TITLE: Antiviral Stockpiling Strategies for Pandemic Influenza: A Review of the Clinical and Cost-Effectiveness

Transcription

1 TITLE: Antiviral Stockpiling Strategies for Pandemic Influenza: A Review of the Clinical and Cost-Effectiveness DATE: 21 January 2011 CONTEXT AND POLICY ISSUES Influenza A viruses can intermittently cause worldwide pandemics with high rates of illness and death. A pandemic is possible at any time with the potential to cause serious illness, death, and substantial social and economic disruption globally. 1 While the characteristics such as the strain of influenza, clinical attack rate (CAR), fatality rate (FR), and hospitalization rate (HR), and duration of a future influenza pandemics are unpredictable, the World Health Organization (WHO) has recommended countries develop pandemic preparedness plans. 2 Part of a pandemic preparedness plan includes the stockpiling of antiviral (AV) agents. 3 The WHO has advised countries with adequate resources to stockpile AV drugs nationally for use at the start of a pandemic. 4 Vaccines could also provide protection against influenza viruses. However, antivirals will be the only specific medical intervention available during the initial pandemic response. 1 A vaccine will take at least four to six months to become available. 1 In Canada, M2 inhibitors (amantadine) and the neuraminidase inhibitors (oseltamivir and zanamivir) are the AV agents available for the prophylaxis and treatment of influenza. Both types of antiviral agents are 70% to 90% effective as prophylaxis and can shorten the duration of illness by 1.5 days when used in treatment. 5 Amantadine, however, is not active against influenza H5N1. 6 In addition, amantadine use leads to the predictable emergence of antiviral resistance when it is used to treat individuals who are in close contact with those on it for prophylaxis. 5 Multiple studies have indicated there is a high rate of amantadine resistance. 3,5,7 The neuraminidase inhibitors available in Canada include oseltamivir (Tamiflu, Roche), which is administered orally, and zanamivir (Relenza, GlaxoSmithKline), which is administered via inhalation Oseltamivir is not approved for use in children under the age of one 11 and zanamivir is not approved for use in children less than seven years old. 12 This report reviews the evidence for the cost-effectiveness of AV stockpiling strategies for pandemic influenza. Disclaimer: The Health Technology Inquiry Service (HTIS) is an information service for those involved in planning and providing health care in Canada. HTIS responses are based on a limited literature search and are not comprehensive, systematic reviews. The intent is to provide a list of sources and a summary of the best evidence on the topic that CADTH could identify using all reasonable efforts within the time allowed. HTIS responses should be considered along with other types of information and health care considerations. The information included in this response is not intended to replace professional medical advice, nor should it be construed as a recommendation for or against the use of a particular health technology. Readers are also cautioned that a lack of good quality evidence does not necessarily mean a lack of effectiveness particularly in the case of new and emerging health technologies, for which little information can be found, but which may in future prove to be effective. While CADTH has taken care in the preparation of the report to ensure that its contents are accurate, complete and up to date, CADTH does not make any guarantee to that effect. CADTH is not liable for any loss or damages resulting from use of the information in the report. Copyright: This report contains CADTH copyright material. It may be copied and used for non-commercial purposes, provided that attribution is given to CADTH. Links: This report may contain links to other information available on the websites of third parties on the Internet. CADTH does not have control over the content of such sites. Use of third party sites is governed by the owners own terms and conditions.

2 RESEARCH QUESTIONS 1. What is the comparative clinical effectiveness and safety of antiviral stockpiling strategies for pandemic influenza? 2. What is the cost-effectiveness and safety of antiviral stockpiling strategies for pandemic influenza? KEY MESSAGE The literature search did not identify any studies that assessed the clinical effectiveness and safety of antiviral stockpiling strategies for pandemic influenza. Evidence from the included economic studies suggested that stockpiling of antivirals for pandemic influenza may be costeffective, assuming that the next influenza pandemic is severe and has similar characteristics as the simulation models used in the included studies. METHODS A limited literature search was conducted on key health technology assessment resources, including PubMed, Ovid EMBASE, The Cochrane Library (Issue 12, 2010), University of York Centre for Reviews and Dissemination (CRD) databases, ECRI (Health Devices Gold), EuroScan, international health technology agencies, and a focused Internet search. The search was limited to English language articles published between January 1, 2000 and December 13, No filters were applied to limit the retrieval by study type. Article selection One reviewer (KQ) screened the titles and abstracts of the retrieved publications and evaluated the full-text publications for final article selection (Table 1). Table 1: Selection criteria for included studies Population All patients of all ages with or exposed to pandemic influenza Intervention Antiviral stockpiling strategies Comparator None Outcomes Clinical benefits and harm, safety and cost effectiveness Study design Health technology assessments, systematic reviews, meta-analyses, randomized controlled trials, non-randomized studies and economic evaluations Quality Assessment The methodological quality of the economic evaluations was assessed using the Drummond Checklist (Appendix 1). SUMMARY OF FINDINGS Antiviral Stockpiling for Pandemic Influenza 2

3 The literature search identified 262 citations (See Appendix 2 for flow chart). From these, 32 potentially relevant articles were retrieved for further examination and two additional articles were identified from hand searching. From these 34 citations, six relevant economic evaluations based upon simulation models were identified. Of the six economic evaluations identified one study was identified that based the model on the population of the following countries: Israel, 13 the United States of America (USA), 14 Australia, 15 the United Kingdom(UK), 16 the Netherlands, 17 and Singapore. 18 Details of the included economic evaluations can be found in Appendix 3. No relevant health technology assessments, systematic reviews and meta-analyses, randomized controlled trials or nonrandomized studies that examined the clinical effectiveness or safety of AV stockpiling strategies for pandemic influenza were identified. Economic evaluations Newall et al. (2010) 15 compared the cost-effectiveness of stockpiling vaccine and AV drugs to mitigate pandemic influenza for the Australian population in a hybrid transmission and decision analytic economic model. Four stockpiling strategies were compared: minimal pharmaceutical intervention, unspecified AV treatment only for the clinically infected, population prepandemic vaccination, and AV treatment combined with prepandemic vaccination. In all strategies, an initial AV containment effort was used for prophylaxis of case contacts and treatment of clinical cases and, after a delay of six months, a match vaccine was distributed to the population. It is assumed that this strategy is the minimal pharmaceutical intervention. The authors used an age-stratified transmission model to calculate CARs and AV drug consumption, then imputed the information into a decision analytic economic model. 15 The authors assumed a small AV drug stockpile (0.2 million courses) to be used in initial containment efforts and prophylaxis of case contacts relative to the estimated Australian population (21,020,222). In addition, it was assumed the efficacy of AV treatment for preventing hospitalization was 59%. 15 All costestimates were made in 2005 Australian Dollars (AUD). Rates of hospitalization and death were estimated relative to the CAR using case- HR and case fatality rate (CFR). Age-specific rates were calculated for three age groups 0 to 19, 20 to 64, and 65 years of age or older. Estimates for the HR ranged from 1.875% to 5.0% and for the CFR from 0.75% to 2.0%. The authors assumed that the stockpile purchase price per AV course was $32 AUD, including storage cost. Costs were discounted by 5% annually. It was assumed AV drugs must be replenished every five years. Results suggested that the AV drug treatment strategy would significantly reduce the number of hospitalizations by 42% and deaths by 42% compared to minimal pharmaceutical intervention. Cost of AV treatment decreased by $6.30 AUD per person compared to minimal pharmaceutical intervention. The incremental cost-effectiveness ratio (ICER) for the AV treatment strategy was $909 per life year saved (LYS). Two sensitivity analyses were conducted, one from a healthcare perspective and one that was a probabilistic sensitivity analysis. The healthcare perspective indicated variation in most parameters did not affect the cost effectiveness of strategies relative to each other. However, when R 0 (basic reproductive number, it represents the number of secondary cases that a person with influenza would infect in a fully susceptible population) was decreased, vaccine efficacy was increased. Also, the percentage of AV available within 48 hours was decreased. As a result, prepandemic vaccination dominated AV treatment alone. Probabilistic sensitivity analyses reported in a scenario where the emergent pandemic strain would have a different subtype than those chosen to stockpile, at a willingness to pay of over Antiviral Stockpiling for Pandemic Influenza 3

4 $12,000 per LYS, AV drug treatment alone was the best strategy. It is important to note that the AV drug resistance rate was not justified. The authors concluded that under the assumption of a severe pandemic occurring in the near future, all the pharmaceutical based mitigation strategies including AV treatment for only the clinically infected was generally estimated to be costeffective. Sander et al. (2009) 14 compared the impact of 17 mitigation strategies including the AV oseltamivir for pandemic influenza on the United States population in a stochastic, individuallevel, microsimulation model that evaluated the cost-utility of alternative strategies from a societal perspective. All results are reported compared to no intervention. The strategies included household targeted AV prophylaxis (HTAP) from stockpiles for different percentages of the population: 25%, 50% and unlimited; full targeted prophylaxis (FTAP) from stockpiles for different percentages of the population: 25%, 50% and unlimited; and treating all cases with oseltamivir. The assumed efficacy of AVs for symptomatic disease given exposure was 0.72 and for infectiousness was 0.62 based on a previous study. Quality adjusted life years (QALYs) were reported based on quality weights estimated from 0 (dead) and 1 (perfect health). Future life years were discounted at 3% per annum. In the base case analysis, the authors assumed a death rate of 2.5% per influenza case. In addition, they assumed that the FTAP cost was three times the amount of HTAP. All interventions reduced the illness attack rate (IAR) and, as a result, morbidity and DR. Full targeted antiviral prophylaxis was the most effective single strategy decreasing the IAR by 54%. Full targeted prophylaxis with a stockpile for 25% of the population decreased the IAR by 4%, reduced DR by 1 per 1000, led to an incremental gain of QALYs 16 per 1000, required 246 courses per 1000 and decreased costs by $10,000 per Full targeted prophylaxis with a stockpile for 50% of the population decreased IAR 10%, reduced the DR by 2 per 1000, had an incremental gain of 34 QALYs per 1000 population, and required 481 courses per 1000, with a decreased cost of $10,000 per 1000 population. Full targeted antiviral prophylaxis with an unlimited stockpile had the greatest effect as it decreased the IAR by 54%, reduced the DR by 8 per 1000, led to an incremental gain of 210 QALYs per 1000, required 2,447 courses per 1000 and decreased cost of $ 70,000 per Treating all cases with AVs reduced the IAR by 2%, reduced the DR by 5 per 1000, led to an incremental gain of 100 QALYs per 1000, required 243 courses per 1000 and led to no change in costs. Household targeted prophylaxis with a stockpile for 25% of the population reduced the IAR by 4%, reduced the DR by 2 per 1000, had an incremental gain of 40 QALYs per 1000, required 250 courses per 1000, and led to no change in costs. Household targeted prophylaxis with a stockpile for 50% of the population reduced the IAR by 16%, decreased the DR by 5 per1000, had an incremental gain of 98 QALYs per 1000, required 498 courses per 1000 and reduced the cost by $20,000 per HTAP with an unlimited stockpile led to an 18% IAR, decreased DR by 6 per 1000, 123 incremental QALYs per 1000, required 651 courses per 1000, and had a decreased cost of $20,000 per Generally, setting the R 0 (for definition see Newall et al above) value at 2.0 or changing the DR did not change the ranking of strategies. Full targeted prophylaxis remained the most effective and least costly, although with higher attack rates FTAP was more effective and with lower attack rates, targeted strategies provided similar effects, at lower costs. The authors concluded that the base-case analysis demonstrated that among the AV interventions, FTAP effectively reduced the burden of pandemic influenza compared to no intervention and was cost-saving from a societal perspective in terms both healthcare costs and productivity losses. The estimated direct costs and indirect costs related to medical treatment in this scenario amounted to a projected $59 billion. 14 Antiviral Stockpiling for Pandemic Influenza 4

5 Lugner et al. (2009) 17 used the outcomes from a dynamic transmission model to develop a costeffectiveness analysis model that compared AV stockpiling to no intervention. The model was set for a 30 year time horizon. The authors assumed there were two options for AV stockpiling: 1) bulk powder oseltamivir with a shelf-life of five to ten years or 2) a combination of two-thirds bulk powder oseltamivir and one-third of the manufactured product (Tamiflu). Tamiflu tablets have a shelf life of five years. Healthcare utilization, sick leave and deaths were estimated for both the intervention and non-intervention scenarios and compared in incremental costeffectiveness, based on a dynamic transmission model. Lost productivity costs were estimated using a friction-costing method. Costs for renewal of the stock were included for the total period irrespective of the occurrence of an outbreak. The authors indicated the convention in the Netherlands for an acceptable cost-effectiveness ratio was 20,000/LYG. Based on a Dutch population of 16.6 million, there would be 10.4 million infected individuals if a pandemic influenza outbreak was uncontrolled. Stockpiling of oseltamivir powder alone was cost-effective if the risk of a pandemic influenza outbreak was perceived to be larger than 23% in the next 30 years. If the stockpile was the combination bulk powder oseltamivir and Tamiflu, then the perceived risk would have to be at least 29%. With the inclusion of productivity losses, the risk would have to be larger than 9% for the oseltamivir powder alone and 11% for the combination alternative. The cost-effectiveness ratio for the combination alternative at the observed risk would be around 15,000/LYG. In the sensitivity analyses, the authors varied the size of the pandemic, percentage of the population receiving AV therapy, CAR, QALYs (gained), pandemic outbreaks in 30 years (two), and cost of distribution and dispensing the drugs. For two scenarios the cost cutoff point was equal or higher than the observed risk if AV drugs were only given to 60% of the symptomatic individuals or if the CAR is as high as 50%(if the stockpile is assumed to treat all symptomatic cases). The authors concluded that cost-effectiveness depends on the risk that there will be an outbreak of pandemic influenza. The analysis indicated that the costeffectiveness ratio of stockpiling a combination of AV drugs falls below the cost-effectiveness cut-off point if the risk of pandemic influenza is about 11%. This study is limited by the assumption that 80% of the population will receive the drug on time (within 48 hours of symptom onset) and the model assumed AV sensitivity will continue to be 100% with no AV resistance. Siddiqui et al. (2008) 16 created a decision analytical model to investigate the cost-effectiveness of stockpiling AV drugs from the perspective of the healthcare provider for the potential of an influenza pandemic in the United Kingdom and the role of near patient testing (rapid diagnostic tests at the point of care) to conserve AV stockpiles. The authors examined three potential strategies for the management of patients with influenza like illness (ILI) with the stockpiled AVs: no intervention, treat all patients with AV drugs, and test then treat only those who test positive for influenza with AV drugs. Baseline scenarios assumed a cumulative CAR of 25% over a wave lasting 15 weeks. The authors examined two fatality scenarios on the basis of the 1918 influenza pandemic (2.3% mortality), and a combined 1957 and 1968/69 influenza pandemic (0.3% mortality). For the base-case analysis the authors assumed the next pandemic would take place within 30 years. Oseltamivir was the AV treatment selected. The authors assumed that for both the treat-only and test-treat strategies 87% of patients with ILI were given AV drugs and were tested until the stockpile ran out. The shelf life of oseltamivir was five years. The base case-analysis assumed the AV drugs stockpile was 14.6 million courses. The cost of AV treatment course was assumed to be 16 for the treat-only option and for test-treat option. The unit cost of a near-patient test was assumed to be 7 and storage of the AV treatment was 1 per unit. All costs were in 2004 pounds sterling. Results indicated under the 1918 scenario there would be about 344,000 deaths without intervention, with the treat only program there would be 236,000 deaths, and with the test-treat program 231,000. The discounted National Health Service (NHS) cost would be 113 million Antiviral Stockpiling for Pandemic Influenza 5

6 without intervention, 1,361 million treat only, and 2,356 million test-treat for the 1918 scenario. Incremental cost per QALY was 1,861 for treat only, and 31,031 for test-treat. For the combined 1957 and 1968/69 scenario, no intervention there would be 44,000 deaths, with the treat only scenario there would be 30,000 deaths, and with the test-treat only scenario there would be 30,000 deaths. Discounted NHS costs were identical to those in the 1918 scenario. However, discounted QALY loss differed million without intervention, million treat only, and test-treat. Incremental cost per QALY was 13,668 treat only, and 227,896 test-treat. A sensitivity analysis conducted on the test-treat program varying the parameters: AV drug efficacy for reducing complications and hospitalizations, timing of pandemic, discount rate, probability of receiving AVs, wastage of AV drugs, and CAR was unlikely to change the result that the test-treat strategy was cost-effective. The authors concluded that stockpiling AV drugs for a treat-only program is likely to be a cost-effective strategy in preparation for a potential influenza pandemic, even if the pandemic occurs many years down the line, under the assumption that AV drugs will reduce the death toll. This study is limited due to the examination of a mass AV strategy. This may not be practical during a pandemic so it could have been useful to include other scenarios that targeted different segments of the population in the model. Lee et al. (2006) 18 compared strategies for stockpiling oseltamivir to treat and prevent influenza in Singapore from the prospective of national planners. The authors conducted Monte Carlo simulations using a decision-based model to perform cost-benefit and cost-effectiveness analyses. The model compared three strategies for use of the stockpiled AVs: no action, early treatment of influenza, and prophylaxis in addition to early treatment. Cost-benefit analyses included direct and indirect economic costs and compared the cost per life saved (LS) for treatment only and prophylaxis compared to no action. The population of Singapore (4,240,300) was divided into three age groups 19 years or younger, 20 to 64, and 65 years of age or older. In addition, the population was subdivided into high and low risk levels. It was assumed that the CAR was 30%, the pandemic duration was assumed to be 12 weeks, costs from the side effects of oseltamivir were negligible, immunity after prophylaxis was 35%, and a shelf-life of oseltamivir of four years. Results indicated if no action was taken there would be 1,105 deaths and the mean economic cost would exceed $1.43 billion Singapore Dollars (SGD). All cost estimates were in 2004 SGD. The maximal economic benefit of stockpiling occurred at a 40% stockpile at a cost of $414 million with 418 lives saved. The cost-benefit and cost-effectiveness changes in the prophylaxis stockpile indicated treatment only, without prophylaxis provided the best economic benefit as it dominated all other approaches with 423 lives saved compared to no action, a cost per life saved of $380,000 and the benefit compared to no action was $379 million saved. Sensitivity analyses were conducted and the costs were most sensitivity to changes in attack rate and case-fatality rate; however, the overall results did not change. The authors concluded that the optimal treatment stockpile is 40% to 60%: 40% maximizes economic benefits and 60% maximizes treatment benefits. In addition, they noted prophylaxis was only economically beneficial compared with no action in subpopulations at high risk. The limitations of this study included the low shelf-life estimate for oseltamivir, assumption of pandemic flu potential appearing every 8 to 10 years (more frequent than other studies and historical estimates), and the disregard of intangible costs in their estimates (for example, societal value of health; indirect effects on national economy and trade). Balicer et al. (2005) 13 conducted a cost-benefit analysis of stockpiling oseltamivir for pandemic influenza for the Israeli population of all ages from the perspective of national pandemic planners. The authors calculated the direct costs to the healthcare system and the overall costs to the economy, including the value of lost workdays. Three strategies for AV use were compared therapeutic use, long-term prexposure prophylaxis use (50 days), and short-term Antiviral Stockpiling for Pandemic Influenza 6

7 postexposure prophylaxis (seven days) using AV stockpiles of oseltamivir powder. The first two strategies could be used for either the entire population or only those at high risk for complications. As a result, the authors compared five strategies with no intervention. It was assumed bulk active powder oseltamivir would have a 10 year shelf-life. The base-case assumptions were for a CAR of 25%, the efficacy of prexposure prophylaxis was 71%, and efficacy of postexposure prophylaxis was 36%. The economic benefit of each strategy was calculated by multiplying each of the reductions in adverse outcomes by the estimated economic value. The cost of each strategy was calculated by multiplying the estimated number of treated person by the discounted cost of a single antiviral course. Results indicated a pandemic would lead to about 1,618,200 cases of influenza, 10,334 hospitalizations, and 2,855 deaths with no intervention. The outcomes would results in healthcare costs of $55.4 million and overall cost to the economy of $523.5 million. All costs were calculated in 2004 US Dollars. The cost-benefit ratios of AV usage varied considerably when compared to no intervention as a reference group. Therapeutic use to all patients yielded a cost benefit ratio of 2.44 to the economy, therapeutic use to patients at high risk was associated with a cost benefit ratio of 3.68 to the economy. Postexposure short-term prophylaxis for all close contact yielded a cost benefit ratio of 2.49 to the economy. Other interventions were not shown to be cost saving (preexposure long-term prophylaxis for the entire population and high risk only). Multivariate sensitivity analyses found the model to be robust. The authors concluded stockpiling is a wise investment that may help mitigate this impending global threat of pandemic influenza. Limitations There are multiple limitations to the economic evaluations included in this rapid review. Because there have not been influenza pandemics for which stockpiling strategies have been used, the literature search did not identify studies based completely upon real-world outbreaks with usage of stockpiles. As a result, all the studies selected for inclusion were simulation studies or models. Each simulation study or model made multiple assumptions about factors for which there could be a large degree of uncertainty such as morbidity, mortality, CAR, feasibility, efficacy of AVs, availability of AVs, AV resistance, compliance, shelf-life of medications, time to the next influenza pandemic and projected costs. Some of these assumptions may be unrealistic. For example, one simulation study assumed 87% of the population would receive AV drugs from a stockpile during a pandemic. 16 In addition, it is not possible to predict the efficacy of AV drugs against the next strain of pandemic influenza. The simulation models in the selected studies predict the next influenza pandemic could occur anywhere from eight to fifty years in the future. At this stage, the wide variability among the estimated parameters appears to be a major limitation to these studies. The size of stockpiles among the economic evaluation models varied from 1% to an unlimited stockpile. The variability in the size of the stockpile available applied to the models make it difficult to compare models. Further, differences in costing years, model inputs and perspectives make it difficult to compare the estimates of cost-effectiveness across studies. A related concern stems from the fact none of the included economic evaluations calculated the size of stockpile and distribution of AV drugs would be needed to contain an outbreak. Prophylaxis to contain an outbreak is a practical consideration that may be a factor during a pandemic. None of the economic evaluations included a cost for lives lost. However, there were no projections based on long-term lost productivity to a nation s economy. This may be a more Antiviral Stockpiling for Pandemic Influenza 7

8 relevant question for a nation evaluating the benefit of stockpiling AVs. It may be difficult to quantify this, but consideration of such factors may increase the generalizability of the models. None of the selected economic evaluations were conducted using the Canadian population as a reference. Although five of the six selected economic evaluations had the populations of developed nations as the reference, it might not be possible to assume similar demographic characteristics and case contacts given the composition and dispersion of the Canadian population. As a result, it is unclear whether the results of the economic evaluations can be generalized to the Canadian population. Further differences in the structure of healthcare systems between countries could limit the generalizability of the included economic evaluations to Canada. A formal quality assessment of the economic evaluations was conducted using the Drummond Checklist. From this critique, it was noted that one study provided no time horizon for cost benefits stated. In addition, only one study accounted for productivity changes in the model. Additional information on the results of the Drummond Checklist can be found in Appendix 3. Knowledge Gaps The included models did not consider estimates of the possible cost of side-effects from AV treatment and prophylaxis. Given the progress in medical treatments, an assessment of the accuracy of using case fatality rates based on 1918 data or 1957 and 1968/1969 needs to be investigated. Comparisons of stockpiling different AV drugs (for example, oseltamivir compared with zanamivir) to examine benefit and harms were not identified. Studies might not be generalizable to a future influenza pandemic in Canada. The length of time necessary to acquire a stockpile from manufacturers at a level that would be clinically or cost-effective was not available from the identified literature. The included models did not account for changes in drug prices over time due to factors such as a loss of patent. The cost-effectiveness, benefits and harms of stockpiling zanamivir was not clear from the included studies. There were no studies identified that evaluated the comparative clinical effectiveness and harms of different strategies for stock-piling AVs. CONCLUSIONS AND IMPLICATIONS FOR DECISION OR POLICY MAKING: Results from the identified studies may help assess the theoretical cost-effectiveness of stockpiling AVs for the Canadian population. Unfortunately, the generalizability of these studies may be limited because they were not modeled based on the Canadian population. Furthermore, none of the included studies identified an instance of pandemic influenza in history when a stockpile of AV drugs was available. There are multiple factors that need to be considered along with the value of stockpiling AVs such as availability of the drugs at the time of a pandemic, timely distribution, and compliance of the population. It is difficult to predict the severity of the impending influenza pandemic. As well, there is uncertainty as to what impact the next influenza pandemic would have without intervention. Overall, models of an influenza pandemic indicated that stockpiling antivirals is a cost-effective strategy, especially if the impending influenza pandemic will be severe in terms of morbidity and mortality. However, it Antiviral Stockpiling for Pandemic Influenza 8

9 should be noted that the actual cost-effectiveness of stockpiling would be dependent on the unknown characteristics of the next influenza pandemic. As no studies were identified that evaluated the comparative clinical effectiveness and harms of different strategies for stockpiling AVs, no conclusions can be made in this regard. PREPARED BY: Canadian Agency for Drugs and Technologies in Health Tel: Antiviral Stockpiling for Pandemic Influenza 9

10 REFERENCES: 1. Public Health Agency of Canada. The Canadian pandemic influenza plan for the health sector [Internet]. Ottawa: Public Health Agency of Canada; 2006 Dec. [cited 2011 Jan 18]. Available from: 2. Gani R, Hughes H, Fleming D, Griffin T, Medlock J, Leach S. Potential impact of antiviral drug use during influenza pandemic. Emerg Infect Dis [Internet] Sep [cited 2010 Dec 13];11(9): Available from: 3. Schneider RB, Benitez JG, d'angelo A, Tyo K. Pandemic influenza: antiviral preparedness and health care workers. Disaster Med Public Health Prep Mar;4(1): Avian influenza frequently asked questions [Internet]. Geneva: World Health Organization; 2005 Dec 5. [cited 2010 Dec 13]. Available from: 5. Monto AS. The role of antivirals in the control of influenza. Vaccine May 1;21(16): Kamps BS, Hoffmann C. Amantadine [Internet]. Paris: Flying Publisher; [cited 2010 Dec 13]. (Influenza Report). Available from: 7. Harrington JE, Jr., Hsu EB. Stockpiling anti-viral drugs for a pandemic: the role of Manufacturer Reserve Programs. J Health Econ May;29(3): Canadian Pharmacists' Association. ecps [Internet]. Ottawa: Canadian Pharmacists' Association; Amantadine: antiviral-antiparkinsonian agent; 2009 Oct [cited 2010 Dec 13]. Available from: Subscription required. 9. Roche. ecps [Internet]. Ottawa: Canadian Pharmacists' Association; Tamiflu oseltamivir phosphate: antiviral agent; 2009 Nov 9 [cited 2010 Dec 13]. Available from: Subscription required. 10. GlaxoSmithKline. ecps [Internet]. Ottawa: Canadian Pharmacists' Association; Relenza zanamivir: antiviral agent; 2009 Nov 12 [cited 2010 Dec 13]. Available from: Subscription required. 11. Guidance for expanded use of oseltamivir (Tamiflu ) in children under one year of age in the context of Pandemic (H1N1) 2009: Important notice [Internet]. Ottawa: Public Health Agency of Canada; 2009 Dec 23. [cited 2010 Dec 13]. Available from: National guidelines for the use of antivirals for control of facility* outbreaks caused by influenza A (H1N1), season [Internet]. Ottawa: Public Health Agency of Canada; 2009 Apr 28. [cited 2010 Dec 13]. Available from: Antiviral Stockpiling for Pandemic Influenza 10

11 13. Balicer RD, Huerta M, Davidovitch N, Grotto I. Cost-benefit of stockpiling drugs for influenza pandemic. Emerg Infect Dis [Internet] Aug [cited 2010 Dec 13];11(8): Available from: Sander B, Nizam A, Garrison LP, Jr., Postma MJ, Halloran ME, Longini IM. Economic evaluation of influenza pandemic mitigation strategies in the United States using a stochastic microsimulation transmission model. Value Health Mar;12(2): Newall AT, Wood JG, Oudin N, MacIntyre CR. Cost-effectiveness of pharmaceuticalbased pandemic influenza mitigation strategies. Emerg Infect Dis [Internet] Feb [cited 2010 Dec 13];16(2): Available from: Siddiqui MR, Edmunds WJ. Cost-effectiveness of antiviral stockpiling and near-patient testing for potential influenza pandemic. Emerg Infect Dis [Internet] Feb [cited 2010 Dec 13];14(2): Available from: Lugner AK, Postma MJ. Investment decisions in influenza pandemic contingency planning: cost-effectiveness of stockpiling antiviral drugs. Eur J Public Health [Internet] Oct [cited 2010 Dec 13];19(5): Available from: Lee VJ, Kai HP, Chen MI, Chow A, Ma S, Kee TG, et al. Economics of neuraminidase inhibitor stockpiling for pandemic influenza, Singapore. Emerg Infect Dis. 2006;12(1): Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews of interventions [Internet]. Version Drummond. Oxford (U.K.): The Cochrane Collaboration; Figure 15.5.a: Drummond checklist (1996). [cited 2010 Dec 13]. Available from: cklist_drummond_1996.htm Antiviral Stockpiling for Pandemic Influenza 11

12 APPENDIX 1: Drummond Checklist Item Yes No Not clear Study design Not appropriate 1. The research question is stated. 2. The economic importance of the research question is stated. 3. The viewpoint(s) of the analysis are clearly stated and justified. 4. The rationale for choosing alternative programmes or interventions compared is stated. 5. The alternatives being compared are clearly described. 6. The form of economic evaluation used is stated. 7. The choice of form of economic evaluation is justified in relation to the questions addressed. Data collection 8. The source(s) of effectiveness estimates used are stated. 9. Details of the design and results of effectiveness study are given (if based on a single study). 10. Details of the methods of synthesis or meta-analysis of estimates are given (if based on a synthesis of a number of effectiveness studies). 11. The primary outcome measure(s) for the economic evaluation are clearly stated. 12. Methods to value benefits are stated. 13. Details of the subjects from whom valuations were obtained were given. 14. Productivity changes (if included) are reported separately. 15. The relevance of productivity changes to the study question is discussed. 16. Quantities of resource use are reported separately from their unit costs. 17. Methods for the estimation of quantities and unit costs are described. 18. Currency and price data are recorded. 19. Details of currency of price adjustments for inflation or currency conversion are given. 20. Details of any model used are given. 21. The choice of model used and the key parameters on which it is based are justified. Analysis and interpretation of results 22. Time horizon of costs and benefits is stated. 23. The discount rate(s) is stated. Antiviral Stockpiling for Pandemic Influenza 12

13 Item Yes No Not clear Not appropriate 24. The choice of discount rate(s) is justified. 25. An explanation is given if costs and benefits are not discounted. 26. Details of statistical tests and confidence intervals are given for stochastic data. 27. The approach to sensitivity analysis is given. 28. The choice of variables for sensitivity analysis is justified. 29. The ranges over which the variables are varied are justified. 30. Relevant alternatives are compared. 31. Incremental analysis is reported. 32. Major outcomes are presented in a disaggregated as well as aggregated form. 33. The answer to the study question is given. 34. Conclusions follow from the data reported. 35. Conclusions are accompanied by the appropriate caveats. Antiviral Stockpiling for Pandemic Influenza 13

14 APPENDIX 2: Selection of Publications 262 citations identified from electronic literature search and screened 230 citations excluded 32 potentially relevant articles retrieved for scrutiny (full text, if available) 2 potentially relevant reports retrieved from other sources (grey literature, hand search) 34 potentially relevant reports 28 reports excluded: - irrelevant outcomes (8) -irrelevant/unknown population (7) -irrelevant intervention (1) -narrative article/no data (5) -irrelevant intervention (3) -letters to the editor (2) -editorials (1) -recommendation/guideline (1) 6 reports included in review Antiviral Stockpiling for Pandemic Influenza 14

15 APPENDIX 3: Economic Evaluation Models of AV Stockpiling Strategies for Pandemic Influenza Name, Year, Objective, Country Newall et al.(2010) 15 Examine the costeffectiveness of stockpiling prepandemic vaccine and AV drugs to mitigate pandemic influenza Australia Methods (with assumptions) Hybrid age-stratified transmission and decision analytic economic model Assumptions: A small stockpile (0.2 million courses) AV prophylaxis treatment of contacts during the initial containment reduced susceptibility and infectiousness by 70% and 60% Efficacy of AV treatment for preventing hospitalization or death of 59% Patient Group Australian population divided into three age groups 0-19, 20-64, >65 years of age AV Stockpiling Strategies Unknown AV treatment of those clinically infected Base-case comparator: Minimum pharmaceutical intervention Clinical and Economic Outcomes/ Results Minimum pharmaceutical intervention Hospitalizations= 782.3/100,000 Deaths= 312.9/100,000 AV drug treatment Hospitalizations = 450/100,00 and Deaths =180/ 100,000 Total discounted healthcare costs for a pandemic in the absence of any intervention was $31.10/person. The cost of the AV treatment program over a five year period was $24.80/person. Quality Assessment: Drummond Checklist 19 The cost of productivity changes or losses was not reported separately 10% of viruses were resistant to AV drugs ICER= $909/LYS 50% of people with clinical influenza would seek medical attention Age-specific hospitalization, mortality 0 to 19= 1.875%, 0.75% 20 to 64= 2.5%, 1% > 65 = 5%, 2% Antiviral Stockpiling for Pandemic Influenza 15

16 Name, Year, Objective, Country Methods (with assumptions) Delay to pandemic 5 years Patient Group AV Stockpiling Strategies Clinical and Economic Outcomes/ Results Quality Assessment: Drummond Checklist 19 Initial AV containment effort and distribution of a matched vaccine $32 per AV course (including storage cost) Discount rate of 5% annually Shelf-life of AV drug 5 years Cost for general practitioner visits Cost attached to lost work days based on average weekly earnings 2.6 work days absent Sander et al. (2009) 14 Determine the cost utility of alternative pandemic influenza mitigation strategies in R 0 = 1.7 Stochastic individual level, microsimulation model Assumptions: AV efficacy for symptomatic disease 0.72, infectiousness 0.62 Stratified population by age and risk USA Population Oseltamivir HTAP25 HTAP50 HTAP FTAP25 FTAP50 No intervention IAR= 50% Deaths/1000=13 QALYs/1000= 21,141 Cost in million$ per 1000= 0.19 HTAP25 IAR= 48% Deaths/1000=11 Incremental QALY= 40 Few details on the older adults and children portion of the simulation No range or confidence intervals provided for estimated outcomes Major outcomes Antiviral Stockpiling for Pandemic Influenza 16

17 Name, Year, Objective, Country the United States taking a societal perspective Canada Methods (with assumptions) 0-4, 5-18, (high and low risk), > 65 years old Future life years discounted by 3% Death Rate of 2.5% 2.5 work days of loss per week Patient Group AV Stockpiling Strategies FTAP Treating all cases with AVs Clinical and Economic Outcomes/ Results Courses/1000= 250 Cost in million $ per 1000= 0.19 HTAP50 IAR=42% Deaths/1000=8 Incremental QALY= 98 Courses/1000=498 Quality Assessment: Drummond Checklist 19 presented as aggregated data only. Oseltamivir costs converted from Euro to US dollars according to Interbank rate as of July 5, 2006 R 0 ranged from 1.5 to 2.6 over 100 runs Cost in million$ per 1000= 0.17 HTAP IAR= 41% Deaths/1000=7 Incremental QALY=123 Courses/1000= 651 Cost in million$ per 1000= 0.17 FTAP25 IAR= 48% Deaths/1000= 12 Incremental QALY=16 Courses/1000=246 Cost in million$ per 1000=0.18 FTAP50 IAR=45% Deaths/1000=11 Incremental QALYs/1000=34 Courses/1000=481 Cost in million$ per 1000=0.18 Antiviral Stockpiling for Pandemic Influenza 17

18 Name, Year, Objective, Country Methods (with assumptions) Patient Group AV Stockpiling Strategies Clinical and Economic Outcomes/ Results Quality Assessment: Drummond Checklist 19 FTAP Lugner et al. (2009) 17 Determine if the stockpiling of AV drugs is cost-effective for therapeutic use in the Netherlands Netherlands Dynamic transmission model of cost-effectiveness Assumptions: 30 year time horizon CAR of 38% based on Asian Flu in 1957 Threshold for acceptable cost effectiveness of 20,000/LYG Population of 16.6 million Dutch population (based on 2007 estimate) Two stockpiling options: Bulk powder (oseltamvir alone) A combination of two thirds bulk powder and one third Tamiflu IAR= 23% Deaths/1000=5 Incremental QALYs/1000=210 Courses/1000=2,447 Cost in million$ per 1000= 0.12 No intervention Approximately 10.4 million infected individuals 22,941 Hospitalizations Deaths 9012 Life-years lost 96,795 Outpatient GP visits No information on the age and risk levels of the Dutch population used for the simulation No explanation of how discounting was conducted Authors do not directly answer the study objective Assume that AV drugs were distributed to 80% of the population Number of doses need to purchases based on estimates from national experts in the Netherlands in pandemic preparedness 33,240,996 OTC drugs and antibiotics due to complications 33,730,543 Hospitalizations 108,334,422 Production losses 2,521,537,242 Antiviral Stockpiling for Pandemic Influenza 18

19 Name, Year, Objective, Country Methods (with assumptions) Patient Group AV Stockpiling Strategies Clinical and Economic Outcomes/ Results Quality Assessment: Drummond Checklist 19 Intervention Infected individuals Approximately 8.6 million infected individuals Hospitalizations 13,851 Deaths 5362 Incremental LYG 38,883 Incremental cost Outpatient GP visits 16,711,863 Incremental cost OTC drugs and antibiotics 8,495,071 Incremental cost Hospitalizations 42,927,786 Incremental cost Production losses 1,885,227,269 Telephone calls to GP 44,077,556 Antiviral Stockpiling for Pandemic Influenza 19

20 Name, Year, Objective, Country Methods (with assumptions) Patient Group AV Stockpiling Strategies Clinical and Economic Outcomes/ Results Quality Assessment: Drummond Checklist 19 Pharmacy fee fro V prescriptions 25,867,972 Stockpiling costs ( 5 million doses): One time purchase, oseltamivir 45,736,179 One time purchase, combination Tamiflu and oseltamivir 56,185,269 Yearly storage costs 51, years stockpiling (present value) (PV): Purchase and storage, oseltamivir 143,658,011 Purchase and storage, combination 176,881,136 Average PV costs during pandemic 1,085,347 Average PV savings including productivity losses during pandemic 1,130,112,974 Average LYG Antiviral Stockpiling for Pandemic Influenza 20

21 Name, Year, Objective, Country Methods (with assumptions) Patient Group AV Stockpiling Strategies Clinical and Economic Outcomes/ Results Quality Assessment: Drummond Checklist 19 31,594 Average quality adjusted life years gained 48,540 The stockpiling of oseltamivir alone is cost effective if the risk of a pandemic influenza outbreak is perceived to be larger than 23% in the next 30 years. If the stockpiling is a combination of oseltamivir and Tamiflu then perceived risk cutoff point would be 29% Including production losses the risk would have to be larger than 9% for oseltamivir along and 11% for the combo alternative Cost effectiveness ratio for the combination alternative at the observed risk would be around 15,000 LYG Siddiqui et al. (2008) 16 Asses the costeffectiveness of stockpiling AV drugs for Decision analytic model Assumptions: Model based on historical pandemic influenza scenarios of 1918, 1957, and 1968/1969 United Kingdom Population 3 stockpiling strategies No treatment Treat all patients with AV drugs 1918 scenario: No intervention =344,000 deaths Treat only program = 236,000 deaths, and Test-treat program = 231,000 deaths Unclear from the characteristics of the United Kingdom population adapted for the model Estimate oseltamivir efficacy Antiviral Stockpiling for Pandemic Influenza 21

22 Name, Year, Objective, Country a possible influenza pandemic and the use of nearpatient testing in management of AV drugs United Kingdom Methods (with assumptions) 30 year time horizon CAR of 25% over 1 wave lasting 15 weeks 15 million cases of influenza in all scenarios Fatality scenarios 1918= 2.3% 1957 or 1968/69= 0.3% Future cost and benefits discounted at 3.5% per year Oseltamivir shelf-life of 5 years Drug stockpile of 14.6 million courses Patient Group AV Stockpiling Strategies Test then treat those who test positive for influenza with AV drugs Clinical and Economic Outcomes/ Results Discounted NHS cost of no intervention = 113 million, Treat-only= 1,361 million Test-treat= 2,356 Discounted QALY loss No intervention= 2.23 million, Treat only= 1.56 million, Test-treat= 1.53 million. Incremental cost per QALY Treat only= 1,861 for treat only, Test-treat= 31, and 1968/69 scenario: No intervention =44,000 deaths, Quality Assessment: Drummond Checklist 19 based on one study. Estimate of efficacy never explicitly stated Cost of AV treatment courses 16 for the treat-only option and for test-treat option Treat only scenario= 30,000 deaths, Test-treat scenario = 30,000 deaths. The unit cost of near-patient test was assumed to be 7 and storage was 1 per unit 87% of patients with ILI are tested and receive AV drugs Discounted NHS costs = 1918 scenario. Discounted QALY loss No intervention= million, Treat only= million, Test-treat= Incremental cost per QALY Treat only= 13,668 Test-treat= Antiviral Stockpiling for Pandemic Influenza 22

23 Name, Year, Objective, Country Lee et al. (2006) 18 Compare economic outcomes of prophylactic treatment with AV drugs to assist national planners with optimal strategies Singapore Methods (with assumptions) Decision based model Assumptions: 3 age groups- 0-19,20-64,>65 years of age Risk groups (for each category)- low, high risk CAR= 30% Pandemic duration= 12 weeks Case fatality rate (based on age groups/ 100,000) Range Hospitalization rate= Patient Group Singapore population from ,240,300 AV Stockpiling Strategies Three strategies; No action Early treatment of influenza with oseltamivir Prophylaxis in addition to early treatment with oseltamivir Clinical and Economic Outcomes/ Results 227,896 No action Deaths= 1,105 Mean economic cost >$1.43 billion SGD. Maximal economic benefit of stockpiling at a 40% stockpile: Economic benefit= $414 million SGD Lives saved =418 The cost-benefit and costeffectiveness changes in the prophylaxis stockpile indicated treatment only, without prophylaxis provided the best economic benefit as it dominated all other approaches. Lives saved= 423 Cost/ life saved= $380,000 SGD and Quality Assessment: Drummond Checklist 19 Did not include cost of productivity changes No discount rate or discounting done Immunity after prophylaxis= 35% Economic benefit= $379 million SGD. Oseltamivir shelf-life = 4 years Balicer et al. (2005) 13 Identify strategies for the use of Potential pandemic every 8-10 years Health related impact estimated from previous influenza pandemics Assumptions: Israeli population Five strategies: Therapeutic use Long-term Health related costs > $55.4 million USD Cost to the economy= $523.5 million Cost-benefit ratio No information on the Israeli population provided No justification for variables in Antiviral Stockpiling for Pandemic Influenza 23