Coronavirus disease 2019 (COVID-19) has had a profound global impact, resulting in over 270 million cases and more than 2 million deaths across Europe as of January 2023. The pandemic significantly strained healthcare systems, with hospitalizations driving up resource use. For example, the UK government has allocated approximately £358 billion to combat the pandemic’s effects, while the International Monetary Fund estimated the worldwide economic toll could reach $13.8 trillion by 2024. Even though the emergency phase has passed, emerging variants like Omicron continue to threaten vulnerable populations, including older adults and those with compromised immune systems or preexisting conditions.
Nirmatrelvir/ritonavir (NMV/r), an oral antiviral, received emergency authorization in late 2021 and full approval from regulatory bodies such as the European Medicines Agency and the U.S. Food and Drug Administration by 2022–2023. It is indicated for mild-to-moderate cases in individuals at high risk of developing severe illness. Clinical evidence shows it reduces hospital admissions and mortality, with real-world data indicating a 79.6% effectiveness in preventing death or hospitalization when administered within five days of symptom onset. Given its potential to alleviate health and economic burdens, assessing its cost-effectiveness is crucial.
This systematic literature review analyzed 22 economic evaluations conducted between January 2022 and September 2024, following PRISMA guidelines. Studies were sourced from databases including Embase, MEDLINE, EconLit, and the Cochrane Library, along with conference abstracts and health technology assessment agency reports. The analysis focused on cost-utility, cost-effectiveness, budget impact, and cost-consequence models across various countries.
Findings revealed that NMV/r was generally cost-effective compared to standard or supportive care in most regions, particularly when long-term outcomes such as reduced hospitalizations and mortality were considered. Ten studies used cost-utility analysis, nine applied cost-effectiveness methods, two developed budget-impact models, and one employed a cost-consequence framework. Most models adopted healthcare system or societal perspectives, with time horizons ranging from one month to a lifetime. A majority of cost-utility analyses found favorable incremental cost-effectiveness ratios relative to local willingness-to-pay thresholds.
In the U.S., budget impact models projected significant savings—up to $2.7 million annually for a one-million-member insurance plan—due to fewer hospitalizations. Including post-COVID conditions further enhanced cost savings. In China, NMV/r showed cost-effectiveness primarily among patients aged 80 and above. Conversely, in the Netherlands, price adjustments were deemed necessary to achieve cost-effectiveness, while in Ghana, Rwanda, and Zambia, results varied by age group and risk profile.
Sensitivity analyses highlighted that treatment cost, efficacy, and hospitalization rates were key drivers of model outcomes. Accounting for long-term sequelae like long COVID improved economic value estimates. Early treatment initiation and inclusion of societal costs also strengthened cost-effectiveness conclusions. However, inconsistencies in reporting—such as missing time horizons, discount rates, or detailed input sources—limited comparability across studies.
Using the Drummond checklist, 59% of the evaluations were rated high quality, with clear research questions and appropriate methodologies. Nevertheless, gaps remain, including limited data on transmission reduction, incomplete representation of vaccinated populations, and reliance on assumptions due to evolving variant dynamics. Future models should incorporate real-world effectiveness, productivity impacts, and updated epidemiological trends to enhance decision-making relevance.
— news from Taylor & Francis Online
— News Original —
Nirmatrelvir/ritonavir treatment for COVID-19: an economic value systematic literature review
INTRODUCTION n nCoronavirus disease 2019 (COVID-19) is a significant cause of morbidity and mortality worldwide [1,2], with over 270 million cases across Europe through 13 January 2023 and over 2 million deaths in Europe since the beginning of the pandemic [3]. During the height of the pandemic (2020-2021), COVID-19 placed substantial pressure on healthcare systems, with hospitalizations being an important driver of increased utilization of healthcare resources [4]. Notably, the United Kingdom (UK) government has spent an estimated £358 billion so far on measures to address the impacts of the COVID-19 pandemic [5], and the International Monetary Fund predicted the global cost of the COVID-19 pandemic to reach $13.8 trillion by 2024 [6]. Although the pandemic has ended, healthcare systems and individuals at high risk of contracting COVID-19—including the elderly, those who are immunocompromised, or those with preexisting health conditions that place them at risk of progression to severe disease—are still burdened by the emergence of new, highly transmissible COVID-19 variants (e.g., Omicron) [7-9]. n nWhile there are many potential therapeutic agents for the treatment of COVID-19 [10], a limited number have been authorized by regulatory agencies [11,12]. Due to the rapid emergence of COVID-19, prompt decision-making regarding treatment options was necessary to contain global mortality rates and alleviate healthcare system pressures. Nirmatrelvir/ritonavir (NMV/r)Footnote*, an oral antiviral drug, was granted Emergency Use Authorization from the end of 2021 onwards in many countries and gained full approval by the European Medicines Agency (EMA) in 2022 and the Food and Drug Administration (FDA) in 2023 for the treatment of mild-to-moderate COVID-19 in patients with an increased risk of progression to severe COVID-19 (i.e., COVID-19 that causes death/require hospitalization [including intensive care]) [13-16]. It has been shown to reduce progression to severe COVID-19, by reducing hospitalizations and all-cause mortality [17,18]. Contemporary real-world evidence has demonstrated NMV/r to be effective against the Omicron variant [19], and the drug has been associated with an estimated effectiveness of 79.6% in preventing hospital admission or death when dispensed within 5 days of symptom onset [20]. Therefore, it has the potential to relieve the health, societal, and economic burdens that are still implicated by highly transmittable COVID-19 variants [21]. n nDue to the constantly evolving evidence base of COVID-19, there is a need to regularly update information, including treatment guidelines, to ensure public health guidelines are in alignment with the most recent data. Several country-specific economic evaluations of NMV/r have recently been conducted, which play a key role in decision-making for patient access to treatment. However, a comprehensive assessment of these evaluations that identifies trends and evidence gaps from an economic perspective is lacking. A need exists for a global systematic evaluation to assess key parameters, structures, and outcomes of existing models/analyses of the economic value of treatment with NMV/r. Thus, the objective of this study was to investigate evaluation methodologies and outcomes of studies assessing the economic impact of NMV/r for mild-to-moderate COVID-19 in adults who are at high risk for progression to severe COVID-19. n nMETHODS n nSearch Strategy n nWe conducted an economic systematic literature review (SLR) following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines to identify relevant articles published in databases between January 2022 and September 2024. Electronic databases searched included Embase (using the Elsevier Platform), MEDLINE and MEDLINE In-Process (using the PubMed platform), EconLit, and the Cochrane Library. Search terms included in the respective searches are detailed in Supplementary Tables S1-S4. Websites of 3 conferences not indexed in Embase were searched to identify conference abstracts (The Professional Society for Health Economics and Outcomes Research [ISPOR and ISPOR Europe], https://www.ispor.org/; Infectious Diseases Week, https://idweek.org/; European Society of Clinical Microbiology and Infectious Disease, https://www.escmid.org/), and 5 websites of international health technology assessment (HTA)/methodological agencies were searched to identify model structures and available utility, resource use, and cost data (National Institute for Health and Care Excellence [NICE], https://www.nice.org.uk/; Canada’s Drug Agency, https://www.cda-amc.ca/; International Network of Agencies for Health Technology Assessment, https://www.inahta.org/; Cost-Effectiveness Analysis registry, https://cear.tuftsmedicalcenter.org/; and Institute for Clinical and Economic Review, https://icer-review.org). Bibliographies of relevant, robust systematic reviews, economic analyses, and HTAs were also searched for further studies of interest. The inclusion and exclusion criteria were based on the PICOS framework (population, intervention, comparator, outcomes, and study type) outlined in Table 1. n nThis study was registered in the International Prospective Register of Systematic Reviews (PROSPERO; CRD42024594675). n nStudy Selection and Data Extraction n nTitles and abstracts of identified publications from the electronic databases, internet searches, and hand searches were first screened using the Nested Knowledge Inc platform (level 1 screening). A second round of screening (level 2 screening) of full-text publications was conducted on publications identified as relevant during level 1 screening. All screening was performed by 2 independent researchers to ensure alignment with the inclusion and exclusion criteria, and any disagreements were resolved by consensus or with a third researcher. For each economic evaluation publication, data were extracted by 1 independent researcher from full-text publications, where available, or from the source used (e.g., abstract or poster). Quality-control procedures included verification of all extracted data with original sources by a researcher who did not perform the primary data extraction. A quality assessment was conducted for each included economic evaluation using the Drummond checklist [22]; one researcher performed the quality assessment, and a second researcher performed a quality-control check of the completed assessment against source documents. Economic evaluation results were summarized as reported in the corresponding publications on the basis of local economic evaluation thresholds for decision-making. n nRESULTS n nThe literature search yielded a total of 320 titles and abstracts (databases = 290; internet searches = 12; hand searches = 18) to screen after the removal of duplicates (Figure 1). Following title and abstract screening, 43 articles (database searches = 34; internet searches = 8; hand searches = 1) were retrieved for full-text screening. Among the 25 identified economic evaluation publications (22 primary and 3 secondary), which met the predefined inclusion criteria (database searches = 18; internet searches = 6; hand searches = 1), 22 were included in the review (Table 2). Of the 22 included economic evaluations, 8 were from the US [21,23-29], and 2 were from the UK (England and Wales) [30,31]. Other evaluations originated from Brazil [32], Spain [33], Japan [34], Hong Kong [35], China [36], South Korea [37], Greece [38], Sweden [39], Canada [40], Malaysia [41], and the Netherlands [42]; one evaluation covered Ghana, Rwanda, and Zambia [43]. n nMethodology of Economic Evaluations n nPopulation and Setting n nEighteen of the economic evaluations identified focused on adults (≥ 18 years old) [21,23-26,28-30,32,34-38,41-44], and 3 focused on older patients (≥ 60 years) [31,33,39]. One included adolescents and adults (defined as ≥ 12 years old) [27]. Five of the economic evaluations provided breakdowns of cost-effectiveness outcomes for different age groups [34,36,37,41,43]. Thirteen evaluations specified that analyses were done in patients with mild-to-moderate COVID-19 who were at high risk for progression, whereas 5 evaluations did not specify the COVID-19 symptom severity of the patients in the analysis but reported patients who were at high risk for progression. Two of the remaining studies reported mild-to-moderate disease severity but did not report the risk level for disease progression; 2 further studies did not report disease severity or risk level of disease progression of included patients. Five studies included patients requiring mechanical ventilation [24,25,28,30,44], whereas 3 studies excluded this information [31,38,41]. Nine evaluations focused on the Omicron variant of COVID-19 [21,23,24,27,29,36,39,41,45] and were the only studies that specified the COVID-19 variant of focus. The population setting was reported in 14 of the included studies and covered 6 different settings: outpatient, nonhospital, primary care clinic, household, community setting, hospitals, and outpatient setting (Figure 2). n nDetails of patient vaccination status were provided in only 13 studies; 1 included unvaccinated patients [43], 3 included vaccinated patients [23,32,34], and 9 included both vaccinated and unvaccinated patients [21,25,27-30,36,37,41]. Two studies incorporated vaccination into scenario analyses or subgroup analyses [41,43]. Edoka et al. [43] conducted a scenario analysis of the cost-effectiveness of NMV/r in COVID-19 vaccinated patients using estimates of effectiveness from an observational study and adjusted the base-case hospitalization rates downwards using vaccine efficacy and coverage for each study country. Low et al. [41] performed subgroup analyses to examine the impact of vaccination status on the cost-effectiveness of NMV/r, but details on how vaccination effect and cost were incorporated into the subgroup analyses were not reported. Zhang et al. [36] provided net monetary benefits (NMBs) for both vaccinated and unvaccinated subgroups but did not describe how vaccination effects and costs were incorporated into the model. Furthermore, Jo et al. [37] did not explicitly include vaccination status in their economic evaluation as a variable, but the epidemiological model on which their economic model was based did incorporate vaccine uptake, with estimates coming from observed data. The epidemiological model was used by Jo et al. [37] to make projections for the peak number of hospital and intensive care unit (ICU) beds. n nComparator Treatments n nComparator treatments covered by studies were standard of care (SoC) (n = 8) [25,26,28,30-32,37,39], best supportive care (BSC) (n = 1) [38], SoC with no antivirals (n = 2) [35,43], symptomatic treatment (n = 2) [33,41], BSC without antivirals or antibody treatments (n = 1) [34], no COVID-19 treatments (n = 2) [24,44], and without NMV/r treatment (n = 4) [23,27,29,36]. One study had a comparator listed as no treatment strategy [21], and another did not report a comparator [42]. Definitions for comparator treatments were not routinely reported, although Fernandes et al. [32] defined SoC as the “best treatment available according to local guidelines except for antivirals,” whereas National Institute for Health and Care Excellence [31] reported that SoC was any treatment widely accepted by the National Health Service. Carlson et al. [24] defined BSC as no anti-severe acute respiratory syndrome coronavirus 2 treatment. n nTypes and Approaches n nMost economic evaluations were either a cost-utility analysis (CUA) (n = 10) [24,25,30-34,38,39,44] or a cost-effectiveness analysis (CEA) (n = 9) [21,27,29,35-37,41-43]; 2 were budget-impact models (BIMs) [23,28], and 1 was a cost-consequence analysis (CCA) [26] (Table 2). Evaluations differed in approaches to analysis; some used an analysis of a short-term decision tree with a long-term Markov model [24,25,34,39,40] or a decision tree only [26,29,31,33,43]. One CEA was a population-level transmission model that estimated the impact of NMV/r-based interventions on reducing the burden of future COVID-19 pandemics [21]. Further, the perspectives of evaluations differed, with many being conducted from the healthcare system perspective (n = 11) [25-27,29,32-34,37,41,43,44]; few studies were conducted from the healthcare and societal perspective (n = 3) [24,30,31], the payer perspective (n = 4) [23,28,38,42], and the societal perspective (n = 3) [21,36,39]; one study did not report perspective [35]. Time Horizons ranged from 28 days to a lifetime; short time horizons were 28 days and 30 days, and longer time horizons were 1 year, 300 days, and 10 years. A lifetime horizon was used in 5 evaluations [25,30,31,34,39], and 2 evaluations reported a 1-year time horizon and a lifetime time horizon [24,38]. For studies that reported an incremental cost-effectiveness ratio (ICER) with a quality-adjusted life-year (i.e., cost-utility studies) [24,25,30-32], it was found that the country reporting a 1-year time horizon (Brazil) had a substantially higher ICER/willingness-to-pay (WTP) ratio compared with all countries that reported a lifetime time horizon (Figure 3). Of the evaluations that reported a discount rate, most used 3% annually applied to costs and benefits (n = 3) [24,25,39] and 2 used 3.5% annually applied to costs and benefits [30,31]. n nModel Inputs: Efficacy, Epidemiology, and Utility Data n nFor treatment efficacy, the most common source used by economic evaluations was the Evaluation of Protease Inhibition for COVID-19 in High-Risk Patients (EPIC-HR) randomized controlled trial (RCT) [46] (n = 7) [21,25,29,31,38,39,44]. The US cohort study by Lewnard et al. [20], which was a real-world evidence study of the Omicron era from a highly vaccinated US population (n = 3) [24,28,39] was another common source. Epidemiological sources were not reported in several included economic evaluations (n = 9) [21,26,27,29,30,33,35,38,42] and varied widely in those that did report their source [23-25,28,31,32,34,36,37,39,41,43,44]. The years of epidemiological data ranged from 2021 through 2023. Frequently cited sources of utility and/or disutility data were Goswami et al. [47] (n = 3) [24,34,39], Sheinson et al. [48] (n = 3) [24,25,44], and Ara et al. [49] (n = 3) [30,31,39]. n nUncertainty and Scenario Analyses n nA total of 15 economic evaluations conducted scenario analyses to explore uncertainties in model input and the impact of additional sources of economic value (Table S-5). The impact of NMV/r on long COVID [24,34], postacute COVID syndrome [39,43], and post-COVID conditions [23,28,44] was commonly explored. In addition, the impact on vaccinated patients [32,37,43] and unvaccinated patients (Institute for Clinical and Economic Review) [25] were explored. The remaining scenario analyses explored changes in various model parameters (utility values, efficacy rates, early treatment initiation, productivity loss, emergency department [ED] and hospitalization duration, probability of hospitalization, and other population parameters). For those economic evaluations reporting treatment impact on long COVID, the symptom duration of long COVID varied, including 3 months [31,34], 6 months [24], and 1 year [39,43]. Some studies did not report symptom duration for long COVID, even though post–COVID-19 conditions were included in the scenario analyses [23,28,44]. Among these, one study did not include symptom duration for long COVID, but this publication was available only in a poster format, limiting the information available [23]. Scenarios for different age groups, alongside the overall population, were considered in the scenario analyses of 3 studies [37,41,43]. In addition, healthcare resource use (HCRU) was a focus in several of the included scenario analyses. One considered the duration of mechanical ventilation [44], while another considered outpatient healthcare resource use [24]. Four economic evaluations considered ED and/or hospitalization [25,26,30,31], 3 included treatment efficacy [30,31,37], and 2 included a societal perspective [24,25]. n nEighteen economic evaluations conducted sensitivity analyses, which varied in type (Table S-6). Several conducted both deterministic and probabilistic sensitivity analyses (n = 6) [24,34,36,38,43,44], and 4 conducted univariate and probabilistic sensitivity analyses [25,32,33,41]. Three conducted both a univariate sensitivity analysis and a 2-way sensitivity analysis [23,28,29], and 1 conducted a univariate sensitivity analysis only [37]. Only 2 economic evaluations stated that they conducted deterministic sensitivity analyses, with no information provided about whether they included univariate, 2-way, and/or multiway sensitivity analyses [30,31]. The remaining 2 studies conducted sensitivity analyses with no detail provided about the specific type [21,27]. n nEconomic Evaluation Outcomes n nOverall, economic evaluations found that NMV/r has the potential to reduce deaths, HCRU (i.e., hospitalizations), and the associated costs in high-risk individuals with COVID-19. Additional benefits of treatment, such as societal value and potential impact on long COVID, were explored in various scenario analyses. Overall conclusions of economic value varied by region, patient demographics, and clinical assumptions and were assessed on the basis of local economic evaluation thresholds. Analyses also varied by country currency and costs-year (Appendix A), so study findings are reported and interpreted according to these variables. n nNMV/r was generally found to be cost-effective across different country-specific WTP thresholds when compared with comparator healthcare alternatives by the majority of CUAs (n = 7/10) (Table 3) [24,25,30,31,38,39,44]; Mizuno et al. [34] found NMV/r to be cost-effective in patients specifically aged 60, 70, and 80 years, where the ICER for NMV/r decreased with older age. However, in some studies, it was reported as not cost-effective under specific conditions, such as high vaccination coverage [32] or in elderly patients with risk factors for COVID-19 progression [33]. Overall, the 9 included CEAs also found NMV/r to be cost-effective compared with comparator treatments in Hong Kong [35], South Korea [37], US [21,27,29], China [36], and Malaysia [41] (Table 4). Results from the transmission dynamic model–based CEA by Bai et al. [21] showed that cost-effectiveness and cost savings depended on transmission rates and treatment coverage; net monetary benefit was highest when both treatment rates and transmission rates were high, reflecting significant averted healthcare costs and improved outcomes compared with scenarios with lower treatment rates and lower transmission rates. Further, a study in China by Zhang et al. [36] reported that NMV/r was cost-effective compared with no NMV/r treatment only in patients aged ≥ 80 years with mild-to-moderate COVID-19 and at high risk of disease progression. In Malaysia, Low et al. [41] reported that NMV/r was only marginally cost-effective compared with symptomatic treatment in patients presenting to a primary care clinic with a confirmed diagnosis of COVID-19. n nIn the Netherlands, it was concluded that treatment price adjustments would be required to demonstrate the cost-effectiveness of NMV/r in patients at high risk of severe COVID-19, with a higher price adjustment required when treatment uptake and treatment effectiveness are low [42]. The Africa-based study by Edoka et al. [43] found NMV/r not to be cost-effective in Ghana, Rwanda, and Zambia when all adults were considered, although it dominated SoC across all 3 countries when only older patients were considered. In adults with other risk factors, Edoka et al. [43] further reported that NMV/r dominated usual care in Rwanda and Zambia, but in Ghana, NMV/r was more costly but also more effective than usual care. n nOf the 2 BIM studies included in this review, both highlighted substantial cost offsets due to reduced hospitalizations and potential overall cost savings for US health plans, emphasizing the economic viability of NMV/r in managing patients at high risk of severe COVID-19. Campbell et al. [23], in a BIM study that was an update of the BIM study previously conducted by Sandin et al. [28], estimated that 29,999 patients were eligible and sought NMV/r treatment over 1 year, assuming a 1-million-member commercial health insurance plan, resulting in 631 fewer hospitalizations and a budget impact of $1,302,304. This translated to $0.11 per member per month (PMPM) and $43 per patient per year (PPPY). Including post-COVID conditions (PCCs), the treatment was cost saving with a budget impact of −$2,345,609, −$0.20 PMPM, and −$78 PPPY. Similarly, Sandin et al. [28] estimated 29,999 eligible patients for treatment based on a hypothetical 1-million-member commercial health insurance plan, with 647 fewer hospitalizations and a budget impact of $2,733,745, or $0.23 PMPM and $91 PPPY. When considering PCCs, the treatment also resulted in cost savings with a budget impact of −$1,510,780, −$0.13 PMPM, and −$50 PPPY. Further, a CCA scenario analysis conducted by Mills et al. [26] reported that NMV/r produced cost savings of $625 per patient when compared with SoC and reduced total hospital days by 0.74 days per patient. n nOverall, the most frequently reported parameters impacting the economic models in the sensitivity analyses were treatment cost [28,32,37,43], effectiveness of treatment [23-25,36], and hospitalization rates [24,31]. The scenario analyses conducted in the included economic evaluations suggested that several factors impact the cost-effectiveness estimates. Consideration of the effects of long COVID was also found to improve cost-effectiveness in 2 studies, reporting a reduced ICER when accounting for long COVID [24,39]. Additionally, the scenario analyses conducted by Edoka et al. [43] indicated that initiating treatment early and considering the postacute impacts of COVID-19 both improved the cost-effectiveness of NMV/r. More favorable ICERs were reported when societal costs were included [24]; 2 studies showed that reducing the treatment price improved the cost-effectiveness [36,44]. n nBai et al. [21] conducted a scenario analysis where they separated direct therapeutic benefits of antiviral treatment from transmission-reducing effects. They reported that higher treatment rates yielded substantial NMB—especially in high transmission scenarios. For example, treating 100% of cases in a low transmission rate (reproduction number = 1.2) resulted in an average NMB of $232.26 billion (95% confidence interval [CI], $107.17-$332.85 billion), while at a high transmission rate with 100% of symptomatic cases, the NMB rose to $1,214.09 billion (95% CI, $885.29-$1,605.70 billion). The economic benefits reflected both therapeutic effects (e.g., reduced hospitalizations and death) as well as benefits from diminished transmission. n nQuality Assessments n nUsing the Drummond checklist [22], 59% (n = 13/22) [24,25,28-32,34,36,37,41,43,44] of the included economic evaluations were classified as high quality, while 41% (n = 9/22) [21,23,26,27,33,35,38,39,42] were considered of moderate quality; no studies were classed as low quality. All the included economic evaluations clearly stated their research questions, and most highlighted the economic importance of the research question (n = 15) [23,24,26,28-30,32-38,41,43]. The viewpoints of the analyses were generally well defined, and the alternatives being compared were usually described; some studies lacked detailed descriptions (n = 9) [23,26,27,33,35,38,39,41,42]. n nThe form of economic evaluation (e.g., CEA or CUA) was usually stated and justified in relation to the research questions in most of the economic evaluations (n = 19) [23-26,28-38,41-44]. Sources of effectiveness estimates were often provided, with some studies using systematic reviews, meta-analyses, or trial data. Primary outcome measures—including ICERs and quality-adjusted life-years—were clearly stated in most studies (n = 18) [21,23-25,27-31,33-37,39,41,43,44]. However, the methods used to value health states and other benefits were frequently not reported (n = 14) [23,26-29,33-39,41,42], which could impact the transparency and reproducibility of the evaluations. Details of the patients from whom valuations were obtained were generally lacking. While some studies reported quantities of resources separately from their unit costs, the methods for estimating these quantities and costs were not always clear. Currency and price data were typically recorded, but adjustments for inflation or currency conversion were often missing. n nThe analysis and interpretation of results across the included economic evaluations were generally thorough, with most studies clearly stating the time horizon for costs and benefits, often adopting a lifetime perspective. Discount rates were typically reported and justified, aligning with local guidelines. Sensitivity analyses were commonly conducted, with the choice of variables and the ranges over which parameters were varied usually justified. Incremental analyses were frequently reported, with major outcomes presented in both disaggregated and aggregated forms. However, some studies lacked details on statistical tests and CIs for stochastic data, which could affect the robustness of the findings. Conclusions generally followed logically from the data reported and were often accompanied by appropriate caveats, enhancing the credibility of the studies. n nLimitations of Economic Evaluations n nSeveral limitations were identified in the included economic evaluations. Due to the dynamic COVID-19 landscape, study findings are not always generalizable, which can be due to limited context-specific factors and the absence of region-specific data [27,43]. Notably, with many infections becoming future reinfections and with the unpredictable uptake of vaccines and the rise of new COVID-19 variants, model data can become out of date [24]. Indeed, the decision problem has evolved with new variants, vaccination programs, and changes in SoC, affecting the relevance and accuracy of estimates [25,30,31,40]. Additionally, the exclusion of certain individuals (e.g., vaccinated individuals or those previously infected with COVID-19) further limits the generalizability of some studies [44]. Notably, the applicability of the model in Mizuno et al. [34] was highlighted by the authors because they relied on clinical parameters from research conducted outside the region under investigation. n nThere were inconsistencies in the elements reported in the economic evaluations, with some studies not reporting time horizons [35-37,42,43] and discount rates [21,27,33,35-38,42]. Some CUAs did not report costs and health outcomes [24,25,30,31,39] and some CEAs did not report costs [21,42] or ICER threshold(s) [27,42]. Although the association between ICER/WTP ratio and time horizon could have been explored in this study, the association between ICER/WTP ratio and study quality could not be examined due to limitations in the available data from the included studies. Additionally, since most studies did not report or perform a number needed to treat analysis, this type of analysis was not included in the present study; it was also not within the scope of this study. The short duration of some studies prevented the assessment of long-term effects and impact on cost-effectiveness, particularly in reducing the risk of long COVID [35,41]; only 1 study explored the potential impact of treatment on transmission [21], which may undervalue the benefits of treatment overall. Further, economic and logistical factors, such as drug supply, budget constraints, out-of-pocket costs, and healthcare access, were not fully considered in some CEAs [21,29], and not every model included potential productivity benefits or societal costs, which are important for a comprehensive analysis [25]. The comparison of cost-effectiveness between NMV/r and other interventions was limited; many studies did not conduct direct comparisons between interventions, relying instead on proxies and assumptions for key parameters due to the lack of specific data [36,37,43]. Specifically, some economic evaluations relied on several assumptions that introduced uncertainty into the cost-effectiveness estimates [25,30,31]. Finally, 2 of the 22 included primary references were publication abstracts [38,39], so limited information was obtained from these studies. To mitigate some of these limitations, economic evaluations commonly conducted sensitivity analyses and/or scenario analyses. Furthermore, economic evaluations adjusted hospitalization rates to account for any decrease from higher natural immunity or milder variants [43] and based relative mortality risk for NMV/r on meta-analyses of real-world evidence [34]. Also, studies were confirmed to be of at least moderate quality following the Drummond checklist. n nDISCUSSION n nThis economic SLR provided a comprehensive assessment of 22 economic evaluations of NMV/r in high-risk patients with mild-to-moderate COVID-19 and identified key trends and evidence gaps from existing economic models. The included economic models in this study revealed heterogeneity in their methodologies, including study setting, country of origin, COVID-19 comparators, and differences in patient demographics and underlying comorbidities. Overall, these findings suggest the need for a standardized model for NMV/r to improve generalizability and model consistency. n nCUAs and CEAs comparing NMV/r with either SoC or BSC were the most common study types. They mostly used the EPIC-HR RCT as a source of efficacy data and employed a hybrid decision-tree Markov approach. Most evaluations incorporated both immediate outcomes and treatment decisions with long-term treatment effects and disease progression, as well as conducted scenario analyses and/or sensitivity analyses to understand the impact of different model inputs, assumptions, and conditions on the economic evaluation results. This helped address some of the concerns surrounding potential changes to clinical and epidemiological inputs that could be impacted by the dynamic nature of COVID-19 disease and changes to population dynamics due to varying levels of acquired immunity. n nThe majority of the economic evaluations showed that the value of treatment with NMV/r was derived from averted deaths and improved patient outcomes, as well as reduced HCRU and associated costs. Most of the CUAs and CEAs found NMV/r to be cost-effective at local cost-effectiveness thresholds when compared with comparator treatments in high-risk patients with mild-to-moderate COVID-19, underscoring its economic value in various countries and high-risk patient populations. The 2 included BIMs highlighted the substantial cost offsets due to reduced hospitalizations and potential overall cost savings for US health plans [23,28]. Similarly, a CCA reported NMV/r to be cost saving and that it reduced total hospital days for these patients [26]. Notably, there was consistency in the sources from which the published models obtained efficacy data, suggesting that the included models were based on comparable efficacy information; however, epidemiological data sources were not often reported and appeared to vary widely. Overall, studies generally emphasized the economic viability of NMV/r in managing patients at high risk of severe COVID-19. n nThe magnitude of the economic value of NMV/r and cost-effectiveness is sensitive to patient cohort demography, such as age; clinical data inputs, such as treatment effectiveness; the inclusion of long-term impacts of COVID-19; healthcare costs; and drug prices. The sensitivity to variation in patient age is likely linked to the correlation between increased risk of developing a severe COVID-19 infection with increasing age [50], and the analysis suggests that the cost-effectiveness of NMV/r increases as the risk of severe outcomes increases in populations such as the elderly. Currently NMV/r is not approved as a treatment for long COVID, although a recent study suggested potential treatment benefits in preventing long COVID [51]. Economic evaluations that considered this potential treatment effect reported an additional economic value of NMV/r. Sensitivity analyses also suggested that treatment cost, as well as treatment effectiveness, are important parameters to consider alongside factors influencing mortality risk, hospitalization rates, and hospitalization risk. n nThis study identified several key evidence gaps from the limitations of the included economic evaluations. First, considering the dynamic nature of the disease, the population demographics, and vaccination status, economic models need to be continually updated to reflect current and new data (e.g., epidemiological data and the impact of long COVID) and provide relevant guidance for decision-making. Second, Fernandes et al. [32] did not model all factors impacting real-world efficacy, nor did they consider adverse events, indicating a need for comprehensive real-world data. The need for contemporary data and additional evidence to support certain model assumptions was also noted by HTA bodies such as National Institute for Health and Care Excellence [31] and the Institute for Clinical and Economic Review [25]. The cost-effectiveness of NMV