Economic Analyses of Respiratory Syncytial Virus Immunoprophylaxis in High-Risk Infants:  A Systematic Review FREE

RESPIRATORY SYNCYTIAL virus (RSV) is the leading cause of lower respiratory tract illness in infants and young children worldwide.¹ Almost all children have been infected with RSV by age 2 years.^2- 4 In the United States alone, there are more than 90 000 hospitalizations and 4500 deaths per year attributable to RSV.^1,5 Infants born prematurely and those experiencing chronic lung disease (CLD), congenital heart disease, or immunodeficiency are at high risk for severe RSV-related disease.^6- 8

Accompanying the heavy medical burden caused by RSV are large monetary costs. The Institute of Medicine projected annual total costs in the United States attributable to RSV infection to be $342 million and placed RSV disease at the second highest level of priority for vaccine development.^1,9 A recent retrospective analysis¹⁰ estimated inpatient charges to be $300 million to $400 million per year for hospitalization owing to RSV pneumonia among children 4 years and younger in the United States. Other costs attributable to RSV include direct costs for outpatients,¹¹ costs for follow-up after hospitalization,¹² and indirect costs of parental time lost from work.^11,13 However, because hospitalizations are the major drivers of overall RSV infection–related costs,¹¹ any prophylactic interventions directed at either preventing hospitalizations or shortening the duration of hospital stays are expected to have an impact on reducing the cost of this illness.

No vaccine is available against RSV. However, 2 products are currently licensed in the United States for passive immunoprophylaxis against RSV: respiratory syncytial virus immunoglobulin intravenous (RSV-IGIV) (RespiGam; MedImmune, Inc, Gaithersburg, Md), containing high-titer RSV antibodies,¹⁴ and palivizumab (Synagis; MedImmune, Inc), a humanized murine monoclonal antibody against RSV.¹⁵ The American Academy of Pediatrics^16- 17 has made recommendations for the use of these agents in certain high-risk infants (Table 1). Children are recommended to receive 1 of the 2 agents once per month throughout the RSV season, which is usually November to April. The dose of RSV-IGIV is 750 mg/kg of body weight¹⁴; palivizumab is given intramuscularly at a dose of 15 mg/kg of body weight.¹⁵ Two large efficacy trials using RSV-IGIV^18- 19 and 1 trial of palivizumab²⁰ have proved the benefits of these agents in reducing RSV hospitalizations and hospital days (Table 2). None of these trials^18- 20 had sufficient power to detect a difference in the mortality rates between the treatment groups.

Although the benefits of administering RSV-IGIV and palivizumab have been demonstrated for high-risk infants, the costs associated with these agents are significant. The average wholesale prices of the drugs alone in 2001 were $816.32 per 50 mL for RSV-IGIV and $1320.77 per 100-mg vial for palivizumab.²¹ Hence, to complete a 5-month prophylactic course in a 4.5-kg infant, RSV-IGIV would cost $5510.16 and palivizumab would cost $4457.60, assuming no wastage of the drug. These high costs have raised considerable debate on the relative value of these agents in terms of their costs and benefits.^22- 28 When the American Academy of Pediatrics first published, in 1998, recommendations for the use of palivizumab and an update on RSV-IGIV use,¹⁶ few cost-effectiveness analyses^29- 32 had been conducted on RSV-IGIV and no published economic data were yet available on palivizumab. In this era of growing demands on the limited health care resources, payers, providers, and policymakers are becoming increasingly cost-conscious. However, use of a single published study on the economic impact of an intervention may be misleading. This review summarizes and critically evaluates all available economic analyses of the 2 RSV immunoprophylactic agents: RSV-IGIV and palivizumab.

Economic analyses evaluating RSV immunoprophylaxis were identified using the MEDLINE and HealthSTAR databases. A search was conducted using various combinations of the following key search terms: respiratory syncytial virus immunoglobulin intravenous, palivizumab, cost, and cost-effectiveness. The search was limited to articles published in English between January 1, 1990, to August 31, 2001. Additional studies were obtained by searching the bibliographies of all relevant identified articles. We contacted several experts asking about knowledge of other studies. We also searched the Cochrane Database of Systematic Reviews, the National Health Service Economic Evaluation Database, the National Institute of Clinical Excellence, and Canadian Coordinating Office of Health Technology Assessment publications to identify any further studies or reviews of these agents. We included only studies that performed an economic analysis of either or both of the agents (RSV-IGIV and palivizumab) in an infant population. Letters to the editor and commentaries containing informal economic analyses were excluded. Once selected, 2 of us (S.K.-B. and J.D.) independently reviewed the articles and extracted summary information using a standardized abstraction form, and differences were resolved by consensus after returning to the original study. The abstraction form was developed by adapting the cost-effectiveness analysis reporting checklist published by the Panel on Cost-effectiveness in Health and Medicine.³³ From each selected study, the following elements were abstracted: author(s), year of publication, drug alternatives being compared, patient population, sample size, type of economic analysis, perspective of the analysis, type of costs included, year in which costs were measured, outcome measures, source of cost data, source of effectiveness data, discount rate, baseline results, sensitivity analysis, author conclusions, and source of funding.

An initial search of the literature yielded 21 potentially eligible articles, 9 of which were excluded from this review: 7 were letters to the editor or commentaries,^{22- 23,26- 28,34- 35} 1 evaluated the cost-effectiveness and cost-benefit of an infection control program to reduce the rate of RSV infection and no prophylactic agent was administered,³⁶ and another evaluated a hypothetical vaccine in the elderly.³⁷Table 3 presents a summary of the key elements of the 12 studies included in this review.^{29- 32,38- 45} Six studies^{29- 32,38- 39} evaluated RSV-IGIV against the no prophylaxis option, 3 studies^40- 42 conducted an economic evaluation of palivizumab, and 3 studies^43- 45 evaluated RSV-IGIV and palivizumab. Four studies^{29,38,40- 41} reported funding from the manufacturer. Selection of study participants and analysis of subpopulations varied from study to study, stratifying the risk groups by gestational age or the presence or severity of CLD, other medical conditions, or both (Table 3). Six studies explicitly reported the weights of the infants being considered in the analysis; some^{31- 32,43,45} conducted an explicit analysis for a range of weights, and others^42,44 assumed a single weight for the entire analysis.

A variety of economic evaluation approaches were used. Seven studies^{29- 30,39,41,43- 45} performed a cost-effectiveness analysis, and another 4 studies^31,38,40,42 conducted an incremental cost analysis but did not express the results in cost-effectiveness terms. One study³² performed a cost-benefit analysis using willingness-to-pay techniques to measure the value of health benefits gained with use of RSV-IGIV in monetary terms. In 8 studies,^{29- 31,39- 41,43- 44} the economic perspective of the analysis was explicitly stated. Most studies used a third-party payer or provider perspective. These studies included only direct medical costs associated with management of RSV infection and direct savings associated with use of the prophylactic agents. Three studies^29,32,43 used a societal perspective, including indirect costs in addition to the direct medical costs. Only 4 studies^29,31,39,43 reported the year in which costs were measured. Two of these studies^29,43 also considered discounting of future costs and benefits using a discount rate of 3% per year. Cost data were obtained from a single center or a few institutions in most studies, with published sources used to supplement the data in a few of these studies. The type of institution varied from academic, tertiary care medical centers to health maintenance organizations. One study⁴⁰ collected cost data reported only from the literature. In most studies, the effectiveness data were based on previously published efficacy results from a single clinical trial. Two studies^32,43 pooled data from clinical trials of RSV-IGIV. One study³⁸ conducted a retrospective analysis using effectiveness data on RSV-IGIV from hospital databases. Respiratory syncytial virus hospitalization rates in the no prophylaxis groups and RSV hospitalization costs varied considerably across the studies. Different outcome measures were used across the studies. Incremental cost per hospitalization prevented and incremental cost per patient were the primary economic measures used. One study⁴⁵ reported incremental costs per hospital day saved, and another⁴¹ calculated incremental costs per RSV infection episode avoided. Three studies^29- 30,43 also reported their results in terms of incremental costs per life-year saved.

To provide a more comprehensive discussion of the important findings and issues raised by the studies included in this review, in the following subsections we present results from studies grouped by the immunoprophylactic agent that was evaluated in each study.

Of the 6 studies that analyzed only RSV-IGIV, 4^29- 32 did not find use of RSV-IGIV to result in cost savings for the entire infant group or any specific subgroup. The study by Atkins et al³⁸ is the only one that reports overall cost savings ($435 994 per 100 infants) with use of RSV-IGIV in all high-risk infants. However, its results should be interpreted with caution. This study compared the cost of RSV-related hospitalization during the 2 winter seasons before and after initiation of routine RSV prophylaxis in an actual clinical setting. A preintervention–postintervention study design lacking a control group is limited by the fact that other factors, such as year-to-year fluctuations in the burden and strain of RSV, rather than RSV-IGIV alone, may have affected the number of RSV hospitalizations. On the other hand, Barton et al³⁹ found that limiting use of RSV-IGIV to a group of infants at highest risk of severe RSV illness and subsequent admission to the intensive care unit would result in cost savings up to $27 000 per hospitalization prevented. However, if RSV-IGIV were administered to all infants meeting Food and Drug Administration–approved indications (Table 1), the cost to prevent 1 hospitalization due to RSV infection would exceed $53 000. Hay et al²⁹ concluded that RSV-IGIV therapy was cost-effective for all high-risk infants ($24 305 per life-year saved), whereas Oelberg et al³⁰ concluded that RSV-IGIV therapy is cost-effective when limited to infants with active bronchopulmonary dysplasia (BPD), at $7764 per life-year saved. However, both of these studies proposed that RSV-associated mortality rates are reduced with RSV-IGIV therapy, an assumption that has not been proved in clinical studies to date. Furthermore, Hay et al²⁹ factored into the calculation of benefits not only the life-years saved of the child but also the life-years of his or her potential children, grandchildren, and so on; this assumption inflates the anticipated benefits from use of RSV-IGIV. O'Shea et al³¹ reported the incremental cost of prophylaxis to be lowest in the subgroup with BPD and younger than 3 months ($1689 per 1.2-kg infant). However, the authors did not present their results in cost-effectiveness terms. Robbins et al³² estimated a threshold number needed to treat (where costs and benefits of preventing hospital admission due to RSV are equal) and concluded that the costs outweigh the benefits of RSV-IGIV treatment for the overall sample of infants and the subgroup of premature infants without BPD. However, based on sensitivity analysis results, this study neither strongly discourages nor promotes use of RSV-IGIV in infants with BPD.

Of the 3 studies^40- 42 that conducted an economic evaluation of RSV immunoprophylaxis with palivizumab, 2^40- 41 used a decision-analytic model populated with data from the IMpact trial and other literature. With a 55% palivizumab therapy–related reduction in RSV hospitalization rates for all infants at risk for lower respiratory tract illness, Marchetti et al⁴⁰ reported results ranging from expected incremental expenses of $3524 per infant to cost savings as high as $34 856 per infant based on ranges of RSV hospitalization rates for infants receiving no prophylaxis (10.6%-42.6%), and of hospital charges per patient ($10 236-$166 375). The study by Lofland et al⁴¹ reported an incremental cost of $2702 to $79 706 per RSV infection episode avoided for a range of RSV infection rates of 10% to 38% in the no prophylaxis group and 5% in the palivizumab group. The analyses indicated cost savings only in the scenario in which the cost of palivizumab therapy was assumed to be much lower ($2500 compared with $4500) considering that all 5 doses would not be administered. That study differed from the study by Marchetti et al⁴⁰ on 2 major premises: source of cost data (single urban academic medical center vs literature) and economic outcome measure (cost per RSV infection episode avoided vs cost per infant). Thomas et al⁴² conducted an incremental cost analysis of palivizumab among infant groups in a large regional neonatal unit in the United Kingdom. For most infant groups categorized to correspond with the American Academy of Pediatrics guidelines on RSV immunoprophylaxis, the potential cost of palivizumab prophylaxis far exceeded the actual cost of hospitalization despite the study being biased in favor of palivizumab prophylaxis for 2 reasons: (1) infants with all lower respiratory tract illnesses, irrespective of whether they are due to RSV, were included in the calculation of hospitalization costs and (2) palivizumab use was assumed to prevent all hospital admissions of patients with RSV.⁴²

Three studies^43- 45 evaluated RSV-IGIV and palivizumab, and all provided mixed results. Joffe et al⁴³ conducted a cost-effectiveness analysis to compare RSV-IGIV and palivizumab with no prophylaxis. Neither of the 2 agents demonstrated cost savings, although the authors reported palivizumab to be both more effective and less costly than RSV-IGIV for all subgroups. Both agents were most cost-effective for the subgroup of infants whose gestational age was 32 weeks or younger, who required 28 days or more of oxygen therapy in the neonatal intensive care unit, and who were discharged from the neonatal intensive care unit between September and November. Palivizumab cost $12 000 per hospitalization averted, and RSV-IGIV cost $25 000 per hospitalization averted in this subgroup. In all other subgroups, these estimates, for both agents, were several-fold larger and varied greatly. Stevens et al⁴⁴ also reported no net cost savings of RSV prophylaxis with either RSV-IGIV or palivizumab. However, in contrast to Joffe et al,⁴³ neither agent was clearly more cost-effective than the other for all subgroups analyzed. Use of RSV-IGIV was more cost-effective among infants requiring respiratory support at 36 weeks' postconceptual age or older, at an incremental cost of $11 468 per hospitalization prevented. Conversely, palivizumab was more cost-effective for infants requiring respiratory support at less than 36 weeks' postconceptual age at an estimate of $32 792 per hospitalization averted. Among other factors, the results of the 2 studies differed primarily because Stevens et al⁴⁴ used subgroup-specific efficacies and Joffe et al⁴³ used overall efficacies for all participants from the clinical trials of RSV-IGIV and palivizumab. Given that palivizumab had significantly higher efficacy in infants without CLD (78%) than in those with CLD (39%) and RSV-IGIV showed a nonsignificant difference in the reverse direction (20% vs 49%), it is clear why the cost-effectiveness estimates in the study by Stevens et al⁴⁴ varied greatly by subgroup and prophylactic agent used. Numa⁴⁵ evaluated the cost-effectiveness of RSV-IGIV and palivizumab using an Australian hospital and intensive care unit database. On the basis of the estimates of costs per hospital day saved, the author indicates that, from an Australian perspective, neither RSV-IGIV nor palivizumab use was cost-effective in any infant group.

Finally, given that both agents are manufactured by the same pharmaceutical company, we stratified the results of the 12 studies included in the review by the presence or absence of manufacturer funding. Studies with some form of funding from the manufacturer were more likely to report the possibility of cost-effectiveness or cost-savings of prophylaxis in the entire high-risk infant population either in their point estimates or in their sensitivity analysis (4/4 vs 0/8; P = .002, Fisher exact test). When the likelihood of reporting favorable cost-effectiveness or cost-savings in either the entire high-risk population or specific infant subgroups was compared across studies, the association with manufacturer funding was not statistically significant (4/4 vs 3/8; P = .08, Fisher exact test).

This review identified wide variability in the results of economic studies of RSV-IGIV and palivizumab prophylaxis, with estimates ranging from cost savings to incremental costs of a high order of magnitude with the use of either agent. In general, neither agent was cost-effective if administered to all infants for whom it is approved, as per the Food and Drug Administration, or for whom it is recommended, as per the American Academy of Pediatrics (Table 1). Based on economic grounds, use of both agents was justified only in specific high-risk subgroups in most studies that performed a subgroup analysis. However, all studies did not concur in their conclusions on the specific high-risk subgroups in which use was considered cost-effective. Economic evaluations have generally reported greater cost-effectiveness of RSV-IGIV prophylaxis in infants with BPD (or CLD), very low-birth-weight infants, and infants at risk of intensive care unit admission.^30- 31,39 Broadly speaking, economic analyses of palivizumab suggest that it would be most cost-effective in the subgroup of premature infants without BPD^40,44 and those with gestational age 32 to 35 weeks^40- 41 because it showed the highest efficacy (78%-80%) in these subgroups in the IMpact trial.²⁰ The few economic studies^43- 45 comparing the 2 agents reported mixed results as to which agent is more cost-effective in different infant subgroups. Economic comparisons of the 2 agents are difficult given that no studies have been conducted specifically to compare the relative efficacy of the 2 agents. However, selection between the 2 agents is also governed by other factors, such as ease of administration, rate of adverse effects, presence of an immunodeficiency, and interference with the schedule of routine vaccines.

The divergent results among the studies can partly be explained by differences in the study methods and assumptions, but they are also a reflection of the poor quality of some of the economic analyses. The differences across studies in the RSV hospitalization rates in the group receiving no prophylaxis and the average cost of RSV-related hospitalization greatly affected the cost-effectiveness ratios. Other factors contributing to the variation are the differing assumptions on the number of doses administered and the weight of the infants across the studies. Given the high acquisition costs of these agents, an assumption of administering less than the routine 5-dose regimen without making adjustments to the efficacy data in the model would bias the cost-effectiveness estimate in favor of the agent. Furthermore, because the dose of the 2 agents depends on the body weight of the infants, cost estimates may significantly vary across studies if investigators do not account for this variable. Half of the studies did not explicitly report the weight of the infants being considered in the analysis. These studies were more likely to report results of favorable cost-effectiveness or possibility of cost savings with use of the agents either in all infants or specific subgroups. One may question whether these positive results would hold if the mean infant weight (not reported) was to be varied. One quarter of the studies calculated cost-effectiveness ratios in terms of cost per life-year saved on the basis of unproved assumptions of reduced mortality with use of these agents. Two of these 3 studies made conclusions of favorable cost-effectiveness in all infants or subgroups based on these ratios. Some of the studies^40- 42 that evaluated palivizumab did not compare it with the appropriate comparator treatment, that is, the available agent RSV-IGIV. An insurer or provider making decisions on whether to use palivizumab will be interested not only in its effectiveness relative to placebo but also in its clinical and cost-effectiveness relative to RSV-IGIV. Nearly one third of the studies evaluated in our review presented results in incremental costs per patient but did not express them in cost-effectiveness terms. Unless there are cost savings, more informed decisions can be made with the knowledge of the amount of additional benefits being gained with the incremental costs. About two thirds of the studies did not state the year in which costs were measured. This omission limits one's ability to inflate cost-effectiveness estimates of multiple studies to a common and more recent year's dollar figures and thereby to perform accurate comparisons. About one quarter of the studies did not perform a sensitivity analysis to test the robustness of the results and report variation around the cost-effectiveness estimates. One third of the studies did not explicitly state the perspective of the analysis. Finally, careful interpretation of all economic analyses is essential. For example, it is important to realize that the expected savings of $34 856 per infant with palivizumab use reported in the study by Marchetti et al⁴⁰ occurs only when the RSV-related hospitalization rates are as high as 42.6% in the no prophylaxis group and charges are as much as $166 375 per hospitalization. These RSV hospitalization rates and charges are several-fold higher than those reported in other studies.

These findings call for stricter adherence to guidelines on conducting and reporting pharmacoeconomic studies to reduce the sources of variation in future analyses and allow for the reliable use of the results in the development of guidelines on use of these agents. For future RSV prophylaxis–specific studies, authors should explicitly state the weight of the infants being considered, incorporate a complete regimen of 5 doses into their base-case calculation, and then, within the sensitivity analysis, test the impact of varying these 2 variables in addition to other variables. The wide geographic and institutional variability in RSV hospitalization rates and costs also necessitates an extensive sensitivity analysis around these variables to allow decision makers to incorporate their local and regional differences in these variables while interpreting the cost-effectiveness of these agents.

Our findings on the association between manufacturer funding and study results of favorable cost-effectiveness or cost savings in the entire high-risk infant population add to the rising skepticism around pharmaceutical company–sponsored economic analyses.^46- 47 Most of the studies in this review were not supported by the manufacturer. Although none of these studies reported cost-effectiveness or cost savings in the entire high-risk population, a few supported the use of these agents in specific subgroups. Our results reaffirm journal policies of promoting full disclosure of all financial interests to further the credibility of economic analyses.

Currently, the relatively high acquisition costs of the agents themselves, compared with the efficacy that they provide, argue for restricting use of these agents to very high-risk infant subgroups. It is clear that if the price of these agents were to be reduced in the future, their cost-effectiveness ratios would noticeably improve, encouraging more widespread use. Furthermore, if future studies were to demonstrate that use of these immunoprophylactic agents reduces mortality from RSV, their cost-effectiveness would potentially improve. Similarly, if the link between avoidance of RSV infection in infancy and reduction in the incidence of reactive airway disease were to be established, the cost-effectiveness of these agents would increase considerably. Future economic studies should incorporate any new evidence that becomes available from ongoing research in this area.⁴⁸ New agents to be introduced in the future should also compare their cost-effectiveness with that of the existing agents (RSV-IGIV and palivizumab) to identify whether the agents are superior in their cost-effectiveness in various infant subgroups. Nevertheless, comparison with the option of no prophylaxis is also important to identify subgroups of infants for which use of neither agent is justified. A recent cost-effectiveness analysis³⁶ of an interdisciplinary RSV infection control program using staff education and other precautionary measures found the program to be highly cost-effective in reducing the rate of RSV nosocomial infections in a hospital. Although the intervention program was not compared with prophylaxis options, the study results underscore the importance of also promoting practical prevention methods, such as avoidance of passive cigarette smoke and frequent hand washing by caretakers, for RSV prevention in low-risk infants and at-risk children outside the hospital.

In light of the issues identified in this review, providers, payers, and health policymakers need to critically appraise and judiciously interpret all cost-effectiveness research on these agents to determine whether their use is cost-effective in their overall infant population or specific subpopulations. However, unlike costs per life-year saved,^49- 51 there are no published standards or benchmarks on what is considered to be an acceptable cost to prevent 1 hospitalization. That decision remains a value judgment and will reflect consideration of many factors based on the perspective of the decision maker.

Accepted for publication May 30, 2002.

Despite the demonstrated benefits of treatment with RSV-IGIV and palivizumab for prevention of RSV in high-risk infants, the costs associated with these immunoprophylactic agents are significant. There is considerable debate on the relative value of these agents in terms of their costs and benefits. When the American Academy of Pediatrics first published, in 1998, recommendations for the use of palivizumab and an update on RSV-IGIV use, few cost-effectiveness analyses had been conducted on RSV-IGIV, and no published economic data were yet available on palivizumab. Since then, several economic studies have been published evaluating RSV-IGIV and palivizumab. In this era of cost-consciousness, health care decision making is increasingly being supplemented by economic data. However, use of a single published study on the economic impact of an intervention may be misleading. Reviews and critical evaluations of all currently available economic studies are essential to aid providers, payers, and health policy makers in this task.

This study reviews and critically evaluates all currently available economic data on these 2 agents. We report high variability in the studies and identify the sources of this variation and other methodological issues. Last, we make several recommendations to be considered in future studies to ensure reliable use of their results by decision makers.

Corresponding author: Sachin Kamal-Bahl, MS, Department of Pharmaceutical Health Services Research, University of Maryland School of Pharmacy, 506 W Fayette St, Suite 113, Baltimore, MD 21201 (e-mail: skama001@umaryland.edu).