Skip to main content

Table 1 Model inputs of the analysis. In the probabilistic sensitivity analysis, the parameter inputs were simultaneously varied within the lower and upper estimates according to the distribution shown. In the deterministic sensitivity analysis, the parameter inputs were varied one-by-one between lower and upper inputs, while the “scenario” column shows inputs based on other plausible assumptions or sources

From: Cost-effectiveness of vaccination of immunocompetent older adults against herpes zoster in the Netherlands: a comparison between the adjuvanted subunit and live-attenuated vaccines

Variable Base case Lower Upper Distribution Scenario Reference (base case/scenario)
Demography
 Cohort size       Statistics Netherlands [29]
  50 years 253,491     
  60 years 222,845     
  70 years 217,058     
  80 years 93,547     
 Background mortality Age-specific      Statistics Netherlands [28]
HZ epidemiology
 HZ incidence per 100,000 person-years       NIVEL [30]/incidence adjusted for immunocompetent population using Schroder [4]
  50–59 years 591 575 607 Beta 461
  60–69 years 857 835 878 Beta 669
  70–79 years 1190 1157 1222 Beta 929
  ≥ 80 years 1481 1435 1527 Beta 1156
 False positive HZ diagnoses (%) 10.0 7.9 12.4 Beta 0 Van Hoek [31]/assuming no false positives
 HZ hospitalization incidence per 100,000 person-years       Dutch Hospital Data [32]/incidence adjusted for the immunocompetent population using Hobbelen [49]
  50–59 years 2.5 2.1 2.9 Beta 2.2
  60–69 years 4.9 4.3 5.5 Beta 4.3
  70–79 years 9.5 8.4 10.6 Beta 8.3
 ≥ 80 years 18.4 16.4 20.4 Beta 16.1
 HZ 1-day hospital admission incidence per 100,000 person-years       Dutch Hospital Data [32]/incidence adjusted for the immunocompetent population using Hobbelen [49]
  50–59 years 3.6 3.1 4.0 Beta 3.1
  60–69 years 9.6 8.8 10.5 Beta 8.4
  70–79 years 21.8 20.1 23.4 Beta 19.1
  ≥ 80 years 28.2 25.8 30.7 Beta 24.8
 HZ mortality incidence per million person-years       Statistics Netherlands [34]/incidence adjusted for the immunocompetent population using Hobbelen [49]
  50–59 years 0.1 0.3 Beta 0.1
  60–69 years 0.4 0.05 0.7 Beta 0.3
  70–79 years 1.8 0.9 2.8 Beta 1.6
  80–89 years 16.5 12.6 20.5 Beta 14.3
  ≥ 90 years 108.9 84.6 133.2 Beta 94.5
 Misclassification HZ as underlying cause of death (%) 47.5 32.0 63.0 Beta 0 Mahamud [35]/assuming no misclassification
QALY loss
 QALY loss per HZ episode       Van Wijck [6]/utilities Van Hoek/QALY loss per HZ episode Van Hoek [31]
  50–59 years 0.040 0.025 0.063 a 0.034/0.067
  ≥ 60 years 0.057 0.039 0.093 a 0.053/0.200
 QALY loss per HZ death Age-specific      Statistics Netherlands [28], Szende [36]
 QALY loss grade 3 adverse event per dose       Excluded/QALY loss for grade 3 adverse events (see Additional file 1)
  HZ/su 0     0.000329
  ZVL 0     0.000022
Costs (€, 2017)
 Health care costs
  GP visit, medication, specialist visit       Based on multiple sources (see Additional file 1)
   50–59 years 158 130 186 a  
   ≥ 60 years 198 163 233 a  
  Hospital admission       
   50–59 years 2856 2490 3222 a  
   60–69 years 3632 3166 4097 a  
   70–79 years 3671 3325 4016 a  
   ≥ 80 years 4504 4093 4915 a  
  One-day hospital admission 282      Hakkaart-van Roijen [38]
  Healthcare costs in gained life years per averted HZ death Age-specific     0 Statistics Netherlands [28], Van Baal [40]/excluding costs in gained life years
  Vaccine administration 11.36      SNPG [41]
 Patient costs       
  OTC medication per HZ episode       Based on multiple sources (see Additional file 1)
   50–59 years 10.42 8.85 12.00 a  
   ≥ 60 years 12.65 10.75 14.56 a  
  Travel costs GP visit, medication, specialist care per HZ episode       Based on multiple sources (see Additional file 1)
   50–59 years 3.33 2.61 4.06 a  
   ≥ 60 years 4.07 3.19 4.95 a  
  Travel costs hospital per HZ hospital visit/hospitalization 5.79      
  Travel costs per vaccination 0.43      
 Productivity losses       
  HZ episode       Based on multiple sources (see Additional file 1)
   50–59 years 398 230 744 a  
   60–69 years 136 78 262 a  
   ≥ 70 years 0     
  HZ death       Friction period of 84 working days [27]
   50–59 years 14,937     
   60–69 years 5074     
   ≥ 70 years     
Vaccine characteristics
 Vaccine uptake (%) 50      Assumption based on Eilers [42]
 Adherence to the second dose of HZ/su (%) 100     90, 70, 50 Assumption
 Efficacy HZ/su over time (linear function)       Function fitted using data from Cunningham [20], Lal [19] and Curran [47]
  Intercept      
   50–69 years 0.981 0.904 1.057c Normal  
   ≥ 70 years 0.992 0.956 1.028c Normal  
  Slopeb − 0.041 0.018 0.065 Beta  
 Efficacy ZVL over time (one-minus-exponential function)d      Additional efficacy against PHN /post-licensure effectiveness against HZ Function fitted using data from Oxman [11], Schmader [46] and Morrison [12]/see Additional file 1
  Intercept − 0.893 − 1.04 − 0.75 Normal
  Slope 0.0807 0.058 0.104 Beta
  Risk ratio of efficacy by age      Estimated using Rohan [45] and Schmader [44]/see Additional file 1
   50–59 years 1.282    
   60–64 years 1.274    
   65–69 years 1.219    
   70–74 years 0.852    
   75–79 years 0.711    
   80–84 years 0.391    
   ≥ 85 years 0.152    
  1. HZ herpes zoster, HZ/su HZ subunit vaccine, PHN post-herpetic neuralgia, QALY quality-adjusted life year, SNPG Stichting Nationaal Programma Grieppreventie, ZVL zoster vaccine live
  2. aAggregated costs from multiple cost items which were varied individually in the probabilistic sensitivity analysis (see Additional file 1)
  3. bThe slope of 0.009 was only used for 50- to 69-year-olds over the first 4 years covered by the trial. After 4 years, the slope of ≥ 70-year-olds was used
  4. cThe efficacy was rounded to 1 during the period that the efficacy function was above 1
  5. dThe efficacy of ZVL over time (in years) was modelled using a one-minus-exponential function 1 − exp(β1 + β2 × years), in which β1 is the intercept and β2 the slope. Risk ratios by age were used to modify the intercept. For instance, the efficacy of 60–64 years at time point zero was 1 − exp(1.274 × − 0.893 + 0 × 0.0807) = 67.9%. In our model, we used the VE of the age group 60–64 for vaccination of 60-year-olds, 70–74 for 70-year-olds, 80–84 for 80-year-olds and ≥ 85 years for the booster for 90-year-olds