Table 1.
Inputs of the decision model for the evaluation of IVIG for severe sepsis/septic shock: parameter values and uncertainty over parameter values
| Parameter | Source | Base case | Notes on scenario analyses undertaken | Notes on subgroup analyses |
|---|---|---|---|---|
| Cohort characteristics | ||||
| Mean age of a severe sepsis patient at admission to hospital | CMPD | 63 years old | Same as base case | Assumed to vary in subgroups defined using age. Sourced from ICNARC database |
| Proportion of males in a severe sepsis population at admission to hospital | CMPD | 0.53 | Same as base case | Assumed to vary in subgroups defined using gender. Sourced from CMPD |
| Short-term outcomes (ST) | ||||
| Probability of dying in hospital when SC is used in the treatment of severe sepsis (baseline risk) | CMPD | 40.6%, 95% CI (40%, 41.2%) | Same as base case | Assumed to vary per subgroup (all). Sourced from CMPD |
| Odds ratio, when IVIG is used to complement SC in the treatment of severe sepsis (based on Model M1) | Evidence synthesis (‘Clinical effectiveness of IVIG’) | 0.75 , 95% CI (0.58, 0.96) | Alternative models tested; see Table 2 | Same as base case |
| Longer-term outcomes (LT) | ||||
| Age specific probability of dying in yearly intervals, conditional on patients having survived up to the start of the year. | Cuthbertson database and general population life tables 2010 | Figure 4. Varies with time. | (1) time horizon | Assumed to vary for subgroups defined using age and APACHE II score. Sourced from CMPD |
| (2) time points at which patients reverted to survival of general population | ||||
| Cost-related parameters | ||||
| Costs of overall IVIG therapy | Non-stochastic, BNF | £5,539.05 | Same as base case | Same as base case |
| Costs of SC, when only SC is used in the treatment of severe sepsis | Non-stochastic, | £0 | Same as base case | Same as base case |
| LOS in ICU for patients remaining alive until discharge from hospital | CMPD | 8.48 (SE = 0.086) | Same as base case | Assumed to vary for all subgroups. Sourced from CMPD |
| LOS in ICU for patients dying in hospital | CMPD | 7.40 (SE = 0.108) | Same as base case | Assumed to vary for all subgroups. Sourced from CMPD |
| Costs associated to a day in ICU for a patient with severe sepsis | Non-stochastic, reference costs [24] | £1,393 | Same as base case | Same as base case |
| Overall hospital LOS for patients remaining alive until discharge from hospital | CMPD | 21.29 (SE = 0.292) | Same as base case | Assumed to vary for all subgroups. Sourced from CMPD |
| Overall hospital LOS for patients dying in hospital | CMPD | 39.07 (SE = 0.325) | Same as base case | Assumed to vary for all subgroups. Sourced from CMPD |
| Costs associated to a day in wards other than ICU for a patient with a severe sepsis episode | Non-stochastic, reference costs [24] | £196 | Same as base case | Same as base case |
| Costs incurred between year t-1 and year t after hospital discharge | Manns [25] | t = 1: £13,654 and t >1: £4,466.5/year | (1) ± 50% of Manns’ estimates | Same as base case |
| (2) average annual per capital NHS cost for the general population | ||||
| Utilities | ||||
| In-hospital HRQoL weight associated to severe sepsis patients | Drabinsky [26] | 0.53 | Same as base case | Same as base case |
| HRQoL weight associated to severe sepsis patients at year t | Cuthbertson database, Drabinsky [26] | t = 1: 0.62 and t >1: 0.6833 | Same as base case | Same as base case |
See NIHR HTA full technical report [10] for a comprehensive reporting of parameter values used to inform the decision model, including those used in subgroup analyses. IVIG, intravenous immunoglobulin; CMPD, Case Mix Programme Database; ICNARC, Intensive Care National Audit & Research Centre; CI, confidence interval; BNF, British National Formulary; SC, standard care; APACHE II, Acute Physiology and Chronic Health Evaluation II; LOS, length of stay; HRQoL, health-related quality of life.