Table 18 Cost-effectiveness Outcomes: Insomnia
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) |
|---|---|---|---|---|---|---|---|---|---|---|
Bonin et al. 2014 [37*] | CEA & CUA | Costs | Intervention: £251.00; Control: £72.00 | ISI scores (mean): | Intervention: Reduced by 17.6% (P < 0.001); Control: Reduced by 3.5% (NS: p0.077) | ICER (cost per 1 point improvement in ISI, per additional person in subclinical ISI, per QALY gained): | Not reported | Bootstrapping Y/N (iterations): | Y, 10,000 | Authors determined the cost-effectiveness was likely but depends on society’s WTP for ISI change. |
QALYs (mean): | Intervention 0.19; Control: 0.17 (QALY gain NS) | WTP Threshold(s) | £150 per 1 point improvement in ISI, £1800 per additional person in subclinical ISI state, £30,000 per QALY gained. | For cost per QALY gained the authors considered the cost-utility to be unclear but there is an error in the abstract (Not likely to have cost utility) | ||||||
Probability cost-effective at WTP | 97% for 1 unit improvement, 80% for additional person in subclinical ISI state, 34% per QALY gained (NB: misinterpreted in abstract). | |||||||||
Thiart et al. 2016 [93*] | CEA | Total employment costs per person at 6 months (absenteeism, presenteeism) | Intervention: €2527.47; Control: €2945.10; difference: €417.63 | Mean (SD) improvement on the ISI scale | Intervention: 9.3 (5.0); Control: 2.6 (4.4) | ICER (cost per every additional participant with a positive treatment response i.e. < 8 points on ISI and reliable change after 6 months) | -€1512 euros (95% CI: − 4493 to 1128). Intervention dominates | Parameters varied | Intervention costs (€100 and €300) | Cost effective |
Intervention costs (per person) | Intervention: €200; Control: N/A | NMB (benefits quantified in monetary terms minus costs of the intervention) | 417.63 (− 593.03 to 1488.70) | Bootstrapping Y/N (iterations): | bootstrapping method with 2500 replications | |||||
Benefit cost ratio (€ gains for every € invested) | 3.09 (−1.97 to 8.44) | WTP Threshold(s) | €0, €761, €1115 for a treatment response to per treatment response. | |||||||
Difference in presenteeism plus absenteeism between groups (€ - not including cost of intervention) | 617.43 | Return on investment (ROI i.e. [(benefit-cost)/(costs × 100)], 95% CI: | 2.08.81 (− 296.52 to 744.35) | Intervention dominates (81-91% of iterations in south east cost effectiveness quadrant); 60-72% | Base case: 87% at WTP of €0, 95% at WTP of €761; Sensitivity analysis (€300 intervention costs): 81% at WTP of €0, 95% at €1115. | |||||
Watanabe et al. 2015 [107*] | CUA | Direct costs | Intervention (CBT-I plus treatment as usual): $702 (SD 175) Control (TAU alone): $448 (SD 115) Non-significantly higher costs with an incremental value of $254 (SD: 203) | QALYs (literature derived and based on depression free days receiving a utility score of 1 otherwise major depressive disorder utility score of 0.59 was used depending on how patients had scored the 17-item HAND - less than 7 was deemed as remission from depression and severe depression was if their score was 27 or more). | Intervention: 0.139 (SD 0.004) QALY. Control: 0.120 (SD 0.004) QALY QALYs were statistically significantly higher (P = 0.002) in the CBT-I-plus-TAU group than in the TAU-alone group. | ICER (cost per QALY gained): | Base case: US$13,678 (95% CI: − 5691 to 71,316). Range in sensitivity analysis: US$5900 (95% CI:2485 to 14,958) to US$42929 (95% CI: 16994 to 163,146). | Parameters varied | Various different “approaches” used to test the results including and excluding hospital stays for depression, weighting utilities for severity versus looking at depression-free days and varying the costs of psychotherapy | adding CBT-I is highly likely to be cost-effective for patients with residual insomnia and concomitant depression (potentially problematic conclusion given we don’t know their methods for identifying literature values for utility with and without depression that were used to derive QALYs) |
Bootstrapping Y/N (iterations): | Y (1000 iterations). | |||||||||
WTP Threshold(s) | US$0 to US$100,000. Authors cite that one QALY is often valued at 50,000–70,000 USD | |||||||||
Probability cost-effective at WTP | 95% if a decision-maker was willing to pay 60,000 USD per QALY gained, and approximately 90% at 40,000 USD |