Cost-Effectiveness of Medically Monitored Ibogaine for Opioid Use Disorder in Tennessee: A State-Specific Semi-Markov Analysis

Opioid use disorder remains a major cause of preventable illness, death, and healthcare use in the United States, and many adults still do not receive medication treatment or remain poorly retained in care. The authors note that this treatment gap has both clinical and economic consequences, particularly in Tennessee, where OUD burden and overdose-related hospital use remain substantial. Although established treatments such as methadone, buprenorphine, and extended-release naltrexone are known to reduce overdose risk and mortality, the state treatment environment is heterogeneous, with many patients untreated or not engaged in care. Previous economic studies have shown that the value of OUD treatment depends strongly on retention, reduced acute-care use, and the balance between upfront treatment costs and later savings. However, there has been limited state-specific economic evidence on how a medically monitored ibogaine pathway might perform within Tennessee’s existing treatment system. The study set out to evaluate whether adding medically monitored ibogaine to heterogeneous standard care in Tennessee could improve health outcomes and reduce costs over time. The authors aimed to model a realistic limited-access ibogaine pathway, not unmonitored ibogaine use, and to compare it with a matched no-ibogaine standard-care reference that reflected the state’s existing mix of treatment and non-treatment pathways. They hypothesised that any benefit would mainly arise through longer-term effects on relapse, hospitalisation, disengagement from care, stable recovery, and mortality rather than through short-term detoxification effects alone. The paper presents a Tennessee-specific decision-analytic economic evaluation intended to inform trial design, implementation planning, and policy discussions rather than to establish ibogaine as routine care.

The authors developed a cohort state-transition model with semi-Markov features to estimate the long-term health and economic effects of adding a medically monitored ibogaine pathway to Tennessee’s opioid use disorder treatment environment. The analysis compared a limited-access ibogaine strategy with a matched no-ibogaine reference. It was framed as a state-specific policy analysis rather than a clinical trial simulation. The primary cost endpoint was cumulative direct medical cost from a healthcare-sector perspective, and the primary health outcomes were quality-adjusted life-years (QALYs), survival, and mortality. Blended total cost, which included indirect costs, was tracked as a secondary outcome. The model was reported as aligned with CHEERS 2022 guidance for economic evaluations. The base-case cohort represented 106,000 Tennessee adults with past-year OUD, based on the 2023-2024 National Survey on Drug Use and Health estimate. The authors also tested denominator sensitivity scenarios using the lower and upper credible-interval bounds and a historical 2022-2023 Tennessee estimate. Tennessee overdose mortality and hospital-discharge data were used as contextual and benchmark inputs rather than direct substitutes for all transition probabilities. The intervention was defined as medically monitored ibogaine within an OUD care pathway, including protocolised administration, cardiac screening and monitoring, safety-escalation capacity, post-acute stabilisation, and follow-up. The comparator was a heterogeneous no-ibogaine basket including methadone, buprenorphine, extended-release naltrexone, non-MOUD treatment, and untreated or not-engaged care. The Tennessee overlay assumed a baseline routing mix of 5.0% methadone, 22.0% buprenorphine, 1.5% extended-release naltrexone, 23.5% non-MOUD treatment, and 48.0% untreated or not engaged, with both arms beginning from the same baseline allocation. The model used weekly cycles over a 10-year horizon, with costs and QALYs discounted at 3% annually. It contained 15 health states and 36 transitions, with tunnel states and time-in-state logic to reflect time-dependent phases such as treatment induction, post-discontinuation risk, monitored ibogaine administration, safety escalation, post-acute stabilisation, early remission, hospitalisation, and relapse. Death was modelled as an absorbing state. Mortality included both background risk and state-specific excess risk, with treatment retention generally protective relative to untreated or recently discontinued care. The authors assigned costs and utilities at the state level and through transition-specific events. Costs included standard-care treatment, hospitalisation and overdose-related acute care, and ibogaine program delivery, monitored care, safety escalation, and aftercare. All cost inputs with source-year information were inflated to 2024 U.S. medical-care dollars. Utilities were based on health-related quality of life in active OUD, treatment engagement, unstable states, remission, and recovery. Input uncertainty was handled using a structured evidence and provenance registry, with parameters classified as literature-based, calibrated, registry-derived, localisable, expert-assumption, or overlay-specific. One-way deterministic sensitivity analysis, probabilistic sensitivity analysis with 5000 simulations, scenario analyses, and threshold analysis were all prespecified. The scenario analyses examined waning efficacy, comparator-mix variation, a conservative stress test, retreatment or re-engagement, short horizon, and denominator sensitivity. The probabilistic analysis varied uncertain parameters jointly across their distributions and evaluated willingness-to-pay thresholds of $50,000, $100,000, and $150,000 per QALY. The model was also checked against calibration targets, benchmark comparisons, and internal validation criteria, including structural consistency, cohort balance, and bounded survival and mortality.

In the 10-year Tennessee base case, the medically monitored ibogaine strategy was associated with lower direct medical cost, higher QALYs, and better survival than the no-ibogaine reference. Cumulative direct medical cost was $5.46 billion with ibogaine and $6.42 billion without it, for an incremental direct-cost difference of -$965.0 million. Total QALYs were 627,371 versus 530,144, an incremental gain of 97,227 QALYs. Ten-year survival was 0.9308 with ibogaine and 0.9098 without, an absolute difference of 2.10 percentage points. Mortality was correspondingly lower in the ibogaine arm, at 0.0692 versus 0.0902. Because the intervention was both less costly and more effective, it was dominant in the base case. Blended total cost, which included indirect costs, was also lower with ibogaine: $17.28 billion versus $27.45 billion, a difference of -$10.18 billion. The mechanistic output was broadly consistent with these aggregate results. The ibogaine arm showed more cumulative stable-recovery entries, fewer hospitalisation entries, and fewer deaths over the model horizon. Stable-recovery entries were 152,492 with ibogaine versus 35,604 without. Hospitalisation entries were 57,581 versus 91,667, and cumulative deaths were 7,338 versus 9,565. These were cumulative transition counts rather than unique individuals, because repeat entries were allowed in the semi-Markov structure. The baseline Tennessee treatment-routing mix was the same in both arms, and most ibogaine entries originated from untreated active use, with smaller contributions from non-MOUD treatment and recent dropout states. Scenario analyses suggested that the overall direction of the result was robust to waning efficacy and comparator-mix variation, but more sensitive to pessimistic assumptions about durability, comparator performance, and programme cost. Under waning efficacy, ibogaine remained less costly and more effective, with incremental direct medical cost of -$152.7 million and a QALY gain of 79,539. Under comparator-mix variation, results stayed close to the base case, with incremental direct medical cost of -$956.2 million. In the conservative stress test, the direct-cost advantage reversed to +$232.7 million, but ibogaine still produced 68,547 additional QALYs and a survival gain of 0.012. The retreatment or re-engagement scenario had minimal effect, with $3.1 million in cost savings and 314 extra QALYs. The short-horizon scenario showed the expected front-loaded cost pattern: ibogaine increased direct medical cost by $13.8 million and produced only 2.7 additional QALYs, indicating that the modelled value emerged over longer follow-up rather than immediately after administration. One-way deterministic sensitivity analysis identified ibogaine programme cost, late-remission relapse, and entry from untreated active use into monitored ibogaine as the most influential parameters. Varying programme cost moved incremental direct medical cost from -$1.89 billion to +$940.6 million, while QALYs were unchanged because only cost was varied. Late-remission relapse shifted incremental direct medical cost from -$1.95 billion to -$393.5 million and QALYs from 109,702 to 72,142. Entry from untreated active use into monitored ibogaine had the largest impact on QALYs, shifting them from 46,054 to 134,802. Other influential inputs included post-acute transition to early remission and monitored safety escalation. In probabilistic sensitivity analysis with 5000 simulations, the mean incremental direct medical cost was -$465.9 million and the median was -$573.3 million; the 2.5th to 97.5th percentile range was -$1.85 billion to +$1.41 billion. Mean incremental QALYs were 93,256, with a median of 91,728 and a 95% interval of 46,964 to 147,408. Across simulations, 72.9% were direct-medical-cost-saving and 100% produced positive incremental QALYs. The strategy was cost-effective in 99.98%, 100%, and 100% of simulations at willingness-to-pay thresholds of $50,000, $100,000, and $150,000 per QALY, respectively. Threshold analysis showed that the economic result was especially sensitive to ibogaine programme cost. At $3,000 per treatment episode, direct medical cost savings were $1.89 billion; at $7,000, savings were $948.4 million; at $11,000, the strategy was approximately cost-saving; and at $15,000, direct medical cost became positive at +$940.6 million. QALYs remained unchanged at 97,227 across this analysis because only cost was varied. Net monetary benefit remained positive across the explored range at a willingness-to-pay threshold of $100,000 per QALY.

The authors interpret the findings as suggesting that a medically monitored ibogaine pathway could plausibly reduce relapse, hospitalisation, disengagement from care, and mortality enough to offset substantial upfront implementation costs in Tennessee’s high-burden OUD treatment environment. They emphasise that the base-case result was economically favourable because ibogaine was both less costly and more effective than the no-ibogaine reference, but they are careful not to claim that it would be cost-saving under all assumptions. They argue that the key pattern is temporal: in the short-horizon scenario, costs rose and health gains were small, whereas the longer-term base case showed benefits emerging over time as more patients entered remission or stable recovery and fewer entered hospitalisation or death. They place these findings in the context of the broader OUD treatment literature, which they say shows that economic value depends more on retention, reduced acute-care use, overdose protection, and sustained recovery than on isolated detoxification effects. They also stress that the comparator was a realistic Tennessee treatment basket rather than untreated care or an optimised buprenorphine-only arm. In their view, the relevant policy question is whether an additional monitored pathway could improve outcomes for people who remain untreated, undertreated, or poorly retained in care, not whether ibogaine should replace established medications such as methadone or buprenorphine. The authors say the scenario analyses show that the result is reasonably robust to comparator-mix variation and waning efficacy, though more vulnerable to pessimistic assumptions about durability, comparator performance, and monitored-care cost. They note that the conservative stress test reversed direct-cost savings but still produced substantial QALY gains and improved survival, which they regard as a meaningful policy signal. They also highlight that cardiac toxicity is a central boundary condition: the intervention modelled here included screening, monitored administration, safety escalation, and follow-up, and they argue that these features are integral to the intervention rather than optional add-ons. A cheaper, less monitored version would be a different intervention and should not inherit the present results. Threshold analysis reinforced the importance of implementation cost, with programme cost emerging as one of the main drivers of economic results. The authors interpret this as showing that the central uncertainty is not only whether ibogaine can help clinically, but whether medically monitored delivery can be organised at a cost compatible with Tennessee’s healthcare priorities. They present the findings as especially relevant because Tennessee and federal policymakers are considering structured pathways for psychedelic and ibogaine-related research. However, they state explicitly that the model does not establish ibogaine as routine OUD care. Several limitations are acknowledged. The ibogaine evidence base remains limited, observational, and protocol-sensitive. Some Tennessee pathway shares were assumption-based, because there is no definitive public source for the full treatment mix. Several costs, including programme, monitoring, escalation, and hospitalisation costs, remain localisable and implementation-sensitive. Calibration and benchmarking were supportive but not perfect: the model reproduced some monitored-safety targets, but the 1-month opioid-free share remained above the calibration target, and some external benchmark comparisons showed residual deviation, including long-horizon survival. The authors kept these discrepancies rather than forcing a tighter fit, presenting them as markers of residual uncertainty. They conclude that the model supports further prospective evaluation of medically monitored ibogaine in Tennessee, ideally with embedded clinical, safety, utilisation, and economic data collection. They suggest future work should compare implementation designs, identify subgroups most likely to benefit, and evaluate whether ibogaine works best as an adjunctive pathway, a rescue pathway after treatment discontinuation, or a bridge into ongoing evidence-based OUD care.

The authors conclude that, in this Tennessee-specific semi-Markov cost-effectiveness analysis, adding a medically monitored ibogaine pathway to heterogeneous standard care was associated with lower direct medical cost, higher QALYs, and improved survival over 10 years. They describe the base-case strategy as dominant, while noting that the size and even the direction of economic benefit depended on assumptions about programme cost, durability, uptake, and comparator performance. Their final implication is not that ibogaine is ready for routine use or should replace established OUD medications. Rather, they argue that a carefully monitored Tennessee ibogaine clinical-trial or implementation pathway could be economically plausible if it improves sustained recovery and reduces downstream hospitalisation, relapse, and mortality. Because cardiac monitoring and escalation capacity are built into the modelled intervention, they say future programmes should collect prospective safety, clinical, utilisation, and cost data to test whether the modelled benefits are realised in practice.