Cost-effectiveness of Favipiravir in moderately to severely ill COVID-19 patients in the real-world setting of Saudi arabian pandemic referral hospitals

Abstract

Purpose

We aimed to evaluate the cost effectiveness of Favipiravir treatment versus standard of care (SC) in moderately to severely ill COVID-19 patients from the Saudi healthcare payer perspective.

Methods

We used the patient-level simulation method to simulate a cohort of 415 patients with moderate to severe COVID-19 disease who were admitted to two Saudi COVID-19 referral hospitals: 220 patients on Favipiravir and 195 patients on SC. We estimated the incremental cost-effectiveness ratio (ICER) of Favipiravir versus SC in terms of the probability to be discharged alive from hospital and the mean time in days to discharge one patient alive. The model was performed twice: first, using unweighted, and second, using weighted clinical and economic data. Weighting using the inverse weight probability method was performed to achieve balance in baseline characteristics.

Results

In the unweighted model, base case (probabilistic) ICER estimates favored Favipiravir at savings of Saudi Riyal (SAR)1,611,511 (SAR1,998,948) per 1% increase in the probability of being discharged alive. As to mean time to discharging one patient alive, ICERs favored Favipiravir at savings of SAR11,498 (SAR11,125). Similar results were observed in the weighted model with savings using Favipiravir of SAR1,514,893 (SAR2,453,551) per 1% increase in the probability of being discharged alive, and savings of SAR11,989 (SAR11,277) for each day a patient is discharged alive.

Conclusion

From the payer perspective, the addition of Favipiravir in moderately to severely ill COVID-19 patients was cost-savings over SC. Favipiravir was associated with a higher probability of discharging patients alive and lower daily spending on hospitalization than SC.

1. Introduction

The coronavirus disease 2019 (COVID-19) pandemic caused by SARS-Cov-2 continues to pose major demands on health care systems worldwide. Symptoms of the disease include fever, cough, dyspnea (Vaira et al., 2020). Atypical symptoms were also seen as well such as olfactory, gustatory and vasculitis-like skin eruption (Vaira et al., 2020). In the severe cases, fetal clinical course can occur in which patients may present with respiratory distress syndrome (ARDS) and severe sepsis (Huang et al., 2020, Zaim et al., 2020).

Several treatments have been proposed recently to treat COVID-19. These treatments are antiviral therapies such as remdesivir, Paxlovid-ritonavir and molnupiravir, monoclonal antibodies such as tocilizumab, baricitinib, steroids (i.e dexamethasone), and anticoagulants such as enoxaparin and unfractionated heparins (National Institutes of Health., 2022). Despite marked progress in the development and deployment of vaccines, therapeutic options remain largely confined to the repurposing of existing drugs and regimens (Hailat et al., 2021; “PhRMA: Biopharmaceutical R&D Working to Fight COVID-19”, 2020; Zaim et al., 2020). This is also the case in Saudi Arabia and real-world clinical data were reported on the use of the antiviral agent Favipiravir in two major medical centers that served as national referral centers for severe COVID-19 cases in the capital of Riyadh (Alamer et al., 2021). In the prior comparative effectiveness study Favipiravir in in the real-world setting versus standard of care (SC), the adjusted discharge ratio of Favipiravir versus SC was significant at 1.96 (95 %CI = 1.56 – 2.46, P < 0.001), reduction of progression to mechanical ventilation was significant at an adjusted hazard ratio (HR) of 0.10 (95 %CI = 0.04 – 0.29, P < 0.001), yet mortality benefit was not significant (Alamer et al., 2021).

Favipiravir antiviral therapy was introduced into the market in 2016 to treat influenza in Japan (Furuta et al., 2013). The drug was repurposed globally as a treatment of COVID-19 with trials showing clinical benefit (Cai et al., 2020, Özlüşen et al., 2021). The drug entered the market under emergency use authorization in various countries including Saudi Arabia, United Arab Emirates (UAE), China, Indonesia, Malaysia, Thailand, India, and Russia (“Adis Insight: Favipiravir,” 2020).

To date, regulatory bodies such as the United States Food and Drug Administration (FDA) and the European Medicines Agency (EMA) have not approved the drug for COVID-19 treatment (Hassanipour et al., 2021). The Saudi Ministry of Health (MOH) protocol for patients suspected of confirmed with COVID-19 includes the use of Favipiravir in moderately to severe COVID-19 illness (“Ministry of Health: Saudi MoH Protocol for Patients Suspected of Confirmed with COVID-19,” 2021).

2. Methodology

2.1. Patient population and setting

We used the patient data from our prior real-world comparative effectiveness study of 234 patients on Favipiravir and 223 patients on SC (Alamer et al., 2021) with moderate to severe COVID-19 to align our economic evaluation with the Saudi MOH protocol for patients suspected of/confirmed with COVID-19. SC is delivered to patients according to the Saudi MOH protocols and guidelines (“Ministry of Health, 2021). Per data availability, 220 patients on Favipiravir and 195 patients on SC alone were eligible to be considered in this economic evaluation. The study was approved by the Institutional Review Boards at King Fahad Medical City and Prince Mohammed Bin Abdulaziz Hospital (institutional review board [IRB] 20-477E, July 2020, note that these two hospitals belong to the second cluster in Riyadh and have the same IRB), with informed consent waived as this was considered an exempt study. The baseline characteristics of patients are in Table S1 (Supplemental Material).

2.2. Model overview

The economic outcomes were modeled on the individual patients data was data from previous literature (Alamer et al., 2021). That is, analyses were conducted based on estimated the clinical and cost consequences of the aggregated real-world data. (Fig. 1) (“Cost‐effectiveness modelling using patient‐level simulation,” 2021). We estimated the cost-effectiveness of Favipiravir versus SC in moderate to severe patients considering two outcomes: likelihood of discharging alive patients (probability) and mean time to discharge one alive patient (days). The model was performed twice, each considering two scenarios: (1) unweighted and (2) weighted clinical and economic data. Weighting was performed to achieve balance in baseline characteristics; and to adjust for confounding using the inverse weight probability method (Desai and Franklin, 2019). As shown in Fig. 1, in the Favipiravir arm, the costs associated with Favipiravir doses (loading and maintenance) and with utilization of mechanical ventilation (MV) in both intensive care unit (ICU) and general ward (GW) settings were included in the weighted and unweighted model scenarios. For patients in the SC arm who were previously also treated in ICU, the costs associated with prior utilization of MV in the ICU and subsequent use in the GW setting were included. Cost data were from published data (Khan et al., 2020). Our study is from the Saudi healthcare system payer perspective and uses a time-horizon of five months.

Fig. 1.

Patient simulation model: SC standard of care, ICU intensive care unit, GW general ward, MV mechanical ventilation.

2.3. Unweighted versus weighted model scenarios

In the unweighted model, we estimated the clinical and economic outcomes using crude patient characteristics data from the Alamer et al study (Alamer et al., 2021). In the weighted model, we used the inverse propensity score method to produce balanced outcome data by generating weights for each included patient. The following covariates were used for matching: age, sex, body mass index (BMI; kg/m2), ethnicity; diabetes, hypertension, dyslipidemia; documented history of cardiovascular disease (stable angina, unstable angina, percutaneous coronary intervention, coronary artery bypass graft surgery, myocardial infarction, or heart failure), cerebrovascular disease (transient ischemic attack, stroke), and/or respiratory disease (asthma, chronic obstructive lung disease); Acute Physiologic Assessment and Chronic Health Evaluation II (APACHE II) score; World Health Organization (WHO) severity categories for COVID-19; the use of angiotensin-converting enzyme inhibitors (ACEIs) or angiotensin enzyme blocker (ARBs); and kidney function as estimated by the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) metric. From technical perspective, the unweighted model is more convenient to conduct than the weighted model. However, from decision-making perspective, the weighted model seems to be more comprehensive than the unweighted model in terms of variable adjustment. Technical details about the methodology are provided in the supplemental material (Supplemental Material).

2.4. Clinical outcomes

In both the unweighted and weighted model scenarios, clinical outcomes of interest were the probability of patients being discharged alive and the mean time in days to discharge a patient alive (Alamer et al., 2021).

2.5. Cost inputs

Table 1 shows the model inputs used for the cost-effectiveness analysis. The cost of Favipiravir was obtained from the Saudi MOH through local expert opinion who provided the costs inputs as per the National Unified Procurement Company (National Unified Procurement Company (NUPCO), 2021.). The NUPCO is the centralized company responsible for providing fixed prices fixed for all governmental hospitals in Saudi. The other cost inputs were estimated from a previously published article that reported the direct medical costs of COVID-19 patients admitted to Saudi hospitals (Khan et al., 2020); specifically, in-hospital medications, personal protective equipment, oxygen therapy, mechanical ventilation (MV), isolation room, physician and medical staff fees, laboratory, and diagnostic tests. All cost estimates were in 2020 Saudi Riyals (SAR) which equals US$0.27. As per Riyal, the cost of hospital care was estimated based on utilization of MV in the intensive care unit (ICU) and in the general ward (GW) using the equation:

$$\mathrm{C}\mathrm{o}\mathrm{s}tofmechanicalventilation=66.5\%\ast \mathrm{c}\mathrm{o}\mathrm{s}tofMVinICU+33.5\%\ast \mathrm{c}\mathrm{o}\mathrm{s}tofMVinGW$$ (1)

$$\mathrm{C}\mathrm{o}\mathrm{s}tofnon-mechanicalventilation=7.9\%\ast \mathrm{c}\mathrm{o}\mathrm{s}tofnon-MVinICU+92.1\%\ast \mathrm{c}\mathrm{o}\mathrm{s}tofnon-MVinGW$$ (2)

Table 1.

Model inputs.

Clinical outcomes	Estimates (lower to upper Bounds)	Reference
Number of Favipiravir moderate to severe patients	220	Alamer et al
Number of SC moderate to severe patients	195	Alamer et al
Probabilities for MV	MV – ICU = 66.5 % MV – GW = 33.5 %	Khan and Alruthia et al
Probabilities for NoMVa	NoMV – ICU = 7.9 % NoMV – GW = 92.1 %	Khan and Alruthia et al
Cost (SAR)
Favipiravir 200 mg tablet	SAR3.75 (±10 %)	NUPCO
Cost of MV – ICUb	SAR11,215 (SAR9,168 to SAR13,262)	Khan and Alruthia et al
Cost of NoMV – ICUc	SAR7,809 (SAR6,516 to SAR9,102)	Khan and Alruthia et al
Cost of MV – GWb	SAR7,307 (SAR5,180 to SAR9,434)	Khan and Alruthia et al
Cost of NoMV – GWc	SAR5,192 (SAR4,570 to SAR5,814)	Khan and Alruthia et al

SC supportive care, MV–ICU mechanical ventilation in intensive care unit, NoMV–ICU no mechanical ventilation used in intensive care unit, MV–GW mechanical ventilation in general ward, NoMV–GW no mechanical ventilation used in general ward. NUPCO: National Unified Procurement Company.

^aWe assumed patients who were non - MV to receive a loading dose of 1800 mg twice a day and then 800 mg twice daily for 10 days.

^bMean cost of MV per patient/ day was calculated as 0.335 * cost of MV - ICU + 0.665 * cost of MV - GW.

^cMean cost of non - MV per patient/ day was calculated as 0.097 * cost of NoMV - ICU + 0.921 * cost of NoMV - GW.

In the Favipiravir arm, Favipiravir was considered add-on cost to the total hospital care cost. The NUPCO price for one 200 mg tablet of Favipiravir was SAR3.75. As per the Saudi MOH protocol for patients with COVID-19, patients who are not on MV receive a 800 mg loading dose of Favipiravir twice a day, followed by 800 mg twice daily for 10 days (“Ministry of Health: Saudi MoH Protocol for Patients Suspected of Confirmed with COVID-19,” 2021).

2.6. Base base analysis

In the base case analysis, we used the mean estimates for both cost and clinical outcomes to estimate the incremental cost-effectiveness ratio (ICER) as follows:

$$\mathit{ICER}=Incrementalcost(cos{t}_{\mathit{Favipiravir}}-cos{t}_{\mathit{SC}})/Incrementaleffect(effec{t}_{\mathit{Favipiravir}}-effec{t}_{\mathit{SC}})$$ (3)

In the unweighted model, cost outcomes were estimated by (1) capturing hospital care cost for each patient in each arm; (2) summing the costs of all patients in each arm; and (3) applying the average cost per patient as the numerator in the incremental cost-effectiveness ratio (ICER) equation. For the clinical outcomes in the denominator, we estimated the (1) probability of a patient being discharged alive and (2) the mean time to discharge of one alive patient.

In the weighted model, we used stabilized inverse propensity score weights to estimate the weighted hospital care costs and applied these as the numerator of the ICER equation. Here too, for the clinical outcomes in the denominator of the ICER equation we estimated (1) the probability of a patient being discharged alive and (2) the mean time to discharge one alive patient.

2.7. Probabilistic sensitivity analysis (PSA)

We performed a probabilistic sensitivity analysis (PSA) with 2000 Monte Carlo simulations as a confirmatory analysis to the base case analysis. As shown in Table 1, we incorporated the lower and upper bounds for cost and clinical inputs; and assumed a gamma distribution for cost inputs, a beta distribution for the probability of being discharged alive, and the log-normal distribution for time to discharge alive. The aggregated patients’ clinical data were summarized in Table S1 (Supplemental Material).

2.8. Cost efficiency analysis

We also included a secondary analysis that estimated the relative amount to spend daily on Favipiravir versus SC to discharge one patient alive, using an equation proposed by Igarashi A et al (Igarashi et al., 2020):

$$\mathrm{R}\mathrm{e}lativeamounttospenddailytodisch\mathrm{a}\mathrm{r}\mathrm{g}eonepatientalive=[(Average\mathrm{c}\mathrm{o}\mathrm{s}tofFavipiravirtreatmentperpatient/Averagetimetodisch\mathrm{a}\mathrm{r}\mathrm{g}ealive)/(average\mathrm{c}\mathrm{o}\mathrm{s}tofSCperpatient/averagetimetodisch\mathrm{a}\mathrm{r}\mathrm{g}ealive)]$$ (4)

In this equation, a ratio < 1 indicates cost-savings while a ratio > 1 suggests cost wastage. We performed this analysis on the base case estimates of the weighted and unweighted models.

2.9. Software

Statistical analyses for the weighted model scenario were conducted in R Core Team (2020) software (R Foundation for Statistical Computing, Vienna, Austria). Cost-effectiveness analyses were conducted using a bespoke model developed in Microsoft® Excel® 365 MSO supporting Visual Basic for Applications (VBA) (Microsoft Corporation, Redmond, WA).

3. Results

The associated clinical and economic outcomes for Favipiravir and SC are shown in Table 2. In the unweighted model, Favipiravir treatment journey, as shown in Fig. 1, was associated with a mean cost of SAR63,191 (95 %CI = SAR56,871 to SAR69,510) per patient; a probability to be discharged alive of 0.97 (95 %CI = 0.94 to 0.99); and a mean duration to discharge alive of 11.0 days (95 %CI = 10.0 days to 11.7 days). SC was associated with a mean cost of SAR134,938 (95 %CI = SAR121,445 to SAR148,432) per patient; a probability to be discharged alive of 0.93 (95 %CI = 0.88 to 0.96); and a mean duration to discharge alive of 17.0 days (95 %CI = 15.0 days to 19.2 days).

Table 2.

The cost and clinical outcomes.

Parameter	Unweighted (lower bound to upper bound)	Weighted by IPSW (lower bound to upper bound)*
Favipiravir
Total cost per patient	SAR63,191 (SAR56,871 to SAR69,510)	SAR64,488 (SAR58,039 to SAR70,937)
Proportion of patients discharged alive	0.97 (95 %CI = 0.94 to 0.99)	0.97 (95 %CI = 0.95 to 0.99)
Mean time to discharge a patient alive (days)	11.0 (95 %CI = 10.0 to 11.7)	11.0 (95 %CI = 9.6 to 12.6)
SC
Total cost per patient	SAR134,938 (SAR121,445 to SAR148,432)	SAR132,491 (SAR119,242 to SAR145,740)
Proportion of patients discharged alive	0.93 (95 %CI = 0.88 to 0.96)	0.93 (95 %CI = 0.90 to 0.95)
Mean time to discharge a patient alive	17.0 (95 %CI = 15.0 to 19.2)	17.0 (95 %CI = 13.3 to 20.2)

SC supportive care, IPSW inverse propensity score weighting method.

^aInverse propensity score weighting (stabilized) for each patient was calculated.

In the weighted model, Favipiravir was associated with a mean cost of SAR64,488 (95 %CI = SAR58,039 to SAR70,937) per patient; a probability to be discharged alive of 0.97 (95 %CI = 0.95 to 0.99); and a mean time to discharge alive of 11.0 days (95 %CI = 9.6 days to 12.6 days). SC was associated with a mean cost of SAR132,491 (95 %CI = SAR119,242 to SAR145,740) per patient; a probability to be discharged alive of 0.93 (95 %CI = 0.90 to 0.95); and a mean time to discharge alive patient of 17.0 days (95 %CI = 13.3 days to 20.2 days).

The base case and probabilistic analyses are shown in Table 3. In the analyses based on the unweighted model, Favipiravir was associated with savings of SAR1,611,511 (base case) and SAR1,998,948 (PSA) over SC per 1 % increase in the probability of being discharged alive. Favipiravir was also associated with savings of SAR11,498 (base case) and SAR11,125 (PSA) over SC per one-day reduction in the mean length of stay for a cost-efficiency ratio of 0.74 (base case and PSA) for Favipiravir over SC.

Table 3.

Pharmacoeconomic model outcomes.

Outcome	Unweighted model	Weighted modela
Proportion of patients discharged alive
ICER base case	SAR1,611,511	SAR1,514,893
ICER probabilistic	SAR1,998,948	SAR2,453,551
Interpretation	The amount saved per one percent increase in probability of live discharge attributable to Favipiravir treatment.
Mean time to discharge a patient alive
ICER base vase	SAR11,498	SAR11,989
ICER probabilistic	SAR11,125	SAR11,277
Interpretation	Saving achieved by treating with Favipiravir for one day reduction in length of stay
Cost efficiency ratio (Favipiravir / SC)
Base case	0.74	0.74
Probabilistic	0.74	0.75
Interpretation	Cost efficiency ratio (Favipiravir / SC) to spend daily during hospitalization to discharge 1 patient alive from the hospital

ICER incremental cost-effectiveness ratio, SC supportive care.

^aWeights were derived from inverse propensity scores.

In the analyses based on the weighted model, Favipiravir was associated with savings of SAR1,514,893 (base case) and SAR2,453,551 (PSA) over SC per 1 % increase in the probability of being discharged alive. Favipiravir was also associated with savings of SAR11,989 (base case) and SAR11,277 (PSA) over SC per one-day reduction in the mean length of stay for a cost-efficiency ratio of 0.74 (base case) and 0.75 (PSA) of Favipiravir over SC.

Fig. 2 illustrates the ICER planes generated by the PSAs for the weighted and unweighted models of the proportion of patients discharged alive. Both models yielded results that are in right lower (“south-east”) quadrant of the cost-effectiveness plane, indicating lower cost and better effect.

Fig. 2.

Probabilistic sensitivity analysis: PSA probabilistic sensitivity analysis.

4. Discussion

Our study showed that Favipiravir treatment was associated with lower cost than SC in both the unweighted and the weighted models. On average, and using the point estimates of the two models, the mean cost of Favipiravir was 48 % less than the mean cost of SC. In addition to this monetary benefit, the Favipiravir was also associated with a higher probability for a Favipiravir patient to be discharged alive compared to a SC patient. On average, and considering the point estimates of the two models, patients treated with Favipiravir had a 4 % greater probability of being discharged alive than SC patients. Moreover, the mean time of hospitalization for one alive discharge was on average 6 days less in Favipiravir treated patients than in patients treated by the SC. Thus, for patients with moderate to severe COVID-19, treatment with Favipiravir is associated with better outcomes while also being cost-saving compared to SC certainly a highly favorable result.

While the two models confirmed each other in terms of findings, the magnitude of the value of Favipiravir varied between the unweighted and the weighted models. Favipiravir was preferred more in the weighted than in the unweighted in terms of the proportion of patients discharged alive, as the denominator in the respective ICERs in the base case analyses suggest. PSA results Fig. 2 showed that the PSA ICER in the weighted model is associated with 18.5 % more savings than the PSA ICER in the unweighted model. The weighting approach in the model aimed to make the two groups comparable in terms of the baseline characteristics, which may affect both effectiveness and cost outcomes.

In the case of mean time to live discharge, despite the trade-offs being made at unknown willingness-to-pay thresholds, the weighted model requires higher thresholds than the unweighted model. This is because the later can be assumed to have greater uncertainty due to the treatment arms not being balanced in terms of baseline characteristics. The PSA ICER for the mean time to discharge a patient alive was lower in the unweighted model compared to the weighted model by approximately 1 %.

Both models yielded comparable cost-efficiency ratios. The cost-efficiency metric quantifies the trade-off positions of the two treatment options at 0.74 in both analyses. It indicates that for a given outcome, the daily cost to have a patient treated with Favipiravir is only 74 % of the daily cost of providing a patient with SC. Our study is most timely from the Saudi payer perspective. The COVID-19 pandemic has not ended and endemic disease is likely to persist in the foreseeable future (Murray C. J. L. 2022). We assumed two models to resolve technical uncertainties about potential confounding effects or interaction between independent and dependent variables.

Our study documented a convincingly strong association between Favipiravir and treatment outcomes, in addition to being cost saving over SC. Thus, our study provides real-world data and actual costs in support of formulary and coverage decision-making. We designed our study as a microsimulation in which we used real-world data to estimate the cost-effectiveness of Favipiravir versus SC. While the microsimulation method is more data demanding than conventional cohort methods, it uses fewer assumptions while approximating the structures of pharmacoeconomic models to reality (Brigggs et al., 2006). The results of this study can be considered generalizable in principle to the Saudi population with moderate to severe COVID-19.

Despite the lack of similar studies in the literature, the findings of this study on Favipiravir are consistent with studies that found the use of antivirals in COVID-19 to be cost-saving. A systematic review of 14 studies concluded that several antiviral treatments similar to Favipiravir, were associated with cost-saving outcomes or were cost-effective at study-specific willingness-to-pay thresholds (Dawoud and Soliman, 2020). A systematic review and meta-analysis (Shrestha et al., 2020) of four studies, found significant clinical and radiological improvements following treatment with Favipiravir in comparison to SC (relative risk 1.29, 95 %CI = 1.08–1.54) but no impact on clinical deterioration (OR 0.59, 95 %CI = 0.30 – 1.14) at the endpoint of the study (7–15 days) (Shrestha et al., 2020).

The importance of our study is not limited to the better outcomes at lower cost generated by Favipiravir treatment versus SC. We also demonstrated the allocative efficiency that can be achieved with Favipiravir versus SC. While the mortality benefit of Favipiravir versus SC has not been established yet (Alamer et al., 2021, Hassanipour et al., 2021), the days needed to discharge a patient alive and progression to mechanical ventilation were reduced significantly with Favipiravir versus SC in severe COVID-19 patients (Alamer et al., 2021). This is important from a decision-making point of view, given the pandemic and the high uptake of health care services. The savings achieved by the uptake of Favipiravir are in parallel with the improvement in outcomes and this was reflected in the 0.74 cost-efficiency metrics estimated in our analyses; that is, the relative cost quantity of one day of hospitalization for a Favipiravir versus SC treated patient. The efficiency was viewed in our study as the relative quantity needed to spend daily during hospitalization to discharge a patient alive.

Our study has limitations, while also identifying areas for future research. Limiting the study to proportion of patients discharged alive and mean time to discharge a patient alive, may not fulfill all unmet needs in real-world settings in Saudi Arabia. The timing of treatment administration is important in assessing the patients’ outcomes not captured in this economic model. However, the original study reported the median time to receive the Favipiravir treatment of 2 days with an interquartile range (IQR) of 1 to 3. This suggests that most patient received the Favipiravir treatment as early as possible after their admission. (Alamer et al., 2021). Quality of life data for patients on Favipiravir and SC are lacking for the Saudi population. Extending this study to a cost-utility analysis was not technically possible but would be desirable in the future. Although we applied the cost-efficiency equations proposed by Igarashi et al(Igarashi et al., 2020), the relative lack of agreed-upon willingness-to-pay thresholds or corridors in Saudi Arabia makes interpretation somewhat difficult. Clinical data on adverse events and cost data on management of adverse events are lacking. Studies in China have suggested that antivirals like Umifenovir (Chen et al., 2021), Baloxavir (Lou et al., 2021), or Lopinavir (Cai et al., 2020) may be effective in moderate to severe COVID-19, these regimens are not approved by the Saudi MOH for use in COVID-19 (“Saudi Food & Drug Authority,” 2022.). Future studies may benefit from including these treatments as comparators.

5. Conclusion

Our real-world patient-level simulation analysis showed that the use of Favipiravir over SC was associated with marked cost-savings, a higher probability of patients to be discharged patients alive, and shorter hospitalization duration and lower cost of hospitalization.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary material

The following are the Supplementary data to this article: