Cost-effectiveness of esketamine nasal spray compared to intravenous ketamine for patients with treatment-resistant depression in the US utilizing clinical trial efficacy and real-world effectiveness estimates

Brendle and colleagues frame treatment-resistant depression (TRD) as a substantial public-health and economic problem within major depressive disorder in the United States. They note that about one third of people with major depressive disorder meet common definitions of TRD and that this subgroup has higher rates of hospitalisation and suicide. Existing pharmacological and somatic therapies have limitations: many antidepressants take weeks to act, somatic therapies can be time consuming or have adverse effects, and there is an unmet need for more rapidly acting treatments. The study sets out to compare the cost-effectiveness of esketamine nasal spray (an FDA-approved, rapid-acting S-enantiomer formulation) with intravenous racemic ketamine (commonly administered off-label in clinics) for adults with TRD. Using both clinical trial efficacy estimates and real-world effectiveness (RWE) drawn from a private outpatient psychiatric clinic, the investigators aim to estimate incremental costs, quality-adjusted life years (QALYs), and incremental cost-effectiveness ratios (ICERs) from both a healthcare sector and a patient perspective over a multi-year horizon to inform decision making about these two treatment strategies.

Using a cohort Markov model with a 1-month cycle length, Brendle and colleagues simulated adults with TRD entering treatment at initiation of third-line antidepressant therapy. The model included four health states: TRD (starting state), response (defined as ≥50% improvement on validated depression scales such as MADRS, HAM-D or PHQ-9), non-response/relapse, and all-cause mortality. A 3-year base-case time horizon was employed to balance capturing accruing costs and benefits with available data; sensitivity analyses varied this from 1 to 5 years. Future costs and QALYs were discounted at 3% annually. The investigators followed CHEERS reporting guidance and ran analyses in TreeAge Pro Suite 2021, with meta-analyses conducted in STATA v16. Inputs combined efficacy estimates from phase 3 esketamine trials and published ketamine trials with patient-level RWE from a single private clinic that provides both therapies. Clinical-trial efficacy for esketamine came from four phase 3 trials (MADRS-measured change), converted to monthly probabilities for response and relapse (SUSTAIN-1 provided a relapse hazard ratio). Ketamine clinical efficacy was derived from a published IV-ketamine meta-analysis focused on TRD with response at 7 days; relapse rates were assumed equal to esketamine due to lack of ketamine relapse data. RWE cohorts included adults ≥18 years with recurrent MDD (ICD-10 F33.2) treated at the clinic; patients receiving other ketamine/esketamine formulations during the timeframe were excluded. Response rates in the RWE cohorts were 31% for esketamine and 33% for IM ketamine, with standard errors from binomial distributions. Cost inputs (2020 USD) incorporated medication WAC prices, administration costs, physician visit fees (CPT 99213 at $78), monitoring staff time, IV supplies for ketamine, and patient time valued by average hourly earnings ($29.37/h). For esketamine medication cost the RedBook WAC per 28 mg device ($324) was used; ketamine medication cost was $21 per 100 mg/mL vial as invoiced by the clinic. Base-case dosing schedules were informed by product labelling for esketamine and clinical consultation for ketamine: initial twice-weekly dosing in month 1 then weekly or twice-weekly maintenance depending on response state, with specified mg/dose assumptions. Two perspectives were modelled: a healthcare sector perspective including medical-system costs, and a patient perspective that replaced provider/medication costs with patient co-payments and included patient time. One-way sensitivity analyses and probabilistic sensitivity analyses (10,000 Monte Carlo simulations) were performed to assess parameter uncertainty; key parameters and distributions are reported in the paper's input table.

Using clinical trial efficacy inputs over a 3-year horizon, esketamine yielded 1.98 QALYs and IV ketamine 2.03 QALYs. From the healthcare sector perspective, total costs for esketamine exceeded ketamine by $176,320, driven predominantly by medication costs (esketamine medication: $179,204; ketamine medication: $2,905), so esketamine was dominated by ketamine (more costly and less effective). From the patient perspective—where esketamine medication costs were modelled as patient co-payments—esketamine was $42,532 less costly than ketamine; this produced an ICER of $867,606/QALY for ketamine compared to esketamine (reflecting small effectiveness differences and large cost differentials). When RWE effectiveness estimates were used, esketamine produced 1.98 QALYs and ketamine 1.99 QALYs over 3 years. Under the healthcare perspective esketamine remained more expensive by $172,919 and again was dominated by ketamine (medication costs: esketamine $177,369; ketamine $3,101). From the patient perspective esketamine cost $43,245 less than ketamine (medication co-payment estimates: esketamine $1,844; ketamine $45,090), yielding an ICER of $7,037,560/QALY for ketamine versus esketamine. Overall, clinical-trial results showed ketamine with +0.05 QALYs versus esketamine, whereas RWE showed +0.01 QALYs for ketamine; these small effectiveness differences combined with large cost differences produced very large ICERs. Sensitivity analyses showed that under the healthcare sector perspective no single parameter variation rendered esketamine cost-effective at a $150,000/QALY threshold. Under the patient perspective the most favourable one-way scenarios still yielded very high ICERs: the lowest attained was $464,389/QALY when applying the lower bound of esketamine co-payment (clinical-trial efficacy), and $712,747/QALY when applying the upper bound of ketamine response probability (RWE). Probabilistic sensitivity analysis indicated that, at a $150,000/QALY threshold, esketamine had essentially zero probability of being cost-effective versus ketamine from the healthcare perspective. From the patient perspective, the probability that esketamine was superior was 0.0055 with clinical-trial efficacy inputs and 0.35 with RWE inputs. Varying the time horizon between 1 and 5 years did not change the conclusion that esketamine was dominated under the healthcare perspective nor did it reduce patient-perspective ICERs below $150,000/QALY.

Brendle and colleagues interpret their decision-analytic modelling to indicate that esketamine nasal spray is unlikely to be cost-effective compared to IV ketamine at a commonly applied US threshold of $150,000/QALY when costs are viewed from the healthcare sector. The dominance of ketamine under that perspective was driven by large differences in medication costs despite similar effectiveness between the two treatments. Under a patient perspective—where insurance coverage and manufacturer copay assistance reduce out-of-pocket costs for esketamine—esketamine can be substantially less costly for patients, although ICERs remain large because overall effectiveness differences between therapies were minimal. The authors situate their findings relative to prior work: previous cost-effectiveness studies comparing esketamine to usual care or oral antidepressants similarly found esketamine unlikely to be cost-effective from a healthcare perspective at prevailing list prices, while an industry-sponsored analysis reported esketamine could lower cost-per-remitter versus oral antidepressants over a short horizon. Brendle and colleagues note that differences between their results and other reports partly reflect choice of comparator and perspective, and that their inclusion of ketamine as a comparator was enabled by available IV-ketamine trials and clinic-level RWE. They also suggest RWE narrowed the apparent efficacy gap because ketamine clinical trials are typically short (one week) whereas esketamine trials reported outcomes at 28 days; thus RWE may better reflect real-world effectiveness for both treatments. Key limitations acknowledged by the authors include small sample sizes in the underlying efficacy data (esketamine trials included a few hundred participants; IV ketamine trials totalled 118 patients), limited longer-term observational or trial data beyond two months, and the possibility that a 3-year horizon fails to capture longer-term costs and benefits. The analysis also excluded broader societal and family-level benefits such as effects on productivity or substance use, and it assumed consistent treatment patterns after relapse which may not reflect real-world clinical pathways. These uncertainties were tested in sensitivity analyses, and the authors report their main conclusions were robust to reasonable parameter variation. In terms of implications, the investigators conclude that cost-effectiveness depends strongly on perspective: payers and policy makers would need substantial price reductions for esketamine to reach value-based pricing from a healthcare sector viewpoint, whereas insured patients may find esketamine financially preferable due to coverage and assistance programmes. The authors recommend shared decision-making between clinicians and patients to weigh access, affordability, and expected benefit when choosing between these treatments.

In this decision-analytic comparison of esketamine nasal spray versus IV ketamine for treatment-resistant depression, Brendle and colleagues conclude that esketamine is unlikely to be cost-effective compared with ketamine from a healthcare sector perspective at a $150,000/QALY threshold. From a patient perspective—reflecting insurance coverage and manufacturer copay assistance—esketamine has similar effectiveness and can be substantially less costly out-of-pocket, making it more accessible for insured patients. The authors emphasise that shared decision-making will be important to determine which treatment patients can access and afford.