Informing evidence-based medicine for opioid use disorder using pharmacoeconomic studies

Abstract

Introduction:

The health and economic consequences of inadequately treated opioid use disorder (OUD) are substantial. Healthcare systems in the United States (US) and other countries are facing a growing healthcare crisis due to opioids. Although effective medications for OUD exist, relying solely on clinical information is insufficient for addressing the opioid crisis.

Areas Covered:

In this review, the role of pharmacoeconomic studies in informing evidence-based medication treatment for OUD is discussed, with a particular emphasis on the US healthcare system, where the economic burden is significantly higher than the global average. The scope/objective of pharmacoeconomics as a distinct scientific discipline is briefly defined, followed by a discussion of existing evidence informed by data from systematic reviews, in addition to a convenience sample of recently published pharmacoeconomic studies, and protocols. The review also explores the need for methodological advancements in the field.

Expert Opinion:

Despite the potential of pharmacoeconomic research in shaping evidence-based medicine for OUD, significant challenges limiting its real-world application remain. How to address these challenges are explored, including how to combine cost-effectiveness and budget impact analyses to address the needs of the healthcare system as a whole and specific stakeholders interested in adopting new OUD treatment strategies.

1.0. Introduction.

Opioid use disorder (OUD) is a complex chronic condition characterized by continued use of opioids despite known adverse consequences on an individual’s physical, psychosocial, and socioeconomic health and well-being.[1] Opioid use, particularly when used intravenously, is associated with increased risk of multiple co-morbid infectious diseases, such as human immunodeficiency virus (HIV), hepatitis C (HCV), and infective endocarditis.[2–4] Inadequate treatment of OUD, including early discontinuation, elevates the risk of fatal opioid overdose.[3] In the United States (US) alone, over 109,000 individuals died of a drug overdose annually since 2021, a two-fold increase from 2015, with approximately 76% involving opioids.[5] Recent increases in the opioid overdose mortality rate in the US, a primary indicator of the severity of the opioid crisis, has been fueled by the rapid proliferation of synthetic opioids,[6,7] and additionally compounded by the rising trend in co-use of stimulants such as methamphetamine among individuals with OUD.[8,9]

While OUD is a growing global crisis,[10,11] it is most severe in North America where the US and Canada face approximately 19 and 8 times the overall health economic burden compared to the global average, respectively, based on 2019 estimates of age-standardized disability-adjusted life years (DALYs).[12] The annual societal costs of OUD in the US is substantial,[13] and the health and economic consequences of OUD and related conditions will likely worsen as the epidemiological composition of new cases increasingly skew towards historically marginalized and minoritized populations who face structural barriers to accessing evidence-based medications.[14] Furthermore, the opioid crises has contributed to higher rates of newborns diagnosed with neonatal opioid withdrawal syndrome from prenatal opioid exposure,[15] and punitive policies for postpartum and pregnant persons who use opioids unnecessarily worsen outcomes and increase foster care burden.[16–19] These intergenerational impacts will likely lead to future adverse consequences on labor market outcomes and overall economic activity,[20] particularly in communities impacted by the successive waves of the opioid epidemic in the US.

Clinical trials and observational comparative effectiveness studies have demonstrated the value of medications for OUD in improving health outcomes and attenuating economic costs compared to non-pharmacological treatment.[11,21–23] Despite evidence of their effectiveness, initiation and adherence to evidence-based medications remain low for high-risk or vulnerable populations with OUD in the US,[4,24,25] such as those involved with the criminal legal-system,[26,27] individuals with low socioeconomic status,[14,28] and racial and gender minorities.[29,30] Persistent differences in the receipt and type of OUD medications,[31,32] as well as disparities in OUD outcomes between racial and ethnic populations may necessitate public health policies aimed at mitigating the potential roles of racism and bias in healthcare systems.[33–35] In addition, enrollment barriers in clinical studies limit evidence generation for optimal treatment modalities for high-risk OUD subpopulations,[36,37] some of whom rely on public resources to cover the costs of care.[3,38] Moreover, OUD and related conditions are associated with incidence of criminal activity, in part due to criminalization of drug use itself, which further strains public resources and therefore raises the need for identifying cost-effective interventions.[39] In the absence of direct pharmacoeconomic evidence, relying solely on clinical, comparative effectiveness research may be insufficient for key stakeholders responding to the opioid crisis (e.g., safety-net hospitals[40]) who face resource constraints for rapid adoption or investment in new treatment strategies for delivering evidence-based medications for OUD.

Accurate pharmacoeconomic assessments, which include evaluating the costs and cost-effectiveness of implementing and sustaining evidence-based medical interventions, are necessary for informed decision-making alongside pharmacologic assessments. The extent to which pharmacoeconomic research and measures, such as quality-adjusted life years (QALYs) and incremental cost-effectiveness ratios (ICERs), tangibly influence medical decision-making by providers and other stakeholders is subject to ongoing debate.[41–45] Nevertheless, results of pharmacoeconomic studies, including value of information from simulation studies and budget impact analyses integrated within clinical trial evaluations,[46] ensures that the health economic consequences of medical decisions and alternative courses of treatment are available to stakeholders during the decision-making process. In other words, pharmacoeconomic studies represent a necessary, though not solely sufficient, condition for informed decision-making.[47–50]

The aim of this review is to discuss how pharmacoeconomic studies can inform contemporaneous and emerging, evidence-based medicine for the treatment of OUD. Medications for the management of OUD are reviewed, highlighting the evolving clinical and pharmacoeconomic research targeting high-risk populations and high-need healthcare settings from a US perspective (e.g., carceral systems[51]). The scope and objective of pharmacoeconomics as a distinct scientific research program is briefly defined, followed by a review and discussion of existing evidence informed by data extracted from systematic reviews,[23,52–54] in addition to a convenience sample of recently published pharmacoeconomic studies, available protocols and opportunities for research associated with the National Institutes of Health (NIH) Helping to End Addiction Long-term (HEAL) Initiative.[54–57] This review is not intended to introduce research designs, measures, or outline analytic methods of pharmacoeconomic studies, which are available elsewhere in the literature.[58,59] While a general understanding of pharmacoeconomics is assumed, specific knowledge about OUD medications and their various formulations, such as daily oral intake and long-acting injectables, or the difference between opioid agonist and antagonist treatment, is not presumed. Thus, the intended audience is interdisciplinary health economists seeking to initiate a research career in pharmacoeconomic analysis of OUD treatment strategies.

2.0. Medications for opioid use disorder.

Effective management of OUD primarily involves evidence-based pharmacotherapies,[60] which include both opioid agonist and antagonist medications: methadone, buprenorphine (with or without naloxone), and naltrexone.[3,61–63] These medications are all recognized as highly efficacious (first-line) treatments for OUD,[64] offering superior efficacy compared to medication supervised withdrawal (i.e., opioid detoxification), and abstinence-based behavioral interventions alone.[3,4,62] Treatment initiation and engagement with pharmacotherapies, however, can be improved in conjunction with behavioral interventions,[65] contingency management,[66] and prescription digital therapeutics.[67,68] Medications that treat symptoms associated with withdrawal (e.g., lofexidine, clonodine) or opioid overdose reversal medications alone (naloxone) that are not then incorporated in evidence-based medication strategies for OUD are not considered in this review.

2.1. Opioid agonist medications (methadone, buprenorphine).

Opioid agonist medications, such as methadone and buprenorphine, bind to and activate opioid receptors, limiting the physical symptoms of opioid withdrawal, reducing cravings for opioids, and in turn lowering the risk of return to disordered opioids use. These medications have also been shown, across multiple studies, to reduce opioids overdose mortality by about half.[69] Opioid agonists are most effective in opioid maintenance treatment, historically known as opioid replacement or substitution therapy; however, these terms are now widely considered inaccurate and stigmatizing. They serve as a safer alternative to illicit opioid use, facilitating physical and psychosocial health and stability during the treatment period (in general, lifelong treatment), with managed tapering considered only for individuals exiting opioid agonist treatment.

Methadone, a full opioid agonist, is taken orally (tablet or oral solution) and is effective in reducing the risk of overdose mortality.[3,4,63] Methadone, while relatively inexpensive, is highly regulated in the US compared to other countries.[70] US regulation requires that methadone be dispensed in licensed opioid treatment programs (OTPs), a process which can constrain access by creating barriers to entry for providers to treat patients with methadone, resulting in an insufficient number of programs overall, and an uneven geographical distribution of available programs.[71,72] While an access point to methadone, primary care and other routine healthcare services are not necessarily integrated within OTPs—arguably a missed opportunity to improve the overall health status of individuals with OUD receiving methadone.[73] More liberal regulatory regimes are common outside the US, and are currently being considered by Congress. The evidence surrounding office-based opioid agonist treatment and pharmacy dispensing is limited, but suggests that adoption of these strategies may expand access without associated adverse health outcomes.[74]

In contrast, buprenorphine, a partial opioid agonist, can be prescribed by providers and dispensed in office-based settings. Until recently, a license to prescribe buprenorphine required a special “waiver” under the Drug Abuse Treatment Act of 2000 (DATA 2000) in the US. Known as an “X-Waiver” since the letter “X” is added to a provider’s Drug Enforcement Administration (DEA) Number, this step was perceived as a barrier to prescribing, and was eliminated as of 2021.[75] Buprenorphine is available in tablet, dissolvable film, and extended-release injectable or surgically implantable (no longer available in the US) formulations—these latter formulations designed to improve medication adherence.[63,76,77] Buprenorphine is available alone or co-formulated with opioid agonist naloxone (hereafter buprenorphine-naloxone); the combined formulation is thought to reduce buprenorphine’s potential for misuse without impacting effectiveness.[78] Transdermal formulations, while available, are generally restricted to treatment of pain and thus not discussed in this review. Buprenorphine treatment traditionally requires significant opioid withdrawal symptoms before initiation to avoid the risk of precipitated withdrawal—whereby the pharmacologic product rapidly displaces other opioids at the receptor and causes severe withdrawal symptoms. Emerging pharmacologic strategies, such as “low dose induction”, may allow for buprenorphine treatment initiation without the need for withdrawal, but clinical evaluations are limited.[79]

2.2. Opioid antagonist (naltrexone) and other medications.

Naltrexone, an opioid antagonist, blocks opioid receptors, preventing the euphorigenic and adverse pharmacological effects of opioids (e.g., respiratory depression) and return to opioid use. Naltrexone is an effective treatment pathway for OUD without the use of opioid agonists.[61,80] However, it requires individuals to have abstained from opioid use for 7 to 10 days prior to medication initiation due to the risks of precipitated withdrawal. In addition, injectable naltrexone cannot be self-administered, requiring frequent healthcare contact. The importance of successful engagement with the healthcare system to maintain continuous treatment has incentivized the development of interventions to improve care navigation and linkage at various points in the OUD care continuum.[81–83] Finally, though effectiveness outcomes related to abstinence are well-described, naltrexone has not shown a consistent mortality benefit,[21,84] possibly due to loss of opioid tolerance compared to treatment with opioid agonists, conferring an increased risk of overdose should individuals return to use. Extended-release formulations of naltrexone in combination with behavioral therapy and rapid initiation processes with ascending doses or with adjunctive opioid and non-opioid pharmacotherapies have been developed to address some of these challenges[85–88]; however, data remains at a preliminary stage.

Other pharmacologic treatments include a wider range of opioids in supervised administration, including hydromorphone, diamorphine (i.e., heroin), and slow-release morphine.[3] Fentanyl-assisted treatment is also being piloted as a potential treatment for individuals with severe OUD when evidence-based medication fail.[89] Utilization of such supervised medications are less common, and as of this writing, not approved by US regulatory authorities and not commonly in use in the US setting, though they are used in some international settings.[90–93]

3.0. Pharmacoeconomic research.

Pharmacoeconomics, a specialized branch of health economics, is concerned with evaluating the comparative clinical, person-centered (relevant to individual patient needs, preferences, and goals[94,95]), and economic (e.g., cost, resource utilization, employment/productivity) outcomes of medical interventions (often relative to standard care or treatment-as-usual) to inform healthcare decisions and policymaking. Naturally, the question arises: ‘Which healthcare decision-maker is being informed?’ The answer to this question—whether it be the healthcare system, third-party payer, society as a whole, or taxpayers—partially dictates data collection, price-weights, and analytic approach (e.g., study evaluation period) chosen by researchers. Pharmacoeconomic research has been defined rather consistently since the advent of the field with commonly used taxonomy of terms,[96] general evaluation methods,[59,97] and principles to demonstrate ‘value-for-money’ for healthcare systems and society while addressing potential biases and comparability issues through the development and periodic updating of reporting guidelines.[98–102] While a helpful guide for early-career researchers, the proliferation of pharmacoeconomic or cost-effectiveness guidelines in the field should not distract from careful consideration of the research design and statistical methods employed in individual studies. Prior definitions of pharmacoeconomics emphasized the analysis of pharmaceutical products and services to distinguish it from broader terms like health economic evaluation or health technology assessment, which can encompass a wider range of studies and health interventions.[103,104]

For the purposes of this review, the scope of pharmacoeconomics is defined as the evaluation of pharmaceutical products, including adjunctive health technologies (e.g., digital therapeutics), concomitant services (e.g., psychosocial counseling/support), and their optimal timing, dosage, and/or delivery setting. This scope includes examining delivery models and linkage to treatment strategies that may enhance the practical effectiveness of these products and services, with an underlying aim to guide resource allocation decisions in healthcare markets and systems. A less commonly conducted, though equally informative aim, is to assess the real and financial resources required to implement and sustain new or competing treatments (e.g., budget impact tools or tailored microcosting analyses). The latter generating data that can directly inform healthcare providers and administrators interested in adopting evidence-based medicine in their respective settings, accounting for the practical financial aspects of treatment delivery. A key limitation of budget impact analyses and related costing tools in informing decision-making, however, is their limited transportability of results to specific payers or providers, unless these tools are developed to be customizable by the end-user.

3.1. Informing medical decisions.

Real resource and budgetary constraints exist in the healthcare system and for payers and providers representing the interest of healthcare consumers and beneficiaries. Pharmacoeconomic analysis can be informative for such stakeholders when cost-conscious medical decision-making can improve the overall health of individuals by distributing healthcare resources more efficiently across the healthcare system’s catchment population—i.e., maximizing aggregate health while minimizing total cost. In resource constraint settings, medication treatment decisions entail trade-offs between obtaining health benefits and incurring costs associated with the utilization of resources (including direct and indirect non-medical resources) that could be employed elsewhere (i.e., opportunity cost). Therefore, a statistic that conveys the relative value of treatment options along with an explicit or presumed decision-criteria can help guide decision-making. For example, a treatment option has economic value for a decision-maker if the net health or monetary benefit is positive (η > 0), or that the incremental cost per additional unit of effectiveness (e.g., incremental cost per QALY) is less than or equal to a maximum value or preference threshold (e.g., λ ≤ willingness-to-pay per QALY). The latter determining whether allocating the next available dollar to a new intervention provides enough value to the decision-maker compared to treatment-as-usual (or the comparator case). Moreover, pharmacoeconomic studies also consider whether the expected budgetary impact of a new treatment is affordable given the local or global budget available to the healthcare system.

In this context, pharmacoeconomic studies inform medical decisions by collecting and analyzing essential data on both the cost and effectiveness outcomes needed to estimate the aforementioned statistics and generate uncertainty around such statistics with respect to a relevant decision-criteria. While the end result of such studies can be summarized within a simplified decision framework (η > 0, λ ≤ willingness-to-pay, etc.), the statistical methods employed to ensure valid and accurate results can often be complex,[27,105,106] requiring pharmacoeconomic specific methods and diagrammatic approaches to interpreting study results.[106,107] This decision-centered approach of reporting statistical results can be highly useful for stakeholders, and differentiates pharmacoeconomic studies from other health services research. However, clear communication of data and translation of results from pharmacoeconomic research into actionable strategies by stakeholders is necessary. As such, pedagogical dissemination and products are high-value activities for pharmacoeconomic researchers.

3.2. Pharmacoeconomic evidence and research priorities.

Data from chronologically continuous systematic review articles are consolidated and summarized along with a narrative review of recently published pharmacoeconomic studies and protocols of planned economic evaluations to provide an assessment of current evidence and emerging opportunities in the literature on medication treatment strategies for OUD. The systematic reviews included the following: Doran (2008),[52] which encompassed studies published up to 2007; Murphy and Polsky (2016),[23] which covered the period from 2007 to 2015; Onuoha et al. (2021),[53] which focused specifically on pharmacological interventions from 2015 to 2019; and Jalali et al. (2020),[54] which undertook a comprehensive review of 20 years of pharmacoeconomic research and planned/in-development studies within the National Institute on Drug Abuse (NIDA) National Drug Abuse Treatment Clinical Trials Network (CTN) since its inception (2000 to 2020). The inclusion of the CTN-focused review by Jalali and colleagues complements the three sequential systematic reviews, as it provides insights into ongoing and future pharmacoeconomic studies in the CTN, a collaborative network that focuses on conducting large-scale clinical trials to evaluate and improve the effectiveness of treatments for substance use disorders and related conditions across various healthcare settings.[108] In addition, NIH-HEAL initiatives supporting pharmacoeconomic research on evidence-based medicine for OUD treatment such as the Justice Community Opioid Innovation Network (JCOIN)[55,109] are included in the discussion, among other initiatives.[51,56,57] Together, these research infrastructures may help to address a significant priority for pharmacoeconomic research into OUD medication strategies, which is to inform medical decisions along the continuum of OUD care and for high-risk populations.[109,110]

3.2.1. Current evidence from systematic reviews.

After excluding detoxification interventions and non-pharmacologic OUD treatment strategies, a total of 85 pharmacoeconomic studies were evaluated by Doran (2008), Murphy and Polsky (2016), and Onuoha et al. (2021). These studies spanned 14 national healthcare systems, including Australia, Canada, China, Greece, Indonesia, Lithuania, Malaysia, The Netherlands, Russia, Spain, Ukraine, United Kingdom (UK, including England), the US, and Vietnam. A plurality of studies was based on observational cohort data, which included 36 primary and secondary analysis of retrospective, as well as single-arm prospective, evaluations. Twenty-four studies derived results from model-based methods (e.g., simulation, decision-analytic, Markov), and 21 based on randomized controlled trials. Only 4 pharmacoeconomic studies included in these reviews focused on budgetary impact analyses.

Doran’s 2008 systematic review was notable for the research community in that it identified a key issue in the literature prior to 2007: a disparity in the quality of evidence generated by pharmacoeconomic and pharmacological evaluations of OUD medication treatment strategies. According to Doran, pharmacological evaluations generally supported the efficacy of opioid agonist treatments over non-pharmacological and detoxification interventions, whereas the pharmacoeconomic literate, until then, revealed a “dearth of good-quality evidence” for OUD treatments.[52] Nevertheless, the review highlighted evidence suggesting methadone, especially when combined with adjunctive psychosocial services or contingency management-type incentive programs, could enhance economic value by improving treatment outcomes without incurring a large cost burden to the healthcare system. It also suggested that opioid agonist treatments, specifically methadone, in carceral settings might be as cost-effective as those in non-carceral settings, though more comprehensive research was required to substantiate those claims at the time.

Important shortcomings in the existing data (included in the 2008 review) was the lack of consistent methodological approaches in pharmacoeconomic studies of OUD treatments and too few focusing on high-risk populations of interest—the review specifically noting pregnant and adolescence populations. Doran recommended that future research explore treatment strategies across various settings, including primary care and criminal-legal systems, and emphasized the need to determine healthcare resource requirements for implementing interventions in these contexts. The review also noted that most studies adopted a limited cost perspective, primarily estimating costs from the healthcare system’s viewpoint. Only a minority of studies analyzed cost consequences comprehensively, with a major criticism being the exclusion of costs associated with criminal-legal system involvement. Since criminal-legal system costs can constitute a significant portion of societal expenses,[13] their inclusion will likely impact the overall estimate of an intervention’s value to society and perhaps taxpayers/policymakers. Doran further emphasized the need to expand pharmacoeconomic measures to include quality of life outcomes in order to calculate incremental QALYs associated with treatment, along with measures that can be used to inform issues related to affordability and equity.

Pharmacoeconomic evidence for opioid agonist treatment for OUD was expanded by subsequent studies from 2007 to 2015, with a growing consensus on the potential net cost-savings to the healthcare system when individuals with OUD were successfully inducted into evidence-based medications. Murphy and Polsky (2016) identified a number of healthcare resource utilization studies concluding that opioid agonist treatment with methadone and buprenorphine-naloxone can lead to significant cost-offsets for healthcare systems by trading increased pharmaceutical and ambulatory care costs near treatment initiation for reduced utilization of high-cost, downstream, inpatient and emergency care services. More recent secondary studies of clinical trials lend further support to these earlier findings.[111] However, much of these results were based on retrospective and observational research designs, which are often subject to well-known causal inference issues (e.g., confounding bias, depletion of susceptibles, etc.[112,113]). In addition, successful initiation itself has become an important area of ongoing research within the pharmacoeconomic literature for OUD.[114]

The cost-effectiveness or net-benefit evaluations in the 2016 review consisted of a larger number focusing on high-risk populations compared to the earlier review; for example, adolescents, and individuals with comorbid HIV.[115–117] Additionally, there was an increase in the number of studies that comparatively analyzed OUD medication treatment modalities with each other, such as buprenorphine versus methadone,[118] naltrexone (including extended-release formulation) versus methadone and/or buprenorphine and buprenorphine-naloxone,[119,120] and methadone versus injectable diamorphine.[121,122] Murphy and Polsky concluded in their review that there was sufficient data to consider opioid agonist treatment with methadone a cost-effective strategy compared to non-pharmacotherapy interventions within historically acceptable willingness-to-pay value ranges (from a societal perspective), but that comparative determination of which medication options were most preferred from an economic perspective was not yet conclusive. Variation in effectiveness measures, study evaluation period, and cost estimation from limited stakeholder perspectives remained a key shortcoming restricting the ability to “generate objective comparisons” across different studies.[23] Opioid antagonist treatment with naltrexone, and further evaluation of buprenorphine-naloxone treatment initiation strategies were underscored as important targets for further pharmacoeconomic research.

The pharmacoeconomic literature on evidence-based medicine for OUD had considerably improved in recent years, and studies have addressed some of the prior observed shortcomings. Studies reviewed by Onuoha et al. (2021), and those recently published are arguably more comprehensive compared to earlier contributions in the literature.[27,53,54,123] These works were more likely to include criminal-legal system associated costs of OUD treatment, and evaluated OUD treatment strategies in specific healthcare settings and among high-risk groups; such as treatment integrated within primary care or office-based settings,[124,125] initiated in emergency departments,[126] among inpatient residential treatment-seeking participants,[127] pregnant women,[128] interventions prior to reentry for individuals with OUD who are incarcerated,[27] as well as higher proportion of pharmacoeconomic studies reporting results using broader, person-centered outcome measures (e.g., QALYs and DALYs). The review by Onuoha et al. provided additional support for concluding that buprenorphine-naloxone, similar to the conclusion on methadone noted earlier by Murphy and Polsky, can be considered as broadly cost-effective compared to non-pharmacological treatment alone. While Onuoha and colleagues highlighted pharmacoeconomic studies evaluating new delivery mechanisms of OUD medications (e.g., subdermal implantable buprenorphine-naloxone[129]) and alternative opioid treatment options (injectable hydromorphone and diacetylmorphine[130]), inference on the comparative economic value of evidence-based medications, and treatment delivery/dosage formulations still remain inconclusive and an active area of research within the literature.[21,131,132]

3.2.2. Emerging opportunities and research priorities.

Expanding the use of medication for OUD, including personalized and targeted approaches for high-risk populations,[133] is an urgent clinical and policy goal in the US and other national healthcare systems dealing with a growing incidence of opioid overdose mortality and OUD prevalence.[3] Pharmacoeconomic research within the NIH-HEAL initiatives will likely play an important and growing role in providing funding opportunities to better inform evidence-based medicine to address the opioid crisis. Of the 138 NIDA CTN study protocols reviewed by Jalali et al. (2020), 14 were identified as incorporating a pharmacoeconomic study. While representing less than 10% of protocols at the time, the inclusion of such studies is a relatively recent phenomenon and have been growing by approximately 1 new pharmacoeconomic study in the CTN per year since 2007. Of the 14 pharmacoeconomic protocols, 8 were noted by Jalali et al. to be in the active or in-development stage at the time (the 6 completed studies were included in the 3 systematic review articles). The 8 pharmacoeconomic study protocols encompassed OUD intervention evaluations in diverse healthcare settings, including multi-site trial designs, and nearly all mentioned the use of trial instruments to generate qualify of life measures of effectiveness; for example, the NIH-supported Patient-Reported Outcomes Measurement Information System (PROMIS)-Preference scoring system (PROPr) to calculate QALYs.[134] The review of CTN protocols also highlighted the standardization of costing instruments for collecting the resources and associated costs of implementing the interventions as a major strength of trends in future/ongoing research. The inclusion of multiple stakeholder perspectives for reporting results once the CTN trials are completed was also noted as an important feature of future research, allowing for greater comparability between studies to generate conclusive statements for decision-makers.

A limitation of the pharmacoeconomic evidence reviewed by Onuoha et al. concerning OUD medication treatment strategies for high-risk populations and specific healthcare settings was the reliance on decision-analytic modeling as the primary method of analysis for many studies. A comparative advantage of trial-based or trial-integrated cost-effectiveness analyses is their ability to estimate, with minimal bias, the complier average causal effect (also referred to as local average treatment effect in the econometric literature) and observe the correlation between cost and effectiveness outcomes at the participant level directly, thus generating more reliable, causal conclusions to inform decisions. As comparative effectiveness evaluations of competing evidence-based medications for OUD are tested in clinical trials, integrating well-designed pharmacoeconomic studies alongside these trials in collaboration with clinical research teams is an important avenue for future research in closing the aforementioned evidence gap observed by Doran in 2007. Active and in-development protocols of clinical trials, consistent with research and institutional practice, are publicly available (see, for example, the NIDA CTN Dissemination Library) at early stages of the study, often prior to enrollment. Engaging with relevant clinical teams and leveraging existing research infrastructures and opportunities should be a continuous priority for pharmacoeconomists.

However, participant randomization designs employed in the CTN and other NIH-HEAL initiatives can be complex (e.g., multi-stage, sequential, and/or adaptive randomization), and pragmatic trials that source data from electronic health databases raise issues related to possible data discontinuity bias, i.e. measurement error when healthcare is consumed outside the system that captures primary data.[135] Jalali et al. anticipated the need for researchers to study and validate the most appropriate econometric research designs for future pharmacoeconomic studies, and argued for increased attention in this domain as pharmacoeconomic studies are increasingly integrated in conjunction with clinical trials.

Notwithstanding the need for econometric methods research, the integration of pharmacoeconomic research alongside or in conjunction with clinical trials is an important research priority for the field as it allows for collecting highly robust, participant-level measures of data to inform evidence-based medicine for OUD. Planned pharmacoeconomic evaluations in other NIH-HEAL initiatives such as JCOIN,[136] represent a potentially ideal scenario for future research. Such evaluations will address some of the issues from prior research covered in the systematic reviews: the lack of comparability of data to generate overall conclusions, the focus on high-risk populations within the care continuum, and integrating budget impact or tailored microcosting analyses with cost-effectiveness analyses to address questions regarding affordability and sustainability of medication strategies. JCOIN is currently planning 10 pharmacoeconomic studies, specifically targeting the high-risk group of criminal-legal system-involved individuals with OUD in the US. The objective is to evaluate the comparative and cost-effectiveness of various medication strategies, as well as to facilitate linkage to treatment. Moreover, these studies will examine medication strategies at multiple points in the criminal-legal system continuum, including 1 study examining linkage to treatment in a judicial supervisory setting (e.g., drug court), 3 studies focusing on pre-release medication treatment, 4 studies on linkage to treatment and retention post-release, and 2 studies on linkages to treatment in community supervision settings (e.g., probation and/or parole).[55] Furthermore, JCOIN economic studies are being developed collaboratively among the economic researchers with published pre-study protocols and significant investment in developing harmonized trial instruments and measures,[55] including a modified version of PROMIS-PROPr for criminal legal system involved populations, following earlier published protocols among this OUD population and for opioid misuse.[51,137] An advantage of data harmonization is that it will allow for pooling of pharmacoeconomic data across trials to answer questions and utilize methods that require larger sample sizes (e.g., latent class analysis[22]), or more appropriately address potential problems in conducting statistical inference when trial participants are lost to follow-up.[105] However, there is currently little research guidance on conducting cross-site and cross-setting pharmacoeconomic evaluations, and generalization of pharmacoeconomic data from JCOIN will be dependent on individual studies achieving enrollment targets and randomization of representative OUD populations.

Finally, identifying the most economically efficient approaches to treating OUD does not address the growing opioid crisis without efforts to curb prevention. Prevention and community level interventions are an important and growing area for pharmacoeconomic research to address OUD in the US. The NIH-HEAL Initiative has encouraged pharmacoeconomic analyses within the HEAL Prevention Cooperative (HPC) and the HEALing Communities Study (HCS).[56,57] To comprehensively address the opioid crisis in the US and North America, and to slow its progression globally, pharmacoeconomic research and preventative-effectiveness research in the HPC and HCS should advance in tandem with the CTN. This coordinated approach is vital to inform evidence-based medicine and evidence-based prevention strategies.

4.0. Conclusion.

Medical decisions that rely solely on information on the efficacy of OUD medication treatment strategies without reference to their health economic consequences limit the ability of healthcare systems to address the growing opioid crisis and expand care access to vulnerable, high-risk populations. Pharmacoeconomic studies provide, for stakeholders, necessary information for informed decision-making in resource constraint settings—though such information is not solely sufficient for decision-making. Opioid agonist medications such as methadone, buprenorphine and/or buprenorphine-naloxone are cost-effective treatment options compared to non-pharmacologic interventions based on review of current evidence in the pharmacoeconomic literature. This conclusion has been supported both by prior, robust trial-based findings, and recent data generated from comprehensive simulation studies.[123] However, the sub-optimal utilization of these medication options, along with observed barriers to initiating and maintaining treatment for individuals with OUD and associated comorbidities (HIV, HCV, mood disorders, etc.), has rightly prompted the addition of pharmacoeconomic studies alongside new comparative effectiveness clinical trials. These trials, if successful, are expected to generate data for identifying the most effective and cost-effective medication delivery formulations, settings, and to which high-risk populations. Primary limitations noted in this review, such as inconsistent outcome measurement across pharmacoeconomic studies, exclusion of criminal-legal system costs, and limited number of stakeholder perspectives, have been progressively addressed over time. In addition, the recognition of incorporating pharmacoeconomic assessments of the sustainability of implementing medication treatment strategies in different healthcare settings through tailored microcosting analyses to directly inform the budgetary interests of providers has been an explicit feature of recent NIH-funded research initiatives.

The future of pharmacoeconomic studies of evidence-based medicine for OUD is promising; it is expected to emphasize prevention and community level interventions alongside treatment across the OUD care continuum to address the opioid crisis more comprehensively. Moreover, recent efforts to harmonize datasets and pharmacoeconomic measures across multiple clinical trial collaboratives will provide expanded opportunities for future research to leverage larger sample sizes and conduct pooled, cross-study evaluations. To ensure that evidence generated from planned and ongoing pharmacoeconomic studies provide robust, causal conclusions for decision-makers, the need for methodological research to accommodate complex clinical trial randomization designs is anticipated.

5.0. Expert opinion.

As noted earlier, the influence of pharmacoeconomic research in informing real-world clinical practice is subject to debate. This review expounded on the use-value of such research in informing evidence-based medicine for OUD, but there are significant barriers outside of methodological and research design issues that reduce their impact. For instance, legal structures regulating national healthcare payment schemes can be hostile to the application of pharmacoeconomic evidence in payment/reimbursement decisions, e.g., the non-interference clause of Medicare (Part D) in the US—though this was recently revised for a limited set of medicines (see Wong (2014) and Darrow et al. (2020) for a historical perspective on US pharmaceutical regulatory history[138,139]). The fragmentation of healthcare delivery and third-party payers in the US can also make it difficult to ensure pharmacoeconomic studies generate data that are representative of the healthcare system as a whole in determining cost-effective national treatment strategies, while simultaneously being informative for healthcare providers weary of the financial uncertainty of adopting new treatment strategies in their specific settings. Implementation scientists have explored the translation of pharmacoeconomic assessments of treatment interventions for real-world application in a variable payment landscape for stakeholders,[140,141] which can be further bolstered in pharmacoeconomic studies by aligning budget impact or tailored microcosting analyses with the implementation frameworks commonly used in OUD clinical trials. One example is the planned pharmacoeconomic evaluation of a hospital-inpatient addiction consultation and linkage to OUD treatment intervention where the budget impact analysis will disseminate a tailored microcosting tool that measures the resources and calculates the associated cost required to sustain the intervention for each component of the Reach, Effectiveness, Adoption, Implementation, and Maintenance (RE-AIM) framework,[83] the most widely applied framework for measuring the impact of an intervention in real world environments.[142] The concept of combining pharmacoeconomic analyses that assess cost-effectiveness from a broader stakeholder perspective coupled with information (or tools) for evaluating sustainability of providing care is not new.[46] However, few pharmacoeconomic studies of evidence-based medication strategies have undertaken this approach. While some of these budget impact tools for OUD have been published,[143] it is important that their dissemination be prioritized outside of traditional academic journals when completed since the target audience is not necessarily affiliated with academic institutions, e.g., in open-access research distribution services and archives (e.g., arXiv, medRxiv, SSRN:MedRN).

A promising area of clinical research for OUD medication strategies is technology-based interventions that aim to promote more effective and equitable access to medications for high-risk populations. Telehealth for OUD care has been expanded during the Covid-19 pandemic through emergency authority,[144] and recent treatment innovations using mobile-based strategies for improving health (often referred to as mHealth) and treatment outcomes have become popular in the addiction literature.[145–148]. However, the financial prospects of continued mHealth innovations are limited since these types of services do not generate revenue for providers through existing fee-for-service mechanisms in the US. Episodic and bundled payment models, as suggested by researchers,[149] may be an appropriate funding mechanism to accommodate these innovations, and budget impact analyses could be key sources of data for informing such payment models by policymakers.[149]

Individuals with OUD fall into a number of overlapping clinical populations, providing pharmacoeconomists with additional opportunities to conduct high priority research to inform medical decision-making.[150] For example, individuals with disordered use of prescribed opioids for pain may have different treatment needs or responses to OUD medications compared to individuals who use recreational drugs such as heroin and fentanyl, despite the same overarching diagnosis. Prior pharmacoeconomic assessments of OUD treatments may not be generalizable to these populations given the variation in costs, unique medication formulations specific for pain (e.g., compounded buprenorphine nasal spray for pain and off-label comorbid OUD and pain treatment), and complexities involved in the management of chronic pain among individuals with comorbid OUD and chronic or cancer-related pain.[151]

It is important to note that ‘real world’ effectiveness and cost-effectiveness of medications for OUD when compared with each other can vary broadly not only by setting, but also by country. For example, comparative effectiveness of opioid agonist treatment with methadone compared to buprenorphine can be different in the US compared to other countries given differences in regulatory structure affecting ease of access.[70,152] Differences in the provision of social services will also significantly impact cost estimates in pharmacoeconomic studies when such data is included in, for instance, a taxpayer perspective. It is therefore the contention of the author that while evidence from pharmacoeconomic studies should, in theory, be invariant across countries, the complexities of healthcare systems, their financing structure, payer and healthcare system revealed preferences’ of value (i.e., willingness-to-pay),[153] including variation in national pharmacoeconomic evaluation guidelines,[154] make synthesis relatively impractical. Consequently, this challenge might be recognized as a general principle in pharmacoeconomics in the context of OUD treatment. While cross-country synthesis of pharmacoeconomic studies in OUD treatments is challenging and potentially impractical, such studies still hold value, but should be interpreted with an understanding of the underlying variability and context-specific factors that influence their outcomes.

Continued investments by federal agencies to advance pharmacoeconomic research is critical to generate causal conclusions to inform evidence-based medicine for OUD. As discussed in this review, the NIH-HEAL initiative has championed this critical research need by funding pharmacoeconomic analysis alongside clinical trials in multiple research infrastructures. However, funding should be expanded to be more favorable to methodological research, which is currently a high-risk activity for pharmacoeconomists dependent on extramural support for career advancement. Nevertheless, high-risk research activity focused on advancing methodologies can yield significant value for both patients and science. Perhaps the development of ‘pharmacoeconometrics’ as a new empirical field is warranted.

Article Highlights.

• Pharmacoeconomics is a specialized branch of health economics concerned with evaluating the comparative clinical, person-centered, and economic outcomes of medical interventions, including adjuvant treatments, products, and services.

• Data generated from pharmacoeconomic studies, such as cost-effectiveness and budget-impact analyses, provide necessary, though not solely sufficient, information for medical decision-making.

• Systematic reviews have determined that opioid agonist medications, such as methadone and buprenorphine, are cost-effective for treating opioid use disorder (OUD) compared to non-pharmacological interventions. However, additional research is needed to identify the optimal medications, formulations, delivery methods, healthcare settings, and treatment linkage strategies for high-risk, vulnerable subpopulations with OUD.

• Continued investment by federal agencies in supporting pharmacoeconomic research conducted alongside or in conjunction with clinical trials is critical for advancing the field. These studies are the primary sources of robust, participant-level health economic data needed to inform evidence-based medicine for OUD.

• There is an anticipated need for researchers to study and validate the most appropriate econometric research designs for future pharmacoeconomic studies. Therefore, federal funding should be expanded to be more favorable to methodological research proposals.