Cost-effectiveness of tiotropium in patients with asthma poorly controlled on inhaled glucocorticosteroids and long-acting beta-agonists
Technical appendix
1WinBUGS model for computation of transitions
A sample of 793 patients with complete sets of ACQ data was used to construct model transition probabilities representing 87% of the 912 patients enrolled in the PrimoTinA-asthma® trials. To obtain the model transition probabilities imputation methods were used to assign each patient to either an asthma control state (controlled asthma, partly controlled asthma or uncontrolled asthma) or to one of three asthma exacerbation health states each week throughout the 48 week trial period. These methods have been described in the main text of this article.
Once this 48-week health state dataset had been created Bayesian methods were used to derive weekly transition probabilities. Two sets of weekly transition probabilities were computed; one set reflected weekly transitions over week 0 to 8 in the trial and one reflected weekly transition over week 9 to 48 in the trial.
WinBUGS (version 1.4) was used to compute the model transition probabilities. WinBUGS is a statistical software for Bayesian analysis using Markov chain Monte Carlo methods. Table 1 presents the code for the WinBUGS model. ‘## Model’ indicates the start and ‘## End of Model’ indicates the last code. Following that the data used in the model is listed. The model was compiled with two chains; inits for these were generated by WinBUGS. The p.plac node or the p.tio node was monitored. The model was run with 50,000 simulations (plus a burn-in of 20,000). Convergence was confirmed visually.
Table 1: WinBUGS model
WinBUGS model## MODEL
# Model is based on Briggs, Ades and Price (Medical Decision Making (2003);23:341-350)
model{
## Multinomial distribution for r,
## i = 6 for non-absorbing states
for (i in 1 : 6){
r.tio[i, 1:6] ~ dmulti(p.tio[i, 1:6], n.tio[i])
r.plac[i, 1:6] ~ dmulti(p.plac[i, 1:6], n.plac[i])
}
## Dirchlet prior distributions for the transition probabilities
for(i in 1:6){
p.tio[i, 1:6] ~ ddirch(prior.tio[i, 1:6])
p.plac[i, 1:6] ~ ddirch(prior.plac[i, 1:6])
}
}
## End of model
##DATA – From Trial 416 day 56 (placebo)a
list(
r.plac=structure(.Data=c(96, 0, 0, 1, 2,
3, 162, 1, 5, 1,
20, 28, 1172, 26, 12,
3, 3, 22, 4, 1,
2, 0, 9, 0, 13,
1, 0, 1, 0, 0),.Dim=c(6,6)),
n.plac=c(99, 172, 1260, 33, 24, 4),
prior.plac=structure(.Data=c(1,1,1,1,1,1,
1,1,1,1,1,1,
1,1,1,1,1,1,
1,1,1,1,1,1,
1,1,1,1,1,1,
1,1,1,1,1,1),.Dim=c(6,6)),
)
## End of Data
aExample for placebo data from trial 416 for the first 8 weeks.
2Model transition probabilities
The two transition matrices (“early response phase” and “late response phase”)are presented inTable 2 and Table 3for the tiotropium and usual care arm respectively.
Table 2: Model transition probabilities, tiotropium arm
toFrom / Controlled asthma / Partly controlled asthma / Uncontrolled asthma / Non-severe exacerbation / Severe exacerbation - without hospitalisation / Severe exacerbation – with hospitalisation
Day 0-56 “Early response phase”
Controlled asthma / 0.945 / 0.004 / 0.008 / 0.025 / 0.013 / 0.004
Partly controlled asthma / 0.057 / 0.889 / 0.025 / 0.020 / 0.007 / 0.002
Uncontrolled asthma / 0.018 / 0.022 / 0.930 / 0.021 / 0.008 / 0.000
Non-severe exacerbation / 0.110 / 0.069 / 0.616 / 0.137 / 0.055 / 0.014
Severe exacerbation - without hospitalisation / 0.041 / 0.042 / 0.417 / 0.042 / 0.437 / 0.021
Severe exacerbation – with hospitalisation / 0.166 / 0.167 / 0.167 / 0.167 / 0.166 / 0.167
Day 57-336 “Late response phase”
Controlled asthma / 0.954 / 0.019 / 0.010 / 0.010 / 0.007 / 0.000
Partly controlled asthma / 0.029 / 0.913 / 0.033 / 0.019 / 0.005 / 0.002
Uncontrolled asthma / 0.006 / 0.016 / 0.939 / 0.030 / 0.010 / 0.001
Non-severe exacerbation / 0.081 / 0.092 / 0.576 / 0.237 / 0.009 / 0.004
Severe exacerbation - without hospitalisation / 0.097 / 0.078 / 0.292 / 0.019 / 0.510 / 0.004
Severe exacerbation – with hospitalisation / 0.062 / 0.125 / 0.281 / 0.063 / 0.031 / 0.438
Table 3: Model transition probabilities, usual care arm
toFrom / Controlled asthma / Partly controlled asthma / Uncontrolled asthma / Non-severe exacerbation / Severe exacerbation - without hospitalisation / Severe exacerbation – with hospitalisation
Day 0-56 “Early response phase”
Controlled asthma / 0.950 / 0.006 / 0.006 / 0.017 / 0.017 / 0.006
Partly controlled asthma / 0.030 / 0.875 / 0.034 / 0.041 / 0.017 / 0.002
Uncontrolled asthma / 0.015 / 0.023 / 0.927 / 0.025 / 0.009 / 0.001
Non-severe exacerbation / 0.060 / 0.107 / 0.631 / 0.143 / 0.036 / 0.024
Severe exacerbation - without hospitalisation / 0.083 / 0.063 / 0.292 / 0.042 / 0.500 / 0.021
Severe exacerbation – with hospitalisation / 0.167 / 0.083 / 0.250 / 0.083 / 0.084 / 0.333
Day 57-336 “Late response phase”
Controlled asthma / 0.943 / 0.015 / 0.008 / 0.023 / 0.010 / 0.001
Partly controlled asthma / 0.031 / 0.901 / 0.027 / 0.030 / 0.009 / 0.001
Uncontrolled asthma / 0.005 / 0.011 / 0.933 / 0.037 / 0.012 / 0.001
Non-severe exacerbation / 0.123 / 0.107 / 0.551 / 0.208 / 0.009 / 0.002
Severe exacerbation - without hospitalisation / 0.066 / 0.061 / 0.352 / 0.017 / 0.499 / 0.006
Severe exacerbation – with hospitalisation / 0.118 / 0.029 / 0.294 / 0.088 / 0.030 / 0.441
3. Parameter ranges used in the one way sensitivity analyses
The impact of variation in the health state cost, rescue medication cost and utility parameters was tested in the one-way sensitivity analysis.
The health state cost and rescue medication cost have been calculated by multiplying the relevant elements of resource use by their unit costs. A lower and an upper bound were fitted around these variables by varying each elements of resource use within its 95% confidence interval but assuming unit costs remain constant. For the utility parameters the 95% confidence interval was calculated from the trial data and this is used as the lower and upper bound for the one-way sensitivity analysis input.
Table 4: Parameter ranges used in the one-way sensitivity analysis
Parameter / Base case value / Lower bound / Upper boundHealth state cost
Health state costs, controlled asthma, per cycle / 7.18 / 1.92 / 41.56
Health state costs, partly controlled asthma, per cycle / 11.61 / 2.95 / 57.48
Health state costs, uncontrolled asthma, per cycle / 41.80 / 17.89 / 98.34
Health state costs, non-severe exacerbation, per cycle / 65.58 / 40.99 / 113.31
Health state costs, severe exacerbations without hospitalisation, per cycle / 83.50 / 64.99 / 109.65
Health state costs, severe exacerbations with hospitalisation, per cycle / 1,021.33 / 620.35 / 1,550.45
Utilities
Utility for controlled asthmastate / 0.937 / 0.923 / 0.952
Utility for partly controlled asthma state / 0.907 / 0.879 / 0.934
Utility for uncontrolled asthma state / 0.728 / 0.707 / 0.749
Utility for mild/moderate exacerbation state / 0.649 / 0.584 / 0.714
Utility for severe exacerbations without hospitalisation state / 0.570 / 0.513 / 0.627
Utility for severe exacerbations with hospitalisation state / 0.330 / 0.297 / 0.363
Rescue medication cost
Rescue medication costs, controlled asthma, per cycle, Tiotropium / 0.06 / 0.05 / 0.06
Rescue medication costs, partly controlled asthma, per cycle, Tiotropium / 0.08 / 0.07 / 0.08
Rescue medication costs, uncontrolled asthma, per cycle, Tiotropium / 0.15 / 0.14 / 0.15
Rescue medication costs, non-severe exacerbation, per cycle, Tiotropium / 0.17 / 0.14 / 0.19
Rescue medication costs, severe exacerbations without hospitalisation, per cycle, Tiotropium / 0.20 / 0.17 / 0.23
Rescue medication costs, severe exacerbations with hospitalisation, per cycle, Tiotropium / 0.21 / 0.19 / 0.23
Rescue medication costs, controlled asthma, per cycle, Usual Care / 0.06 / 0.05 / 0.06
Rescue medication costs, partly controlled asthma, per cycle, Usual Care / 0.09 / 0.08 / 0.09
Rescue medication costs, uncontrolled asthma, per cycle, Usual Care / 0.17 / 0.16 / 0.18
Rescue medication costs, non-severe exacerbation, per cycle, Usual Care / 0.15 / 0.13 / 0.17
Rescue medication costs, severe exacerbations without hospitalisation, per cycle, Usual Care / 0.22 / 0.19 / 0.25
Rescue medication costs, severe exacerbations with hospitalisation, per cycle, Usual Care / 0.21 / 0.19 / 0.23
4Distributions around model parameters used in probabilistic sensitivity analyses
The impact of uncertainty in health state related resource use, rescue medication resource use, utilities and model transition probabilities was tested in the probabilistic sensitivity analysis,
The efficacy parameters take the form of a transition matrix where there are 36 possible transitions. All transitions are therefore inter-related due to the nature of the transition matrix structure. Due to this inter-dependence, the uncertainty around the transition probabilities in the transition matrix was captured in the model by using a Dirichlet distribution. As there is no built-in function for the Dirichlet distribution in Excel, it was approximated using a Gamma distribution (or a Normal distribution in cases where the number of patient transitions were more than 340) as described by Briggs et al. 2003[1]. The six possible transitions from each health state were based on observed patient numbers in the RCT data which weresubsequently divided by the row sum of all six transitions to derive the transition probability. The numbers of transitions (rather than the probabilities) were sampled in the probabilistic sensitivity analysis, and the probability was again then computed by dividing by the row sum of all six possible transitions from each health state. Each transition approximated using the Gamma distribution had an alpha parameter which was equal to the ‘number of transitions that had been observed in the trial plus 1’ and beta was set to ‘1’. The transitions approximated by the Normal distribution were specifiedusing the mean, which was equal to the number of transitions observed, and the standard deviation, which was the square root of the number of transitions observed. Thus, the degree of uncertainty in the Dirichlet distribution directly depends on the number of transitions observed in the trial as well as the total number of transitions from one particular health state.Table 5 and Table 6 display each transition probability alongside the parameters of the distributions used to sample the associated number of transitions for the tiotropium and usual care arms respectively.
The uncertainty incost estimates wascaptured by uncertainties in resource use while the unit costs of the resources were assumed to be fixed. Distributions were fitted around the elements of resource use used to calculate the health state specific costs and rescue medication costs. Distributional information was available because resource use parameters were either obtained from the survey of 15 health care providers or from the RCT data. The Lognormal distribution was used to represent the uncertainty in outpatient visit (Table 7), home visit (Table 8), test and procedure resource use (Table 9) and rescue medication use (Table 12) variables. The beta distribution was used to represent uncertainty in co-medication use (Table 10) and hospitalisation resource use (Table 11). A beta distribution was fitted to the utility parameters (Table 13).
Table 5: Model transition probabilities and distributional information, tiotropium arm
toFrom / Controlled asthma / Partly controlled asthma / Uncontrolled asthma / Non-severe exacerbation / Severe exacerbation - without hosp. / Severe exacerbation – with hosp.
Day 0-56 “Early response phase”
Controlled asthma / Probability / 0.945 / 0.004 / 0.008 / 0.025 / 0.013 / 0.004
Distribution / Gamma / Gamma / Gamma / Gamma / Gamma / Gamma
Distribution parameters / α: 224 β: 1 / α: 1 β: 1 / α: 2 β: 1 / α: 6 β: 1 / α: 3 β: 1 / α: 1 β: 1
Partly controlled asthma / Probability / 0.057 / 0.889 / 0.025 / 0.020 / 0.007 / 0.002
Distribution / Gamma / Normal / Gamma / Gamma / Gamma / Gamma
Distribution parameters / α: 25 β: 1 / μ: 391 sd:19.8 / α: 11 β: 1 / α: 9 β: 1 / α: 3 β: 1 / α: 1 β: 1
Uncontrolled asthma / Probability / 0.018 / 0.022 / 0.930 / 0.021 / 0.008 / 0.000
Distribution / Gamma / Gamma / Normal / Gamma / Gamma / Gamma
Distribution parameters / α: 43 β: 1 / α: 52 β: 1 / μ: 2223 sd: 47.1 / α: 51 β: 1 / α: 20 β: 1 / α: 1 β: 1
Non-severe exacerbation / Probability / 0.110 / 0.069 / 0.616 / 0.137 / 0.055 / 0.014
Distribution / Gamma / Gamma / Gamma / Gamma / Gamma / Gamma
Distribution parameters / α: 8 β: 1 / α: 5 β: 1 / α: 45 β: 1 / α: 10 β: 1 / α: 4 β: 1 / α: 1 β: 1
Severe exacerbation - without hosp. / Probability / 0.041 / 0.042 / 0.417 / 0.042 / 0.437 / 0.021
Distribution / Gamma / Gamma / Gamma / Gamma / Gamma / Gamma
Distribution parameters / α: 2 β: 1 / α: 2 β: 1 / α: 20 β: 1 / α: 2 β: 1 / α: 21 β: 1 / α: 1 β: 1
Severe exacerbation – with hosp. / Probability / 0.166 / 0.167 / 0.167 / 0.167 / 0.166 / 0.167
Distribution / Gamma / Gamma / Gamma / Gamma / Gamma / Gamma
Distribution parameters / α: 1 β: 1 / α: 1 β: 1 / α: 1 β: 1 / α: 1 β: 1 / α: 1 β: 1 / α: 1 β: 1
Day 57-336 “Late response phase”
Controlled asthma / Probability / 0.954 / 0.019 / 0.010 / 0.010 / 0.007 / 0.000
Distribution / Normal / Gamma / Gamma / Gamma / Gamma / Gamma
Distribution parameters / μ: 3448 sd: 58.7 / α: 70 β: 1 / α: 36 β: 1 / α: 35 β: 1 / α: 26 β: 1 / α: 1 β: 1
Partly controlled asthma / Probability / 0.029 / 0.913 / 0.033 / 0.019 / 0.005 / 0.002
Distribution / Gamma / Normal / Gamma / Gamma / Gamma / Gamma
Distribution parameters / α: 84 β: 1 / μ: 2679 sd: 51.8 / α: 97 β: 1 / α: 55 β: 1 / α: 14 β: 1 / α: 5 β: 1
Uncontrolled asthma / Probability / 0.006 / 0.016 / 0.939 / 0.030 / 0.010 / 0.001
Distribution / Gamma / Gamma / Normal / Gamma / Gamma / Gamma
Distribution parameters / α: 49 β: 1 / α: 133 β: 1 / μ: 8014 sd: 89.5 / α: 252 β: 1 / α: 82 β: 1 / α: 9 β: 1
Non-severe exacerbation / Probability / 0.081 / 0.092 / 0.576 / 0.237 / 0.009 / 0.004
Distribution / Gamma / Gamma / Gamma / Gamma / Gamma / Gamma
Distribution parameters / α: 37 β: 1 / α: 42 β: 1 / α: 262 β: 1 / α: 108 β: 1 / α: 4 β: 1 / α: 2 β: 1
Severe exacerbation - without hosp. / Probability / 0.097 / 0.078 / 0.292 / 0.019 / 0.510 / 0.004
Distribution / Gamma / Gamma / Gamma / Gamma / Gamma / Gamma
Distribution parameters / α: 25 β: 1 / α: 20 β: 1 / α: 75 β: 1 / α: 5 β: 1 / α: 131 β: 1 / α: 1 β: 1
Severe exacerbation – with hosp. / Probability / 0.062 / 0.125 / 0.281 / 0.063 / 0.031 / 0.438
Distribution / Gamma / Gamma / Gamma / Gamma / Gamma / Gamma
Distribution parameters / α: 2 β: 1 / α: 4 β: 1 / α: 9 β: 1 / α: 2 β: 1 / α: 1 β: 1 / α: 14 β: 1
Source: PrimoTinA-asthma® clinical trials.
α: alpha parameter of gamma distribution, β: beta parameter of gamma distribution, μ: mean of normal distribution, sd: standard deviations, hosp.: hospitalisation.
Table 6: Model transition probabilities and distributional information, usual care arm
toFrom / Controlled asthma / Partly controlled asthma / Uncontrolled asthma / Non-severe exacerbation / Severe exacerbation - without hosp. / Severe exacerbation – with hosp.
Day 0-56 “Early response phase”
Controlled asthma / Probability / 0.950 / 0.006 / 0.006 / 0.017 / 0.017 / 0.006
Distribution / Gamma / Gamma / Gamma / Gamma / Gamma / Gamma
Distribution parameters / α: 170 β: 1 / α: 1 β: 1 / α: 1 β: 1 / α: 3 β: 1 / α: 3 β: 1 / α: 1 β: 1
Partly controlled asthma / Probability / 0.030 / 0.875 / 0.034 / 0.041 / 0.017 / 0.002
Distribution / Gamma / Normal / Gamma / Gamma / Gamma / Gamma
Distribution parameters / α: 14 β: 1 / μ: 406 sd: 20.1 / α: 16 β: 1 / α: 19 β: 1 / α: 8 β: 1 / α: 1 β: 1
Uncontrolled asthma / Probability / 0.015 / 0.023 / 0.927 / 0.025 / 0.009 / 0.001
Distribution / Gamma / Gamma / Normal / Gamma / Gamma / Gamma
Distribution parameters / α: 36 β: 1 / α: 57 β: 1 / μ: 2253 sd: 47.5 / α: 61 β: 1 / α: 21 β: 1 / α: 3 β: 1
Non-severe exacerbation / Probability / 0.060 / 0.107 / 0.631 / 0.143 / 0.036 / 0.024
Distribution / Gamma / Gamma / Gamma / Gamma / Gamma / Gamma
Distribution parameters / α: 5 β: 1 / α: 9 β: 1 / α: 53 β: 1 / α: 12 β: 1 / α: 3 β: 1 / α: 2 β: 1
Severe exacerbation - without hosp. / Probability / 0.083 / 0.063 / 0.292 / 0.042 / 0.500 / 0.021
Distribution / Gamma / Gamma / Gamma / Gamma / Gamma / Gamma
Distribution parameters / α: 4 β: 1 / α: 3 β: 1 / α: 14 β: 1 / α: 2 β: 1 / α: 24 β: 1 / α: 1 β: 1
Severe exacerbation – with hosp. / Probability / 0.167 / 0.083 / 0.250 / 0.083 / 0.084 / 0.333
Distribution / Gamma / Gamma / Gamma / Gamma / Gamma / Gamma
Distribution parameters / α: 2 β: 1 / α: 1 β: 1 / α: 3 β: 1 / α: 1 β: 1 / α: 1 β: 1 / α: 4 β: 1
Day 57-336 “Late response phase”
Controlled asthma / Probability / 0.954 / 0.019 / 0.010 / 0.010 / 0.007 / 0.000
Distribution / Normal / Gamma / Gamma / Gamma / Gamma / Gamma
Distribution parameters / μ: 3081 sd: 55.5 / α: 49 β: 1 / α: 26 β: 1 / α: 75 β: 1 / α: 32 β: 1 / α: 4 β: 1
Partly controlled asthma / Probability / 0.029 / 0.913 / 0.033 / 0.019 / 0.005 / 0.002
Distribution / Gamma / Normal / Gamma / Gamma / Gamma / Gamma
Distribution parameters / α: 78 β: 1 / μ: 2233 sd: 47.3 / α: 68 β: 1 / α: 75 β: 1 / α: 22 β: 1 / α: 3 β: 1
Uncontrolled asthma / Probability / 0.006 / 0.016 / 0.939 / 0.030 / 0.010 / 0.001
Distribution / Gamma / Gamma / Normal / Normal / Gamma / Gamma
Distribution parameters / α: 44 β: 1 / α: 104 β: 1 / μ: 8556 sd: 92.5 / μ: 340 sd: 18.4 / α: 114 β: 1 / α: 10 β: 1
Non-severe exacerbation / Probability / 0.081 / 0.092 / 0.016 / 0.237 / 0.009 / 0.004
Distribution / Gamma / Gamma / Normal / Gamma / Gamma / Gamma
Distribution parameters / α: 79 β: 1 / α: 69 β: 1 / μ: 354 sd: 18.8 / α: 134 β: 1 / α: 6 β: 1 / α: 1 β: 1
Severe exacerbation - without hosp. / Probability / 0.097 / 0.078 / 0.292 / 0.019 / 0.510 / 0.004
Distribution / Gamma / Gamma / Gamma / Gamma / Gamma / Gamma
Distribution parameters / α: 24 β: 1 / α: 22 β: 1 / α: 127 β: 1 / α: 6 β: 1 / α: 180 β: 1 / α: 2 β: 1
Severe exacerbation – with hosp. / Probability / 0.062 / 0.125 / 0.281 / 0.063 / 0.031 / 0.438
Distribution / Gamma / Gamma / Gamma / Gamma / Gamma / Gamma
Distribution parameters / α: 4 β: 1 / α: 1 β: 1 / α: 10 β: 1 / α: 3 β: 1 / α: 1 β: 1 / α: 15 β: 1
Source: PrimoTinA-asthma® clinical trials.
α: alpha parameter of gamma distribution, β: beta parameter of gamma distribution, μ: mean of normal distribution, sd: standard deviations, hosp.: hospitalisation.
Table 7: Outpatient visit variables tested in PSA
Variable / Applied distribution / Expected mean / Standard deviationOutpatient visits: GP, controlled asthma, per cycle / Lognormal / 0.031 / 0.0382
Outpatient visits: GP, partly controlled asthma, per cycle / Lognormal / 0.039 / 0.0452
Outpatient visits: GP, uncontrolled asthma, per cycle / Lognormal / 0.138 / 0.1397
Outpatient visits: GP, non-severe exacerbation, per exacerbation / Lognormal / 0.597 / 0.4775
Outpatient visits: GP, severe exacerbation without hospitalisation, per exacerbation / Lognormal / 1.367 / 0.8013
Outpatient visits: GP, severe exacerbation with hospitalisation, per exacerbation / Lognormal / 0.590 / 0.4936
Outpatient visits: nurse, controlled asthma, per cycle / Lognormal / 0.050 / 0.0414
Outpatient visits: nurse, partly controlled asthma, per cycle / Lognormal / 0.068 / 0.0497
Outpatient visits: nurse, uncontrolled asthma, per cycle / Lognormal / 0.158 / 0.2060
Outpatient visits: nurse, non-severe exacerbation, per exacerbation / Lognormal / 0.430 / 0.4199
Outpatient visits: nurse, severe exacerbation without hospitalisation, per exacerbation / Lognormal / 0.903 / 0.7316
Outpatient visits: nurse, severe exacerbation with hospitalisation, per exacerbation / Lognormal / 1.380 / 1.7493
Outpatient visits: specialist, controlled asthma, per cycle / Lognormal / 0.016 / 0.0209
Outpatient visits: specialist, partly controlled asthma, per cycle / Lognormal / 0.033 / 0.0414
Outpatient visits: specialist, uncontrolled asthma, per cycle / Lognormal / 0.094 / 0.0930
Outpatient visits: specialist, non-severe exacerbation, per exacerbation / Lognormal / 0.094 / 0.1613
Outpatient visits: specialist, severe exacerbation without hospitalisation, per exacerbation / Lognormal / 0.339 / 0.5004
Source: Survey of UK HCPs (IMS Health)
Table 8: Home visit variables tested in PSA
Variable / Applied distribution / Expected mean / Standard deviationHome visits: GP, controlled asthma, per cycle / Lognormal / 0.001 / 0.0008
Home visits: GP, partly controlled asthma, per cycle / Lognormal / 0.010 / 0.0258
Home visits: GP, uncontrolled asthma, per cycle / Lognormal / 0.025 / 0.0551
Home visits: GP, non-severe exacerbation, per exacerbation / Lognormal / 0.034 / 0.1027
Home visits: GP, severe exacerbation without hospitalisation, per exacerbation / Lognormal / 0.220 / 0.3730
Home visits: GP, severe exacerbation with hospitalisation, per exacerbation / Lognormal / 0.102 / 0.1733
Home visits: nurse, uncontrolled asthma, per cycle / Lognormal / 0.001 / 0.0020
Home visits: nurse, severe exacerbation without hospitalisation, per exacerbation / Lognormal / 0.003 / 0.0129
Home visits: nurse, severe exacerbation with hospitalisation, per exacerbation / Lognormal / 0.005 / 0.0136
Source: Survey of UK HCPs (IMS Health)
Nurse home visits for the following health states were not sampled in the PSA, as no events were assumed to occur in the base case: controlled asthma, partly controlled asthma, non-severe exacerbation.
Table 9: Test and procedure resource use variables tested in PSA
Variable / Applied distribution / Expected mean / Standard deviationSpirometry tests: controlled asthma, per cycle / Lognormal / 0.026 / 0.0331
Spirometry tests: partly controlled asthma, per cycle / Lognormal / 0.028 / 0.0329
Spirometry tests: uncontrolled asthma, per cycle / Lognormal / 0.049 / 0.0422
Spirometry tests: non-severe exacerbation, per exacerbation / Lognormal / 0.287 / 0.4486
Spirometry tests: severe exacerbation without hospitalisation, per exacerbation / Lognormal / 0.288 / 0.4477
Spirometry tests: severe exacerbation with hospitalisation, per exacerbation / Lognormal / 0.463 / 0.6061
Desensitisation of asthma use: controlled asthma, per cycle / Lognormal / 0.005 / 0.02
Desensitisation of asthma use: partly controlled asthma, per cycle / Lognormal / 0.008 / 0.02
Desensitisation of asthma use: uncontrolled asthma, per cycle / Lognormal / 0.009 / 0.02
Source: Survey of UK HCPs (IMS Health)
Flu vaccination was not sampled in the PSA, as it was assumed that patients receive one flu vaccination per year.
Table 10: Co-medication use variables tested in the PSA
Variable / Applied distribution / Expected mean / Alpha / BetaPrednisolone use: controlled asthma, per cycle, per patient / Beta / 0.290 / 29 / 71
Prednisolone use: partly controlled asthma, per cycle, per patient / Beta / 0.340 / 34 / 66
Prednisolone use: uncontrolled asthma, per cycle, per patient / Beta / 0.510 / 51 / 49
Prednisolone use: non-severe exacerbation, per exacerbation, per patient / Beta / 0.570 / 57 / 43
Prednisolone use: severe exacerbation without hospitalisation, per exacerbation, per patient / Beta / 0.810 / 81 / 19
Prednisolone use: severe exacerbation with hospitalisation, per exacerbation, per patient / Beta / 0.870 / 87 / 13
Amoxicillin use: uncontrolled asthma, per cycle, per patient / Beta / 0.300 / 30 / 70
Amoxicillin use: non-severe exacerbation, per exacerbation, per patient / Beta / 0.540 / 54 / 46
Amoxicillin use: severe exacerbation without hospitalisation, per exacerbation, per patient / Beta / 0.710 / 71 / 29
Amoxicillin use: severe exacerbation with hospitalisation, per exacerbation, per patient / Beta / 0.700 / 70 / 30
Singulair use: uncontrolled asthma, per cycle, per patient / Beta / 0.830 / 83 / 17
Hydrocortisone IV use: severe exacerbation with hospitalisation, per exacerbation, per patient / Beta / 0.810 / 81 / 19
Magnesium IV: severe exacerbation with hospitalisation, per exacerbation, per patient / Beta / 0.740 / 74 / 26
Source: Survey of UK HCPs (IMS Health)
IV: Intravenous
Table 11: Hospitalisation and hospital related resource use parameters tested in the PSA
Variable / Applied distribution / Expected mean / Alpha / BetaAsthma related unscheduled hospitalisations:controlled asthma, per cycle, per patient / Beta / 0.003% / 2 / 1224
Asthma related unscheduled hospitalisations:partly controlled asthma, per cycle, per patient / Beta / 0.004% / 2 / 1100
Asthma related unscheduled hospitalisations: uncontrolled asthma, per cycle, per patient / Beta / 0.006% / 12 / 4094
Regular hospitalisations: severe exacerbation with hospitalisation, per exacerbation, per patient / Beta / 39.1% / 18 / 28
A&E visits + hospitalisation, per exacerbation, per patient / Beta / 41.3% / 19 / 27
Ambulance use with hospitalisation, per exacerbation, per patient / Beta / 2.2% / 1 / 45
Ambulance use with A&E visit and hospitalisation, per exacerbation, per patient / Beta / 4.3% / 2 / 44
Hospitalisations incl. ICU stay, per exacerbation, per patient / Beta / 13.0% / 6 / 40
A&E visits, severe exacerbation without hospitalisation, per exacerbation, per patient / Beta / 5.8% / 28 / 451
Source: PrimoTinA-asthma® clinical trials.
A&E: Accident and emergency department
Table 12: Rescue medication use variables tested in PSA