Table 4.
Cost–utility analysis (deterministic models)
| Scenarios* | Control | Intervention | ∆ in cost | ∆ in effect† | ICER | Interpretation‡ |
| Healthcare services payer perspective | ||||||
| Basecase | £68.86 | £72.92 | £4.07 | −0.00005 | −£81 400 | Not cost-effective |
| Scenario A | £56.76 | £63.93 | £7.17 | −0.00010 | −£71 700 | Not cost-effective |
| Scenario B | £69.22 | £73.37 | £4.15 | −0.00012 | −£33 850 | Not cost-effective |
| Scenario C | £53.99 | £61.90 | £7.92 | 0.00160 | £4950 | Cost-effective |
| Scenario D | £78.36 | £80.25 | £1.89 | −0.0015 | −£1260 | Not cost-effective |
| Scenario E | £57.58 | £77.18 | £19.60 | 0.0030 | £6533 | Cost-effective |
| Scenario F | £68.86 | £72.92 | £4.07 | 0.00001 | £407 000 | Not cost-effective |
| Scenario G | £68.86 | £72.92 | £4.07 | 0.00042 | £9690 | Cost-effective |
| Scenario H§ | £68.86 | £72.92 | £4.07 | −0.0038 | −£1071 | Not cost-effective |
| Societal cost perspective | ||||||
| Scenario I | £133.85 | £126.42 | −£7.43 | −0.00005 | £1 48 600 | Not cost-effective |
| Scenario J | £167.36 | 154.72 | −£12.64 | −0.00005 | £2 52 800 | Not cost-effective |
| Scenario K | £120.04 | £117.51 | −£2.53 | −0.00010 | £25 300 | Not cost-effective |
| Scenario L | £135.51 | £127.59 | −£7.92 | −0.00005 | £158 400 | Not cost-effective |
| Scenario M | £135.23 | £127.44 | −£7.79 | −0.00005 | £155 800 | Not cost-effective |
| Scenario N | £120.32 | £120.14 | −£0.18 | 0.00160 | −£112 | Cost-effective and cost saving |
| Scenario O | £142.49 | £130.59 | −£11.90 | −0.00150 | £7933 | Not cost-effective |
| Scenario P | £171.35 | £62.68 | −£108.67 | 0.0030 | −£36 223 | Cost-effective and cost saving |
| Scenario Q | £133.85 | £126.42 | −£7.43 | 0.00001 | −£743 000 | Cost-effective and cost saving |
| Scenario R | £133.85 | £126.42 | −£7.43 | 0.00042 | −£17 690 | Cost-effective and cost saving |
| Scenario S§ | £133.85 | £126.42 | −£7.43 | −0.0038 | £1955 | Not cost-effective |
*Full scenario details are presented in the (online supplementary file 1).
†Changes in effect have been adjusted for baseline differences for each model and are representative of an annual timeframe (see table 2 for more details).
‡Not cost-effective is suggested if the effect is negative and therefore the ICER is negative; not cost-effective may also be suggested when the ICER is positive due to both a negative cost and effect that is, positioned in the Southwest quadrant of the cost-effectiveness plane, depending on the WTP threshold. As the stated WTP threshold is £20 000 per QALY gain, all positive ICERs due to positive costs and effects that are over £20 000 are also deemed not cost-effective. Also note that CIs were not reported as the analysis are deterministic and non-linear; therefore CIs could not be meaningfully interpreted.
§Average unadjusted EQ-VAS scores across baseline to day 7 are presented in the (online supplementary appendix). After adjustments for imbalance at baseline, the incremental effect was negative at −0.174 at day 7. The change in effect presented in the table above has been adjusted to represent an annual timeframe consistent with cost per QALY interpretation.
EQ-VAS, EuroQol-Visual Analogue Scale; ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life year; WTP, willingness to pay.