Table 22 Cost-effectiveness Outcomes: Mixed Studies
Authors, Year | Type of Analysis/ | Type of Costs | Costs | Type of effectiveness outcome | Effectiveness | ICER (definition); | ICER (results) | Sensitivity Analysis (definitions) | Sensitivity analysis (results) | Authors conclude (Reviewer comments where these differ from authors) |
|---|---|---|---|---|---|---|---|---|---|---|
Larsen et al. 2016 [65*] | CUA | Mean (SD) total costs (including healthcare, intervention and work absenteeism): | Intervention: €4212 (5931); Control: €5992 (7948) | General QoL | no persistent impact on general QOL | ICER (cost per QALY gained): | Intervention is dominant using DLQI | Bootstrapping Y/N (iterations): | Bootstrapping (1000 replications) | Cost effective |
Mean (SD) healthcare costs (excluding work absenteeism): | Intervention: €1606 (SD 1281); Control: €2708 (3928); | 15D (utilities) | no significant impact of intervention regarding QALY | WTP Threshold(s) | €62,500 for a health gain | |||||
Mean cost per participant for the intervention | €243 | ICER (cost per QALY gained): | Intervention is extendedly dominated using 15D | Probability cost-effective at WTP | a threshold value of zero, At a WTP threshold of zero, there was a 95% probability that MI was cost-effective. At the WTP threshold, 66.3% bootstrapped iterations were dominant. | |||||
Maes et al. 2014 [73*] | CUA | Total costs per person | Intervention: $7392; Control: $7035 | HUI | Intervention: 0.63 (baseline) to 0.65 (follow up); Control: 0.64 (baseline) to 0.61 (follow up) | ICER (cost per QALY gained): | $10,456 per QALY (health-care perspective); $24,580 per QALY (societal perspective) | Bootstrapping Y/N (iterations): | Y; (1000 replications) | Cost effective |
Healthcare costs | Intervention: $4034; Control: $3882 | |||||||||
Patient/family costs | Intervention: $106; Control: $135 | WTP Threshold(s) | $45,000 for a QALY gain | |||||||
Productivity losses | Intervention: $3252; Control: $3018 | Probability cost-effective at WTP | 68% from the healthcare perspective; from societal perspective 58% (52% in complete case analysis). | |||||||
Parry et al. 2012 [85*] | CEA and CUA | Costs: | Resource/service use indicators used as a proxy for costs. Text reports “no statistical differences between the groups on any of the service use indicators, although there was a slight increase in GP consultations in the CBT group during the treatment period” | Change in ASC panic-fear score | Intervention: −5.04 (SD 6.20); Control: −2.43 (SD 5.54) | ICER (cost per QALY gained): | Economic evaluation not conducted. No ICERs for change in ASC score. Data suggest intervention would be dominated due to EQ-5D scores being reduced in treatment arm. | No reported conclusion on cost effectiveness (unlikely to be cost-effective). | ||
Average cost of intervention (per participant) | £378-£798 | Difference in EQ-5D score between intervention and control group at 6 months (ANCOVA) | ITT analysis: −0.11 (95% CI: − 0.20 to − 0.03; p0.012); Complete case analysis: 0.12 (95% CI: 0.25 to 0.02); Intervention group had significantly lower scores. | |||||||
Rolving et al. 2016 [89*] | CEA and CUA | Average total costs at 12 months follow up: | Intervention: €52,492; Control: €52,580 | Change in ODI score at 3 months | Intervention: −14.8 (−18.7; −10.9); Control: −4.0 (−10.3; − 2.3); | ICER (cost per QALY gained): | Not reported as costs intervention dominant (costs less and more effective than control). | Sensitivity Analysis Used | different imputation strategies | Cost effective per QALY gained (for ODI it depends on society’s WTP for 15 point gain) |
Extra costs related to the intervention: | Intervention: €610 (production loss), €630 (intervention costs) and €116 (travel expenses); Control: N/A | Change in ODI score at 6 months | Intervention: −15.2 (− 18.8; − 11.6); Control: −8.4 (− 14.6; − 2.2) | Bootstrapping Y/N (iterations): | Y; 10,000 replications | |||||
Change in ODI score at 9 months | Intervention: − 14.9 (− 18.4; −11.5) Control: − 10.0 (−16.6; −3.3); | WTP Threshold(s) | €40,000 for one additional QALY, €10,000 per 15 point gain in Oswerty Disability Index (ODI) | |||||||
Change in EQ-5D utility score at 12 months from baseline [NB: may have been measured at other time points as reportedly only significant at the 3 month time point] | Intervention: 0.135; Control: 0.129 | Probability cost-effective at WTP | 70% per QALY gained, 90% per 15 point gain in ODI | |||||||
Change in QALYs at 12 months from baseline | Intervention: 0.71; Control: 0.636 | |||||||||
Tyrer et al. 2014 [94*] | CEA and CUA | Mean total health and social care costs over 24 months | Intervention: £7314.20; Control: £7727.40 | Mean (SD) improvement from baseline on the HAI | Intervention: 5·90 (7·54); Control: 3·66 (6·57) | ICER (cost per 1 point improvement in HAI scale) | £55.86 | Bootstrapping Y/N (iterations): | Y; (number of iterations not reported) | Unclear |
Mean costs (range) of the intervention (mean of 6 sessions) | Intervention: £421.51 (£0-£2383); Control: N/A | Mean gain in utility score (EQ-5D) from baseline at 24 months | Intervention: 0.085; Control: 0.065 | |||||||
Mean QALY gain from baseline to 24 months | Intervention: 1.108 QALYs; Control: 1.097 QALYs 95% CI: 95% CI is −0.091 to 0.087; p = 0.964. | ICER (cost per QALY gained): | £14,169 per QALY gained (however as QALY 95% Cis include zero there was no evidence on cost-effectiveness plane that CBT-HA is cost-effective in terms of health-related quality of life) | WTP Threshold(s) | £20,000-£30,000 per QALY gained | |||||
Probability cost-effective at WTP | The probability that the intervention is cost effective exceeds 50% if society’s willingness to pay for a 1 unit change in HAI is at least £53 or more. There is a slightly higher probability of standard care being more cost-effective than CBT-HA. This finding is due to variability in the data, resulting in wide confidence intervals, and very small differences in QALYs. | |||||||||
van der Aa et al. 2017 [96*] | CEA and CUA | Mean (SE) patient costs over 24 months | Intervention: €21,931 euros (€2035); Control: €22,808 euros (€2956). Mean difference not significant (−€1154; 95% CI − 7708 to 4328). | Mean change (SE) in HADS-A score; Difference between groups (95% CI) | Intervention: 1.88 (0.47); Control 0.45 (0.51); Mean difference 1.43 (95% CI 0.10 to 2.77) | Intervention dominant (ICER: − 613). Conclusions do not change with healthcare only perspective. | Sensitivity Analysis Used | Varying perspectives (healthcare only and human capital approach to include productivity losses); | Cost-effectiveness depends on willingness to pay threshold of decision makers | |
Mean change in CES-D score | Intervention: 6.40 (1.05); Control: 3.67 (0.99); Mean difference 2.73, 95% CI −0.28 to 5.74, not statistically significant. | Intervention dominant (ICER: − 321). Conclusions do not change with healthcare only perspective; | Bootstrapping Y/N (iterations): | Y; (5000 iterations) | ||||||
Incidence of depression/anxiety at 24 month follow up | Intervention: 0.29; Control: 0.46; Mean difference: 0.17 which was statistically significant (95% CI 0.06 to 0.29). | Intervention dominant (ICER negative: − 5159) indicating that to prevent one case of depression or anxiety €5159 is saved in the stepped-care group as compared to usual care. Conclusions do not change with healthcare only perspective; | WTP Threshold(s) | Change in score (HADS-A and CES-D):€0 - €4000; Per disorder prevented: €0 - €33,000; €0 - €20,000 per QALY gained | ||||||
QALYs gained (SE); Difference between groups (95% CI) | Intervention 1.32 (0.04); Control 1.28 (0.04); Mean difference 0.03 (95% CI −0.09 to 0.15), not statistically significant | Intervention dominant (ICER of −29,233 euros per QALY). Conclusions do not change with healthcare only perspective | Probability cost-effective at WTP | For the CES-D and the HADS-A, 60% at €0 per point improvement on the CES-D/HADSA; this increased to 95% or more at a WTP of €2500 per point improvement on the CES-D and €4000 per point improvement on the HADS-A; Per disorder prevented, 59% aat €0. At a WTP of €10,000 this was 77%, and at€20,000 it was 88%, and increased to 95% or more at a WTP of €33,000 per disorder prevented; Per QALY gained, 59% at a threshold of €0, this increased to 65% or more at a willingness-to-pay of €20,000 per QALY. |