 2-year period was $29.67 per ILI fever day avoided for individual-based vaccination scenario and dominant (both cost saving and more effective than placebo) for group-based vaccination scenario (Table 4). In the individual-based vaccination scenario, 45% of lost productivity was attributable to the dedicated visit to be vaccinated.
The vaccine/administration cost below which its use would be cost saving was $28 for group-based vaccination from the societal perspective. For the individual-based vaccination from the societal perspective, the break-even vaccination cost was $4.93. From the third-party payer perspective, both individual-based and group-based vaccination resulted in the same break-even cost of $10.29 as this perspective includes direct medical costs but not costs attributable to caregiver lost productivity as a result of ILI or transportation for vaccine administration.
Additional Sensitivity Analyses
The results of 1-way sensitivity analyses are shown in Table 4. Considering the third-party perspective, the CE was $19.10 per ILI fever day avoided for both the individual-based and group-based vaccination, because this perspective includes only direct medical costs. The largest changes in CE for the individual-based vaccination scenario resulted from increasing the vaccine cost. From the societal perspective, the CE ranged from $10 per ILI fever day avoided at a vaccine/administration cost of $10/dose to $69 at $40/dose. Considering direct medical costs only, the CE ranged from $0.52 at $10/dose to $58.47 at $40/dose. As shown in Table 4, group-based vaccination results in “dominance” for the vaccine in the vast majority of sensitivity analyses; that is, the vaccine program is both cost saving and more effective than no vaccine use.
The results of 2-way sensitivity analyses for the individual-based and group-based vaccination scenarios are shown in Table 5. In our main analyses, 80% of children (in the age group studied) received 1 dose of vaccine and 20% received 2 doses per year. In the 2-way sensitivity analysis, 10% and 0% rather than 20% of children were assumed to receive 2 doses of vaccine per year. Because no significant difference in vaccine efficacy was noted for 1 versus 2 doses in the clinical trial and there were more febrile illnesses with 2 doses than with 1 dose in the clinical trial, we held vaccine effectiveness constant. From the societal perspective in the individual-based scenario, if 10% of children receive 2 doses per year, then the resulting cost per ILI fever day avoided ranged from $4.68 to $63.84 as the vaccine/administration costs ranged from $10 to $40, respectively (Table 5). When all children received only 1 dose of vaccine, cost per ILI fever day avoided ranged from dominance to $40.59 at $10 to $40/dose, respectively. From the third-party perspective, similar trends were noted except that CE was dominant at the lowest vaccine cost per dose ($10).
Table 5 indicates the results of these analyses for the group-based vaccination scenario as well. From the societal perspective, CE results remain dominant for the vaccine at $10 to $25/dose when 10% of children receive 2 doses per year; at $40/dose, the cost per ILI fever day avoided is $18.52. When all children receive 1 dose per year, CE remains dominant for the vaccine at $10 to $30/dose; at $40/dose, the cost per ILI fever day avoided is $11.96. Results from the third-party payer perspective for group-based vaccination are identical to those shown in Table 5 for individual-based vaccination.
Our data suggest that use of intranasal influenza vaccine in young children may be cost-effective when reduction in ILI fever days is used as the primary outcome measure. Our results are derived from a large sample of children and include a broad range of outcomes, including utilization of outpatient and inpatient clinical services (as a result of vaccination and/or ILI) and caregivers' lost productive time from usual activities. This study is the first to evaluate an influenza vaccination strategy that targets young children and uses data from a randomized clinical trial. Such data allow for more precise estimates of vaccine efficacy and disease-associated direct medical and lost productivity costs. Our findings support earlier economic modeling analyses of influenza vaccine immunization programs that suggested that vaccination of healthy children would have substantial economic benefits to society.14,,15
Increasingly, national bodies that formulate vaccine policy consider not only vaccine safety and efficacy but also vaccine cost-effectiveness. The relative usefulness of our CE estimates, therefore, should be considered in the context of similar analyses of other vaccines targeted toward prevention of disease in children. These include pneumococcal conjugate vaccine,38 hepatitis B,39 varicella,29 rotavirus,40Haemophilus influenzae type b,41H influenzae type b–hepatitis B combination,42 and maternal immunization for prevention of neonatal group B streptococcal disease.43 Varicella and rotavirus are most similar to influenza in that disease incidence in children is high but the likelihood of severe complications and mortality is relatively low. Lieu et al38 noted that routine varicella vaccination of infants was cost saving from a societal perspective and cost ∼$2 per case prevented from the third-party payer perspective. Tucker et al40 noted that routine rotavirus vaccination of infants cost $103 per case prevented from a third-party payer perspective. These data are similar in magnitude to the results of our CE analysis, not considering the potential CE advantages in the sensitivity analyses noted above; that is, for individual-based vaccination, a cost of ∼$30 per ILI fever day avoided from the societal perspective and ∼$19 from the third-party perspective and dominance for group-based vaccination from the societal perspective. Our data for influenza vaccine also are similar to the above-mentioned studies regarding the break-even vaccine costs. Mere comparison of CE estimates, however, is inadequate to appreciate fully a particular vaccine's general worth to individuals, third-party payers, or society. Such decisions require careful integration of many factors, many of which are unquantifiable.
The main CE outcome measure for our study was the cost per ILI fever day avoided. This was chosen to allow for an estimation of the clinical impact of influenza vaccination. The total costs included direct medical, direct nonmedical, and lost productivity. The CEs generally were less advantageous from the societal perspective than from the third-party payer perspective for individual-based vaccination, a finding not usually seen in CE analyses of this type. This effect resulted from the relatively high lost productivity of a caregiver as a result of acquiring the vaccine on an annual basis (in the individual-based setting), compared with modest gains in productivity for the caregiver resulting from lower disease incidence and/or severity of the disease for vaccinated children. If vaccine were delivered in a setting that minimized caregiver lost productivity and transportation costs incurred for receipt of the vaccine itself (eg, in child care, school, or community groups), the CE of this vaccine would improve to the level of being cost saving to society under most circumstances.
The sensitivity analyses in this study demonstrate the impact of specific variables on our main CE analyses of influenza vaccination. The most important are cost of vaccine, the proportion of children who require 2 doses of vaccine in any given year, and the vaccination setting (individual vs group based), which defines the degree of caregiver lost productivity in obtaining the vaccine. The most advantageous scenario involves receipt of 1 dose of vaccine annually, administered in a group-based setting. One other study that evaluated the economic impact of targeting healthy pediatric populations also found that addition of indirect costs to the analysis and vaccine cost was an important cost driver.14 In that study, when indirect costs were excluded from the analysis, only group-based vaccination strategies remained cost saving. Although we believe that intranasal influenza vaccination of healthy young children is cost-effective, influenza vaccine currently is not recommended for children who do not have chronic health conditions. To this end, our sensitivity analyses may be useful for extrapolation to circumstances encountered locally and/or nationally if and when the intranasal vaccine is licensed and vaccination programs are developed.
A hallmark of influenza is its ability to disrupt the community by causing child care, school, and workplace absenteeism.9,,44,45 Infection rates in young children are comparable to or higher than in adults,46–48 and lost productivity as a result of influenza at all ages results in substantial costs to society.14 Preschools are now an integral part of American life and (as with schools) increase the risk of infection for participating children and for secondary transmission to family members.49 As such, an influenza vaccination program that targets all healthy children. rather than only those with chronic health conditions, may increase its clinical and economic impact.
Our study has certain limitations that affect the final CE estimates. Several assumptions that we made might bias toward a lower final CE. For example, we assumed that 80% of the cohort of children (younger than 6 years) who receive vaccine will receive only 1 dose per year. If the intranasal vaccine is licensed for general use for a broader age range than that included in our clinical trial, then a greater proportion will receive only 1 dose per year, therefore increasing its CE. We also did not include potential cost savings associated with vaccine-based herd immunity because it is unknown what proportion of children would require vaccination to induce such immunity in unvaccinated individuals, should it occur. Similarly, we did not include the results from a stochastic modeling study, which showed that there would be a >95% decrease in the likelihood of a community-wide influenza epidemic if even as few as 70% of young children were vaccinated.50 Conversely, we included the economic impact of decreased transmission of influenza within families, and some of our data were from secondary data sources. In addition, the results of this study are based on only 2 years of vaccine experience and, thus, do not necessarily reflect “average” vaccine efficacy and attack rates. Furthermore, potential added costs to account for changes in logistics (eg, clinic freezer space) were not included in the analysis. However, we believe, on the whole, that our analyses underestimate the potential cost-effectiveness of this vaccine.
The intranasal influenza vaccine has been shown to be safe, immunogenic, and effective in preventing influenza in young children.17 This vaccine also may provide some advantage over the inactivated vaccines currently in use. Perhaps most important, its mode of delivery (nasal mist) likely will be preferred over intramuscular injection, currently a barrier to immunization with the licensed inactivated influenza vaccines.1 Vaccination also results in less severe clinical disease among vaccinees compared with controls and in year 2 afforded protection against a heterologous strain not included in the vaccine.17,,20 In addition to serum antibody, intranasal influenza vaccine induces mucosal (IgA) immunity, which may contribute to more efficient clinical protection.20 These characteristics, considered with our CE analysis in different logistic and payment settings, may facilitate its potential acceptance as a routine vaccine for young children.
Our data suggest that vaccination of young children has the potential for economic benefits to society. This impact may be maximized if vaccination is performed in group-based settings such as child care and elementary schools.
We received funding for this study from Aviron and the National Institute of Allergy and Infectious Diseases, National Institutes of Health (contract NO1-AI45249).
We thank the following for their contributions to this study, without whom it would not have been completed: Joel Ward, MD; Susan Partridge; Jessica Boring; Ann Vannier, MD; Swei-ju Chang; and Mellie Badar.
- Received January 5, 2001.
- Accepted March 26, 2001.
Reprint requests to (B.R.L.) MEDTAP International, 7101 Wisconsin Ave, Ste 600, Bethesda, MD 20814. E-mail:
- CE =
- cost-effectiveness •
- ILI =
- influenza-like illness, ICD-9-CM, International Classification of Diseases, Ninth Revision, Clinical Modification
- Foy HM,
- Cooney MK,
- Allan I
- ↵Glezen WP, Six HR, Frank AL, et al. Impact of epidemics upon communities and families. In: Kendal AP, Patriarca PA, eds.Options for the Control of Influenza. New York, NY: Alan R. Liss; 1986:63–74
- ↵Centers for Disease Control and Prevention. Prevention and control of influenza: recommendations of the Advisory Committee on Immunization Practices. Advisory Committee on Immunization Practices (ACIP).MMWR Morb Mortal Wkly Rep. 1999;48(RR-4):1–34
- Glezen WP,
- Decker M,
- Joseph SW,
- Mercready RG
- Longini I,
- Koopman JS,
- Monto AS,
- Fox JP
- Frank A,
- Taber LH,
- Glezen WP,
- Geyer EA,
- McIlwain S,
- Paredes A
- ↵White T, Lavoie S, Nettleman MD. Potential cost savings attributable to influenza vaccination of school-aged children. Pediatrics. 1999;103(1). Available:http://www.pediatrics.org/cgi/content/full/103/1/e73
- Cohen GM,
- Nettleman MD
- ↵Gold MR, Siegel JE, Russell LB, Weinstein MC, eds.Cost-Effectiveness in Health and Medicine. New York, NY: Oxford University Press; 1996
- ↵Haddix AC, Teutsch SM, Shaffer PA, Dunet DO. Prevention Effectiveness: A Guide to Decision Analysis and Economic Evaluation. New York, NY: Oxford University Press; 1996
- Couch RB,
- Kasel JA,
- Glezen WP,
- et al.
- ↵National Center for Health Statistics. National Ambulatory Medical Care Survey . CD-ROM Series 13, No. 11, SETS Version 1.22a, July 1997
- ↵National Center for Health Statistics. National Hospital Ambulatory Medical Care Survey (NHAMCS) [CD-ROM]. Series 13. No. 10. SETS Version 1.22a. Rockville, MD: National Center for Health Statistics; 1997
- ↵Luce B, Manning WG, Siegel JE, Lipscomb J. Estimating costs in cost-effectiveness analysis. In: Gold MR, Siegel JE, Russell LB, Weinstein MC, eds. Cost-Effectiveness in Health and Medicine. New York, NY: Oxford University Press; 1996:200–203
- Lieu TA,
- Cochi SL,
- Black SB,
- et al.
- Preblud SR,
- Orenstein WA,
- Koplan JP,
- Bart KJ,
- Hinman AR
- ↵HealthCare Consultants of America, Inc. 1998 Physician's Fee and Coding Guide: A Comprehensive Fee and Coding Reference. Augusta, GA: Healthcare Consultants of America, Inc; 1998
- ↵Agency for Health Care Research and Policy (AHCPR), Center for Organization and Delivery Studies. Healthcare Cost and Utilization Project (H-CUP3). Rockville, MD: AHCPR; 1998
- ↵Prospective Payer Commission (PROPAC). Annual Report. 1997
- ↵Bureau of Labor Statistics. http://www.bls.gov, June 1999
- Grabowsky M,
- Markowitz L
- Adams PF,
- Marano MA
- Khan AS,
- Polezhaev F,
- Vasiljeva R,
- et al.
- ↵Schoenbaum SC. The economic impact of influenza: the individual's perspective. Am J Med. 1987;82(suppl 6A):26–30
- Monto AS,
- Ohmit SE,
- Margulies JR,
- Talsma A
- Monto AS,
- Koopman JS,
- Longini IM Jr.
- Fox JP,
- Hall CE,
- Cooney MK,
- Foy HM
- Goodman RA,
- Churhill RE,
- Sacks JJ,
- Addiss DG,
- Osterholm MT,
- Cardinale V, ed. 1998 Redbook. Montvale, NJ: Medical Economics Company; 1998
- Copyright © 2001 American Academy of Pediatrics