Several COVID-19 treatments have strong evidence behind them, but the right drug depends almost entirely on when in the illness it is used and how sick the patient is. Paxlovid and remdesivir reduce hospitalizations dramatically when given early to high-risk outpatients, while dexamethasone saves lives in hospitalized patients who need oxygen. At the same time, some of the most widely discussed drugs from the pandemic’s early years, including hydroxychloroquine and ivermectin, have failed repeatedly in rigorous trials. The landscape has also shifted as new variants rendered an entire class of treatment, monoclonal antibodies, effectively obsolete.
Paxlovid and How It Became the Front-Line Outpatient Drug
Paxlovid (nirmatrelvir combined with ritonavir) is the most studied oral antiviral for COVID-19 and remains a first-choice outpatient treatment for people at higher risk of severe disease. In the original trial of unvaccinated high-risk adults, those who started Paxlovid within three days of symptoms had roughly an 89% lower risk of hospitalization or death compared with placebo. All 13 deaths in the trial occurred in the placebo group.1PubMed Central. Oral Nirmatrelvir for High-Risk, Nonhospitalized Adults with Covid-19 Real-world data have held up: a large electronic health record study found that Paxlovid-treated patients had roughly a third the odds of hospitalization compared with untreated patients.2PubMed Central. Effect of Nirmatrelvir/Ritonavir (Paxlovid) on Hospitalization among Adults with COVID-19: an EHR-based Target Trial Emulation from N3C
The catch with Paxlovid is timing. It needs to be started within five days of symptom onset to be effective, and sooner is better. The ritonavir component, which is included not to fight the virus but to slow down your body’s breakdown of nirmatrelvir so the active drug stays at useful levels, also brings a heavy load of drug interactions. That ritonavir issue matters more than most people realize, and we will return to it below.
Molnupiravir and How It Compares
Molnupiravir (sold as Lagevrio) is the other oral antiviral available for outpatient COVID. It works by a completely different mechanism: the drug gets incorporated into the virus’s genetic material during replication and introduces errors, essentially scrambling the viral copies so they can no longer function.3Nature Structural & Molecular Biology. Mechanism of molnupiravir-induced SARS-CoV-2 mutagenesis It works, but not as well as Paxlovid. In a head-to-head trial, viral clearance was about 25% slower with molnupiravir than with Paxlovid, and molnupiravir was eventually removed from that trial platform for meeting the threshold for inferiority.4PubMed. Antiviral efficacy of molnupiravir versus ritonavir-boosted nirmatrelvir in patients with early symptomatic COVID-19 (PLATCOV)
There is also a lingering concern unique to molnupiravir. Because it works by introducing mutations into viral RNA, some researchers have observed that patients treated with the drug can shed virus carrying more new mutations than usual. In the PLATCOV trial, three of nine patients with persistent infections after molnupiravir showed increased mutations in follow-up samples, whereas none of the patients on Paxlovid or no treatment did.4PubMed. Antiviral efficacy of molnupiravir versus ritonavir-boosted nirmatrelvir in patients with early symptomatic COVID-19 (PLATCOV) Whether this contributes meaningfully to the emergence of new variants is still being debated, but it is one reason many clinicians prefer Paxlovid when both options are available. Molnupiravir remains useful as a fallback, especially for patients who cannot take Paxlovid because of drug interactions.
Remdesivir as an Outpatient and Hospital Treatment
Remdesivir (Veklury) was the first antiviral authorized for COVID-19 and is given intravenously, which originally limited its use to hospitalized patients. It was later tested in outpatients and performed impressively: a three-day course in high-risk, non-hospitalized patients reduced hospitalization or death by about 87% compared with placebo.5PubMed. Early Remdesivir to Prevent Progression to Severe Covid-19 in Outpatients A cohort study found similar results, with remdesivir-treated outpatients having a dramatically lower rate of hospitalization or death than untreated patients.6PubMed Central. Early Outpatient Treatment With Remdesivir in Patients at High Risk for Severe COVID-19: A Prospective Cohort Study
The practical limitation is that remdesivir requires intravenous infusions on three consecutive days, making it far less convenient than a pill. It tends to be reserved for patients who cannot take Paxlovid or molnupiravir, or occasionally for hospitalized patients early in their course. Its benefits in patients already on ventilators have been more modest, which tracks with a broader pattern in COVID treatment: antivirals are most effective when given before the disease has shifted from a virus-driven phase to an inflammation-driven one.
Why Timing Matters So Much
A modeling study early in the pandemic helped explain something that has held true for nearly every COVID drug tested since: if you give an antiviral after the virus has already peaked in the body, you can shorten how long the patient sheds virus and tamp down the immune response, but you cannot undo the viral damage that has already accumulated. Potent therapy given before the viral peak, during the presymptomatic or very early symptomatic window, has the greatest potential to change outcomes.1PubMed Central. Oral Nirmatrelvir for High-Risk, Nonhospitalized Adults with Covid-19 This is why the five-day treatment window for Paxlovid exists and why every major antiviral trial has shown the best results in patients treated within the first few days of symptoms.
Dexamethasone and the Inflammation Phase
If antivirals are the early-phase drugs, dexamethasone is the quintessential late-phase drug. This cheap, widely available steroid became one of the pandemic’s most important discoveries thanks to the RECOVERY trial. Among hospitalized patients on ventilators, dexamethasone cut deaths by about a third. Among those receiving supplemental oxygen without a ventilator, it reduced deaths by roughly a fifth. But in patients who were hospitalized but did not need any respiratory support, there was no benefit, and a trend toward possible harm.7PubMed. Dexamethasone in Hospitalized Patients with Covid-19
The logic is straightforward: in severe COVID, much of the damage comes from the body’s own inflammatory response rather than the virus itself. Suppressing that response with a steroid helps when inflammation is the main threat. But in someone whose immune system has not yet ramped up excessively, a steroid just dampens the very defense the body needs. This makes dexamethasone one of the clearest examples of a treatment whose benefit is proportional to how sick the patient is.8PubMed Central. Role of Dexamethasone on Oxygen Requirement, Mortality, and Survival Incidence among COVID-19 Patients
Tocilizumab and Baricitinib for Hospitalized Patients
Dexamethasone is not the only immune-modulating drug with evidence behind it. Tocilizumab, which blocks a signaling molecule called IL-6, had a rocky path through clinical trials. An early single-center study of 20 patients in China reported rapid improvements, but a larger placebo-controlled trial found no difference in death at 28 days.9PubMed Central. Tocilizumab in Hospitalized Patients with Severe Covid-19 Pneumonia The RECOVERY trial, which was much larger, ultimately showed that tocilizumab reduced 28-day mortality and helped patients leave the hospital sooner, especially when given alongside steroids.10The Lancet. Tocilizumab in patients admitted to hospital with COVID-19 (RECOVERY) The discrepancy between trials likely comes down to patient selection and whether patients were also receiving corticosteroids. Tocilizumab seems to add benefit on top of steroids in patients with rapidly worsening inflammation, not as a standalone therapy.
Baricitinib, a drug originally developed for rheumatoid arthritis, works by blocking a different set of inflammatory signals. A meta-analysis of randomized trials found it reduced 28-day mortality by about 31% compared with standard care.11PubMed Central. Baricitinib in hospitalised patients with COVID-19: A meta-analysis of randomised controlled trials The large RECOVERY trial confirmed a benefit, though the researchers noted it was somewhat smaller than earlier, smaller trials had suggested.12The Lancet. Baricitinib in patients admitted to hospital with COVID-19 (RECOVERY) In a trial combining baricitinib with remdesivir, patients on the combination recovered a day sooner on average, and those who started on high-flow oxygen or noninvasive ventilation recovered about eight days sooner.13PubMed. Baricitinib plus Remdesivir for Hospitalized Adults with Covid-19 Both tocilizumab and baricitinib are now recommended in guidelines for hospitalized patients with significant oxygen needs, typically layered on top of dexamethasone.
The Rise and Fall of Monoclonal Antibodies
Early in the pandemic, monoclonal antibodies were a major part of outpatient treatment. Products like casirivimab-imdevimab (Regeneron’s cocktail), sotrovimab, and tixagevimab-cilgavimab (Evusheld, used for prevention in immunocompromised people) worked by binding to the virus’s spike protein and blocking it from entering cells. They were effective against the variants circulating at the time, but the virus kept evolving.
By the time Omicron subvariants emerged, virtually every authorized monoclonal antibody had lost its potency. In lab studies, casirivimab and imdevimab showed dramatically reduced ability to neutralize Omicron sub-lineages across the board. Sotrovimab held up somewhat longer but lost efficacy against BA.2, BA.4, and later subvariants. Tixagevimab-cilgavimab likewise failed against XBB and related lineages.14PubMed Central. Impact of SARS-CoV-2 variant mutations on susceptibility to monoclonal antibodies and antiviral drugs A broader review concluded that monoclonal antibodies developed against earlier variants now have “little prospect” of clinical use.15PubMed Central. In Vitro Efficacy of Antivirals and Monoclonal Antibodies against SARS-CoV-2 Omicron Lineages XBB.1.9.1, XBB.1.9.3, XBB.1.5, XBB.1.16, XBB.2.4, BQ.1.1.45, CH.1.1, and CL.1 The virus’s relentless mutation of its spike protein means that any monoclonal antibody is likely to have a limited shelf life unless it targets a more conserved region of the virus. All previously authorized monoclonal antibodies for COVID have been deauthorized or marked as inactive against current strains in the United States.
Hydroxychloroquine and Ivermectin Do Not Work
No discussion of COVID treatments would be complete without addressing the two drugs that generated the most public controversy and the least clinical benefit. Hydroxychloroquine was tested extensively in the RECOVERY trial, and the hydroxychloroquine arm was stopped early for futility. Death within 28 days was slightly higher in the hydroxychloroquine group than in usual care, and patients on hydroxychloroquine were less likely to be discharged alive. Those not on ventilators at baseline who received hydroxychloroquine were actually more likely to end up on a ventilator or die.16PubMed. Effect of Hydroxychloroquine in Hospitalized Patients with Covid-19
Ivermectin has been tested in multiple large, well-designed randomized trials. In a U.S. trial of over 1,500 outpatients with mild-to-moderate COVID, ivermectin at standard dose did not significantly improve time to recovery and showed no effect on hospitalization or death.17JAMA. Effect of Ivermectin vs Placebo on Time to Sustained Recovery in Outpatients With Mild to Moderate COVID-19 A follow-up trial tested a higher dose for a longer course and reached the same conclusion: no meaningful benefit, and less than a 0.1% probability that ivermectin shortened symptoms by even 24 hours.18JAMA. Effect of Higher-Dose Ivermectin for 6 Days vs Placebo on Time to Sustained Recovery in Outpatients With COVID-19 A separate trial in Japan and Thailand confirmed no difference in time to recovery between ivermectin and placebo.19PubMed. Efficacy and safety of ivermectin in patients with mild COVID-19 in Japan and Thailand The collective evidence is clear: neither hydroxychloroquine nor ivermectin helps COVID patients, and hydroxychloroquine may cause harm.
Blood Thinners in Hospitalized Patients
COVID-19 causes abnormal blood clotting in many hospitalized patients, which led to trials of therapeutic-dose blood thinners (heparin) beyond the low preventive doses typically given to hospital patients. The results split sharply by severity. In moderately ill hospitalized patients who needed only low-flow oxygen or none, full-dose heparin reduced the combined risk of death or progression to ventilation. But in critically ill patients who already required ICU-level care or organ support, therapeutic-dose heparin did not help and may have increased bleeding risk without benefit.20JAMA. Heterogeneous Treatment Effects of Therapeutic-Dose Heparin in Patients Hospitalized for COVID-19
A meta-analysis of randomized heparin trials confirmed this pattern: in moderately ill patients, therapeutic heparin reduced the combined outcome of death or blood clots, but in severely ill patients there was no evidence of benefit and a statistically significant interaction between illness severity and treatment effect.21PubMed Central. Randomized trials of therapeutic heparin for COVID-19: A meta-analysis The HEP-COVID trial found a similar split, with therapeutic heparin cutting the primary outcome roughly in half among non-ICU patients but having no effect in those already in the ICU.22JAMA Internal Medicine. Efficacy and Safety of Therapeutic-Dose Heparin vs Standard Prophylactic or Intermediate-Dose Heparins for Thromboprophylaxis in High-risk Hospitalized Patients With COVID-19 This mirrors the pattern seen with dexamethasone: the same drug can help or be futile depending on where a patient falls on the severity spectrum.
Convalescent Plasma and Its Very Narrow Niche
Giving patients antibody-rich plasma from people who have recovered from COVID was one of the first therapies tried. The evidence shows it can work, but only under a specific set of conditions. A meta-analysis of over 22,000 patients found that when convalescent plasma was given early (within three days of hospitalization or early in disease), it significantly reduced mortality, but when given later to patients in severe or critical condition, it had no effect.23iScience. Convalescent Plasma Therapy for COVID-19: Meta-Analysis of 22,591 Patients A systematic review synthesizing results across trial designs concluded that plasma is effective only when given early, with high antibody levels, to patients who have not yet developed their own antibody response, particularly those who are immunocompromised or seronegative.24PubMed Central. Timing, Titre, and Host Determinants of Convalescent Plasma Efficacy in COVID-19: A Systematic Review With Design-Stratified Narrative Synthesis
With the availability of oral antivirals and the loss of monoclonal antibody options, high-titer convalescent plasma has retained a small role for immunocompromised patients who cannot mount their own immune response. For the general population, it has been largely displaced by more convenient and reliable treatments.
Paxlovid’s Drug Interaction Problem
Paxlovid’s ritonavir component powerfully inhibits an enzyme system in the liver that metabolizes a huge number of common medications. This means that starting a five-day course of Paxlovid can abruptly raise the blood levels of drugs a patient is already taking, sometimes to dangerous levels. The French Society of Pharmacology and Therapeutics published recommendations covering 171 individual drugs and drug classes, categorizing them from “Paxlovid possible without changing anything” all the way to “contraindicated.”25PubMed Central. Management of drug-drug interactions with nirmatrelvir/ritonavir in patients treated for Covid-19
In practice, managing these interactions is tricky because the treatment window is short and there is little time for elaborate monitoring. For many medications, the practical approach is to temporarily pause the interacting drug for the five days of Paxlovid treatment, then resume it afterward.26PubMed Central. Recommendations for the Management of Drug-Drug Interactions Between the COVID-19 Antiviral Nirmatrelvir/Ritonavir (Paxlovid) and Comedications A Japanese study found that about 85% of patients who could not take Paxlovid because of drug interactions were switched to molnupiravir instead.27PubMed Central. Evaluation of the drug-drug interactions management system for appropriate use of nirmatrelvir/ritonavir Medications that commonly cause problems include certain heart drugs, blood thinners, cholesterol-lowering statins, anti-seizure medications, and immunosuppressants. If you take multiple prescription medications, the drug-interaction issue is worth raising with your doctor before you get sick, not after.
Treating COVID in Immunocompromised Patients
People with weakened immune systems, including those on cancer chemotherapy, organ transplant recipients, and patients treated with drugs that deplete certain immune cells, face a different version of COVID. They often cannot clear the virus on their own timeline, leading to prolonged or relapsing infections that do not respond to a standard five-day course of a single antiviral. Case reports and small series have increasingly described using combinations of antivirals, such as Paxlovid plus remdesivir, or remdesivir plus molnupiravir, for longer durations. In one series, combination antiviral therapy succeeded in about 92% of treatment courses compared with 50% for single-drug therapy.28PubMed. Antiviral combination therapy in severely immunocompromised patients with persistent or recurrent COVID-19
Some of these patients also benefit from passive antibody supplementation. A case report described a lymphoma patient who had received CD20-depleting therapy (which eliminates the immune cells that produce antibodies) and experienced repeated COVID relapses on antivirals alone. Adding intravenous immunoglobulin with high antibody levels against the virus finally resolved the infection.29PubMed. Successful management of persistent COVID-19 using combination antiviral therapy and intravenous immunoglobulin transfusion in an immunocompromised host This remains a niche approach, and larger studies are still needed, but the pattern is consistent: immunocompromised patients often need more aggressive and prolonged treatment than the general population.30PubMed. Antiviral combination regimens as rescue therapy in immunocompromised hosts with persistent COVID-19
Drugs with Interesting but Incomplete Evidence
A few repurposed drugs have shown enough promise to merit attention without having landed firmly in standard-of-care guidelines. Fluvoxamine, an antidepressant, reduced the rate of emergency visits or hospitalization in the TOGETHER trial, with treated patients about a third less likely to reach those outcomes than placebo patients.31The Lancet Global Health. Effect of early treatment with fluvoxamine on risk of emergency care and hospitalisation among patients with COVID-19 The proposed mechanisms include anti-inflammatory action and effects on platelet activity, both of which could help if the drug is given before the immune response spirals.32Open Forum Infectious Diseases. The STOP COVID 2 Study: Fluvoxamine vs Placebo for Outpatients With Symptomatic COVID-19 However, with effective purpose-built antivirals now widely available, fluvoxamine has not been widely adopted.
Inhaled budesonide, a common asthma steroid, showed promise in two UK trials. The smaller STOIC trial found a large reduction in urgent care visits or hospitalization with early budesonide use, and patients recovered about a day faster.33PubMed Central. Inhaled budesonide in the treatment of early COVID-19 (STOIC) The larger PRINCIPLE trial confirmed faster recovery (about three days sooner on average) but fell just short of showing a statistically convincing reduction in hospitalizations.34The Lancet. Inhaled budesonide for COVID-19 in people at high risk of complications in the community in the UK (PRINCIPLE) Budesonide is inexpensive and widely available, which makes it a practical option in resource-limited settings even if it does not replace antivirals.
Metformin, the widely used diabetes drug, showed no benefit for acute COVID outcomes in the TOGETHER trial.35PubMed Central. Effect of early treatment with metformin on risk of emergency care and hospitalization among patients with COVID-19 Its most intriguing finding came from a different study looking at longer-term outcomes, where participants who received metformin during acute COVID had a lower incidence of long COVID symptoms over 10 months.36The Lancet Infectious Diseases. Outpatient treatment of COVID-19 and incidence of post-COVID-19 condition over 10 months (COVID-OUT) That finding rests on a single trial where nearly half the participants were unvaccinated, so the evidence is too thin to recommend metformin broadly for this purpose, but it is being studied further.
Can Treating Acute COVID Prevent Long COVID?
One of the most pressing questions is whether early antiviral treatment can reduce a patient’s risk of developing long-lasting symptoms. The evidence so far is mixed and somewhat disappointing. A large target-trial emulation using data from the N3C and RECOVER consortia found that Paxlovid treatment during acute COVID had no effect on overall long COVID incidence or new respiratory symptoms, though there was a small effect on cognitive and fatigue symptoms.37PubMed Central. Effect of Paxlovid treatment during acute COVID-19 on Long COVID onset A separate study of hospitalized patients found that Paxlovid reduced the risk of long COVID symptoms at six months after discharge.38PubMed. Paxlovid reduces the risk of Long COVID in patients six months after hospital discharge These differences may reflect the study populations: hospitalized patients represent a more severe baseline, where antivirals might have a larger impact on downstream inflammation and tissue damage.
The honest assessment is that we do not yet have a reliable way to prevent long COVID through acute treatment alone. Antivirals clearly reduce hospitalization and death, and there are hints they may help with certain post-acute symptoms, but no one should assume that taking Paxlovid during an acute infection guarantees protection against prolonged symptoms.
Global Access Barriers
Having effective drugs means little if patients cannot get them. Pfizer struck a licensing deal with the Medicines Patent Pool that allows 95 low- and middle-income countries to access generic versions of Paxlovid. But those generics have been slow to arrive due to bioequivalence testing requirements, and the countries covered by the deal account for only about 10% of global COVID diagnoses. Roughly 39% of diagnoses come from middle-income countries that were excluded from the deal and face full pricing.39PubMed Central. Barriers to Worldwide Access for Paxlovid, a New Treatment for COVID-19 Even in wealthier countries, getting a prescription filled within the narrow treatment window requires having quick access to testing, a provider visit, and a pharmacy with stock on hand, a chain that breaks easily for people without reliable healthcare access.