关 键 词：

老年人 使用 流感疫苗 成本 效益 外文 文献 英文

资源描述：

Cost effectiveness of quadrivalent versus trivalent infl uenza vaccine for elderly population in China Minghuan Jiang a b c Pengchao Lia b c Weihua Wangd Mingyue Zhaoa b c Naveel Atifa b c Shan Zhua b c Yu Fang a b c aDepartment of Pharmacy Administration and Clinical Pharmacy School of Pharmacy Xi an Jiaotong University Xi an 710061 China bCenter for Drug Safety and Policy Research Xi an Jiaotong University Xi an 710061 China cShaanxi Center for Health Reform and Development Research Xi an 710061 China dDepartment of Non communicable Chronic Disease Control and Prevention Shaanxi Provincial Center for Disease Control and Prevention Xi an 710054 China a r t i c l ei n f o Article history Received 6 June 2019 Received in revised form 13 November 2019 Accepted 18 November 2019 Available online 29 November 2019 Keywords Quadrivalent infl uenza vaccine Trivalent infl uenza vaccine Cost effectiveness Elderly China a b s t r a c t Background Infl uenza associated excess death occurred most in the elderly We aimed to assess the cost effectiveness of quadrivalent infl uenza vaccine QIV versus trivalent infl uenza vaccine TIV for prevention of infl uenza infection among elderly population in China Methods A decision analytic model was developed to compare 1 year clinical and economic outcomes of three infl uenza vaccination options no vaccination TIV and QIV in a hypothetical cohort of Chinese elderly aged 69 years Outcome measures included cost infl uenza infection rate infl uenza related mor tality rate quality adjusted life years QALY loss and incremental cost effectiveness ratio ICER from societal perspective Sensitivity analyses were performed to examine the uncertainty of model inputs Results Base case results showed no vaccination was dominated more costly at higher QALY loss by TIV and QIV QIV was more costly USD56 29 versus USD54 28 with lower infl uenza infection rate 0 608 versus 0 623 mortality rate 0 00199 versus 0 00204 and QALY loss 0 01213 versus 0 01243 than TIV QIV was cost effective compared to TIV with ICER of 6 700 USD QALY below the willingness to pay threshold 29 580 USD QALY One way sensitivity analysis found the cost effectiveness of QIV was subject to the relative risk of vaccine effectiveness of QIV versus TIV and TIV would be cost effective if the relative risk was below 1 05 In 10 000 Monte Carlo simulations the probabilities of QIV TIV and no vaccination to be cost effective were 86 3 13 7 and 0 respectively Conclusion QIV appears to be a cost effective option compared to TIV and no infl uenza vaccination for elderly population in China 2019 Elsevier Ltd All rights reserved 1 Introduction Infl uenza pandemic was one of the ten threats to global health as unveiled by the World Health Organization WHO in 2019 1 Infl uenza is an acute respiratory infectious disease caused by infl uenza virus and results in substantial mortality especially among elderly population 2 It was estimated that the global infl uenza related respiratory excess mortality rate ranged from 2 9 to 233 5 per 100 000 individuals 3 Infl uenza vaccine is the most effective way to prevent infl uenza infection and avert infl uenza related complications and medical expenditures 4 In China trivalent infl uenza vaccine TIV and quadrivalent infl uenza vaccine QIV are both available since 2018 TIV contains two strains of infl uenza A viruses H1N1 and H3N2 and one of the two lineages of infl uenza B virus Yamagata or Victoria In contrast QIV consists of four infl uenza viruses A H1N1 A H3N2 B Yamagata and B Victoria QIV had similar effi cacy against the two strains of infl uenza A viruses and one lineage of infl uenza B virus included in TIV yet superior immunogenicity against the non TIV B lineage 5 https doi org 10 1016 j vaccine 2019 11 045 0264 410X 2019 Elsevier Ltd All rights reserved Abbreviations QIV quadrivalent infl uenza vaccine TIV trivalent infl uenza vaccine QALY quality adjusted life years ICER incremental cost effectiveness ratio WHO World Health Organization GDP gross domestic product CI confi dence interval WTP willingness to pay Corresponding author Department of Pharmacy Administration and Clinical Pharmacy School of Pharmacy Xi an Jiaotong University Xi an 710061 China E mail addresses jiangmh2017 M Jiang lipengchao1996 stu P Li 495928425 W Wang mingyue0204 M Zhao naveel atif87 N Atif zoey0323 S Zhu yufang Y Fang Vaccine 38 2020 1057 1064 Contents lists available at ScienceDirect Vaccine journal homepage China is stepping into an aging society with elderly aged over 60 years accounting for 17 9 230 86 million of overall popula tion 6 and 86 of annual infl uenza associated excess death occurred among elderly 65 years 7 The Technical Guidelines for Seasonal Infl uenza Vaccination in China 2018 2019 recom mended elderly aged over 60 years to be one of the priority groups for infl uenza vaccination 8 However the infl uenza vaccination coverage rate in elderly was found to be low and varied seasonally 4 3 40 0 9 Currently infl uenza vaccine is not included in the National Immunization Schedule in China and individuals need to completely pay out of pockets for immunization 10 Both the economic and health impacts of infl uenza vaccination were war ranted to be assessed Most previous pharmacoeconomic studies found QIV to be cost effective compared with TIV in developed countries or regions 11 14 Yet there is a scarcity of data compar ing these two vaccination strategies in China Therefore we aimed to evaluate the cost effectiveness of QIV versus TIV among elderly population in China 2 Methods 2 1 Decision analytic model We designed a 1 year decision analytic model Fig 1 to simu late three vaccination options no vaccination TIV and QIV in a hypothetical cohort of Chinese elderly aged 69 year old The age of elderly simulated in the model was similar to the mean age of Chinese elderly aged over 60 years 15 Individuals with infl u enza vaccine contraindicated were excluded from the present model We assumed patients who had history of infl uenza vacci nation had no or similar effects on the risk reduction of infl uenza infection in all three options of the model Direct and indirect costs infl uenza infection rate infl uenza related mortality rate and quality adjusted life year QALY loss in each option were the primary outcomes simulated from the societal perspective of China In TIV and QIV options of the model individuals could either accept or reject the infl uenza vaccination All elderly vaccinated or not were possibly at high or low risk of infl uenza infection and might be infected by infl uenza during the year The high risk elderly were those who had history of chronic diseases such as coronary heart disease hypertension diabetes and tuberculosis Infectedpopulationwere defi nedas laboratory confi rmed infl uenza patients and they might seek for inpatient or outpatient care Patients with inpatient care might survive or die after treatment 2 2 Clinical inputs Model inputs were listed in Table 1 We performed literature search on Medline and Embase over the period of 2000 2019 using the keywords infl uenza infl uenza vaccine vaccine acceptance vaccine effectiveness mortality elderly utility and Chinese The study selection criteria included 1 research articles written in English and meta analyses were preferred 2 infl uenza infec tion was confi rmed by laboratory testing 3 inpatient or outpa tienthealthcareutilizationhadlowtransferabilityamong countries and corresponding data from China were needed All data retrieved for model inputs were screened and assessed for eli gibility and quality independently by two authors The coverage rate of infl uenza vaccination 26 7 among elderly was derived from a meta analysis on estimation of pooled infl uenza vaccination rate from 126 studies in China 9 The pro portion of high risk population among elderly 53 99 was derived from data reported in 2018 Chinese Health Statistics Yearbook 15 The incidence of infl uenza infection in unvacci nated individuals was retrieved from a meta analysis of 32 ran domized clinical trials N 13 329 on evaluation of annual infl uenza attack rate 16 The odds ratio of infl uenza infection in unvaccinated high versus low risk individuals was retrieved from an epidemiological analysis in 19 hospitals of global infl u enza surveillance network 17 The infl uenza infection rate in unvaccinated low risk elderly was further estimated by incorpo rating the data of proportion of high risk elderly the infl uenza infection rate in all unvaccinated population and the odds ratio of infl uenza infection in unvaccinated high versus low risk indi viduals The vaccine effectiveness of TIV was retrieved from a meta analysis of 95 trials N 2 504 162 on assessment of the effi cacy effectiveness and safety of infl uenza vaccination 4 The relative risk of vaccine effectiveness by QIV versus TIV was derived from a meta analysis on immunogenicity comparison of QIV versus TIV in elderly over 60 years 5 The infl uenza infec tion rate in vaccinated elderly was calculated by the following formula Infection rate of unvaccinated individuals 1 vaccine effectiveness The hospitalization rates of high or low risk individuals after infl uenza were estimated from a retrospective study on evaluation of infl uenza related healthcare resource use in 395 Chinese elderly Influenza High risk No influenza Survived Low risk Survived Inpatient Outpatient Dead Elderly in China QIV TIV Accept Reject Accept Reject No vaccination Fig 1 Simplifi ed decision analytic model of infl uenza vaccination options among the elderly in China TIV trivalent infl uenza vaccine QIV quadrivalent infl uenza vaccine 1058M Jiang et al Vaccine 38 2020 1057 1064 18 The mortality rate of inpatient care 12 was estimated by pooling data from infl uenza surveillance hospitals in China 19 20 2 3 Utility and costs inputs The utility score of healthy elderly was derived from a community based study on assessment of health related quality of life in 10 257 Chinese elderly 60 years by using EQ 5D 3L instrument 21 The utility scores of outpatient and inpatient care were retrieved from a retrospective telephone survey with EQ 5D 3L instrument on 529 outpatients and 249 inpatients in China 22 The length of outpatient care 7 days range 3 10 days was sim ilar to the duration of illness which was derived from a clinical study on evaluation of infl uenza vaccine effi cacy 23 The length of inpatient care 9 days range 4 14 days was derived from stud ies that evaluated the economic and hospitalization burden in Chi nese elderly with infl uenza infection 24 26 The QALY loss from inpatient or outpatient care was further calculated by the utility of inpatient or outpatient care multiplied by the corresponding length of health care The age specifi c life expectancy of Chinese elderly was estimated from data reported in 2018 Chinese Health Statistics Yearbook 15 The QALY loss from death was calculated by age specifi c time loss for death and utility value of healthy elderly discounted by an annual rate of 3 In the present study direct medical and non medical and indi rect costs productivity costs were included for cost analysis 1USD 6 61 Chinese Yuan Costs of vaccination and seeking out patient or inpatient care were the primary components of direct medical cost Direct non medical costs were costs of transportation and accommodation of caregivers Indirect costs were defi ned as productivity loss by income of patients and caregivers 25 The cost of TIV per dose USD11 and incremental cost factor of QIV versus TIV 2 2 range 2 0 2 4 were derived from local price in Shaanxi Provincial Procurement Platform 27 Direct medical costs of outpatient and inpatient care for high and low risk infected population were derived from the same source as mentioned above for hospitalization rate of infected patients 18 The direct non medical cost and indirect costs were both retrieved from a retro spective study on evaluating the economic burden of infl uenza infection among 529 outpatients and 254 inpatients in China 25 All costs from published literature were adjusted to 2019 data by using the infl ation rate of the annual consumer price index of medical care from the National Bureau of Statistics of China 28 2 4 Cost effectiveness analysis and sensitivity analysis Cost effectiveness and sensitivity analyses were performed by TreeAge Pro 2018 TreeAge Software Inc Williamstown MA and Microsoft Excel 2013 Microsoft Corporation Redmond WA USA In base case analysis the expected values of overall cost infl uenza infection rate infl uenza related mortality rate and QALY loss were calculated by utilizing the base case values of model variables One option was dominated when it was more costly with higher QALY loss than another option When QIV was more effec tive lower QALY loss at higher cost compared to TIV the incre mental cost effectiveness ratio ICER of QIV would be calculated with the following equation ICER CostQIV CostTIV QALY lossTIV QALY lossQIV As recommended by WHO one option was considered to be cost effective if the ICER was below 3 fold gross domestic product GDP per capita 29 The GDP per capita in China was obtained from the National Statistics Bureau for year 2017 28 which was adjusted to year 2019 USD9 860 GDP per capita using the projected GDP growth rate from 2017 to 2019 reported by the International Monetary Fund 30 Therefore the willingness to pay WTP threshold applied in the present study was 29 580 USD QALY 3 fold GDP per capita Sensitivity analyses were performed to examine the uncertainty of all model inputs One way sensitivity analysis was conducted Table 1 Model inputs Base case valueRangeDistribution typeReference Clinical Inputs Coverage rate of infl uenza vaccination0 2670 043 0 400Triangular9 Proportion of high risk elderly0 53990 4319 0 6479Triangular15 Odds ratio of infl uenza infection in unvaccinated high versus low risk elderly1 291 05 1 58Lognormal17 Infl uenza infection rate in unvaccinated low risk elderly0 0620 055 0 070Triangular15 17 Vaccine effectiveness of TIV0 490 33 0 62Triangular4 Relative risk of vaccine effectiveness of QIV versus TIV1 161 00 1 34Lognormal5 Hospitalization rate of high risk elderly with infl uenza infection0 39340 3147 0 4721Beta18 Hospitalization rate of low risk elderly with infl uenza infection0 08980 0718 0 1078Beta18 Mortality rate of inpatient care0 120 096 0 144Beta19 20 Utility Inputs Utility score of healthy elderly0 80360 8 1Triangular21 Utility score of outpatients0 57330 4586 0 6880Triangular22 Utility score of inpatients0 41280 3302 0 4954Triangular22 Age of elderly 69 Triangular15 Length of outpatient care days 73 10Triangular23 Length of inpatient care days 94 14Triangular24 26 Cost Inputs USD Cost of TIV per dose11 27 Incremental cost factor of QIV versus TIV2 22 0 2 4Lognormal27 Cost of inpatient care Direct medical cost high risk elderly 2 9442 355 3 533Lognormal18 Direct medical cost low risk elderly 1 4891 191 1 787Lognormal18 Direct non medical cost254203 304Lognormal25 Indirect cost226181 271Lognormal25 Cost of outpatient care Direct medical cost high risk elderly 55 644 5 66 7Lognormal18 Direct medical cost low risk elderly 63 650 9 76 4Lognormal18 Direct non medical cost11 18 9 13 3Lognormal25 Indirect cost63 951 1 76 8Lognormal25 1USD 6 61CNY TIV trivalent infl uenza vaccine QIV quadrivalent infl uenza vaccine In probabilistic sensitivity analysis the age of elderly was sampled by a triangular distribution 69 years range 60 90 years M Jiang et al Vaccine 38 2020 1057 10641059 over variable ranges 95 confi dence interval CI or 20 of base case values to identify the infl uential parameters on base case results To assess the impacts of all variables simultaneously prob abilistic sensitivity analysis was performed for 10 000 Monte Carlo simulations by randomly drawing each model input with a specifi c probability distribution as shown in Table 1 3 Results 3 1 Base case analysis The base case results were shown in Table 2 No vaccination was dominated more QALY loss at higher cost by TIV and QIV which was excluded from further cost effectiveness analysis Com pared with TIV QIV was more costly USD56 29 versus USD54 28 with lower infl uenza infection rate 0 0608versus0 0623 infl uenza related mortality rate 0 00199 versus 0 00204 and QALY loss 0 01213 versus 0 01243 QIV was therefore found to be cost effective compared to TIV with ICER of 6 700 USD QALY far below the WTP threshold of 29 580 USD QALY 3 fold GDP per capita 3 2 Sensitivity analysis One way sensitivity analysis suggested that the base case results were subject to the relative risk of vaccine effectiveness of QIV versus TIV 1 16 range 1 00 1 34 If the relative risk was below 1 05 TIV would be the cost effective option The variation of cost and QALY in each option against the relative risk of vaccine effectiveness of QIV versus TIV were shown in Fig 2a and 2b respectively Throughout the range of the relative risk no vaccina tion was dominated more costly at higher QALY loss by TIV and QIV remained more effective lower QALYs loss at higher cost compared to TIV QIV was cost effective compared to TIV when the relative risk was higher than 1 05 because the ICERs of QIV ver sus TIV were below the WTP threshold When the relative risk was less than 1 05 the ICERs of QIV versus TIV exceeded the WTP threshold and TIV would be the preferred option Except the relative risk of vaccine effectiveness of QIV versus TIV the impacts of one way variation of the remaining parameters on ICERs of QIV versus TIV were shown in Fig 3 The vaccine effec tiveness of TIV was the most infl uential factor To further examine the impact of vaccine effectiveness of TIV 0 49 range 0 33 0 62 Table 2 Base case results Option Infl uenza infection rate Infl uenza related morality rate Cos USD QALY loss ICER USD QALY saved TIV0 06230 0020454 280 01243 QIV0 06080 0019956 290 012136 700 Novaccination0 07170 00235 56 930 01430Dominated TIV trivalent infl uenza vaccine QIV quadrivalent infl uenza vaccine ICER incre mental cost effectiveness ratio QALY quality adjusted life year No vaccination was dominated by TIV or QIV due to more QALY loss at higher cost Fig 2 One way sensitivity analyses a Cost and b QALY loss of each option against relative risk of vaccine effectiveness of QIV versus TIV TIV trivalent infl uenza vaccine QIV quadrivalent infl uenza vaccine QALY quality adjusted life year 1060M Jiang et al Vaccine 38 2020 1057 1064 we extended its range to 0 1 Results showed TIV would be pre ferred over QIV when the vaccine effectiveness was below 0 16 because the ICERs of QIV versus TIV exceeded the WTP threshold of 29 580 USD QALY In addition to explore the impact of vaccina tion cost on base case results we extended the range of incremen tal cost factor of QIV versus TIV from 2 0 to 2 4 to 1 0 2 4 low limit QIV and TIV cost the same QIV was found to be dominant gain higher QALY at lower cost compared to TIV when the factor was below 1 51 Probabilistic sensitivity analysis was performed by 10 000 Monte Carlo simulations Comparing with TIV Fig 4a QIV was more costly by USD1 762 95 CI USD1 742 1 782 p 0 05 and more effective by saving 0 000277 QALY 95 CI 0 000273 0 0000282 QALY p 0 05 Of 10 000 simulations QIV saved more QALY at lower cost in 2 79 of time QIV was more effective lower QALY loss at higher cost with ICERs below and above the WTP threshold of 29 580 USD QALY in 83 87 and 10 85 of time respectively The remaining 2 49 simulations showed QIV to be less effective higher QALY loss as compared to TIV Comparing with no vaccination Fig 4b QIV was cost saving by USD0 499 95 CI USD0 458 0 541 p 0 05 and gained more QALY by 0 00198 QALY 95 CI 0 00196 0 002 QALY p 0 05 QIV saved QALY at lower cost in 56 23 of time The remaining 43 77 of 10 000 simulations showed QIV was more effective and costly with ICERs below the WTP threshold The probability of each option to be cost effective

内容简介：

-