Azithromycin for COVID-19: real-time meta analysis of 20 studies
Abstract
Significantly lower risk is seen for mortality and hospitalization. 8 studies from 8 independent teams in 6 countries show significant
benefit.
Meta analysis using the most serious outcome reported shows
16% [5‑27%] lower risk. Results are similar for higher quality studies and worse for Randomized Controlled Trials.
No treatment is 100%
effective. Protocols combine safe and effective options with individual
risk/benefit analysis and monitoring.
Other treatments are more effective.
All data and sources to reproduce this analysis are in the appendix.
Azithromycin for COVID-19 — Highlights
Azithromycin reduces
risk with very high confidence for pooled analysis, high confidence for mortality and hospitalization, and very low confidence for ICU admission and cases.
Real-time updates
and corrections with a consistent protocol for 112 treatments. Outcome specific analysis and combined evidence from all
studies including treatment delay, a primary confounding factor.
Figure 1.
A. Random effects
meta-analysis. This plot shows pooled effects,
see the specific outcome analyses for individual outcomes.
Analysis validating pooled outcomes for
COVID-19 can be found below.
Effect extraction is pre-specified, using the most serious outcome reported.
For details see the appendix.
B. Timeline of results in azithromycin studies.
SARS-CoV-2 infection primarily begins in the upper respiratory
tract and may progress to the lower respiratory tract, other tissues, and the
nervous and cardiovascular systems, which may lead to cytokine storm,
pneumonia, ARDS, neurological injury1-12 and
cognitive deficits4,9, cardiovascular
complications13-15, organ failure, and death.
Minimizing replication as early as possible is recommended.
SARS-CoV-2 infection and replication involves the complex interplay of 50+
host and viral proteins and other factorsA,16-22, providing many
therapeutic targets for which many existing compounds have known activity.
Scientists have predicted that over 8,000 compounds may
reduce COVID-19 risk23, either by
directly minimizing infection or replication, by supporting immune system
function, or by minimizing secondary complications.
We analyze all significant
controlled studies of
azithromycin
for COVID-19.
Search methods, inclusion criteria, effect extraction criteria (more serious
outcomes have priority), all individual study data, PRISMA answers, and
statistical methods are detailed in Appendix 1. We present random
effects meta-analysis results for all studies, studies within each treatment stage, individual outcomes, Randomized Controlled Trials (RCTs), and higher quality studies.
Figure 2 shows stages of possible treatment for
COVID-19. Prophylaxis refers to regularly taking medication before
becoming sick, in order to prevent or minimize infection. Early
Treatment refers to treatment immediately or soon after symptoms appear,
while Late Treatment refers to more delayed treatment.
Figure 2. Treatment stages.
Preclinical research is an important part of the development of
treatments, however results may be very different in clinical trials.
Preclinical results are not used in this paper.
Table 1 summarizes the results for all stages combined, for Randomized Controlled Trials, after exclusions, and for specific outcomes.
Table 2 shows results by treatment stage.
Figure 3 plots individual results by treatment stage.
Figure 4, 5, 6, 7, 8, 9, 10, 11, and 12
show forest plots for random effects meta-analysis of
all studies with pooled effects, mortality results, ventilation, ICU admission, hospitalization, progression, recovery, cases, and viral clearance.
| Improvement | Studies | Patients | Authors | |
|---|---|---|---|---|
| All studies | 16% [5‑27%] ** | 20 | 42,142 | 5,723 |
| After exclusions | 18% [6‑28%] ** | 19 | 41,563 | 5,650 |
| Randomized Controlled TrialsRCTs | 2% [-6‑10%] | 7 | 9,867 | 5,515 |
| Mortality | 18% [3‑31%] * | 15 | 37,912 | 5,665 |
| VentilationVent. | 3% [-48‑36%] | 4 | 8,874 | 5,441 |
| ICU admissionICU | 52% [-35‑83%] | 2 | 1,231 | 36 |
| HospitalizationHosp. | 20% [3‑35%] * | 3 | 1,544 | 55 |
| Recovery | 1% [-54‑36%] | 2 | 1,627 | 37 |
| Cases | 19% [-13‑43%] | 2 | 0 | 15 |
| RCT mortality | 2% [-6‑10%] | 5 | 8,442 | 5,478 |
| RCT hospitalizationRCT hosp. | 20% [3‑35%] * | 3 | 1,544 | 55 |
| Early treatment | Late treatment | Prophylaxis | |
|---|---|---|---|
| All studies | 29% [16‑40%] **** | 15% [-5‑31%] | 14% [3‑23%] * |
| After exclusions | 29% [16‑40%] **** | 17% [-3‑33%] | 14% [3‑23%] * |
| Randomized Controlled TrialsRCTs | -33% [-486‑70%] | 2% [-6‑10%] | |
| Mortality | 40% [2‑60%] * | 14% [-6‑31%] | 34% [-28‑66%] |
| VentilationVent. | 3% [-48‑36%] | ||
| ICU admissionICU | 52% [-35‑83%] | ||
| HospitalizationHosp. | -33% [-486‑70%] | 21% [3‑36%] * | |
| Recovery | -100% [-685‑49%] | 7% [-5‑19%] | |
| Cases | 19% [-13‑43%] | ||
| RCT mortality | 2% [-6‑10%] | ||
| RCT hospitalizationRCT hosp. | -33% [-486‑70%] | 21% [3‑36%] * |
Loading..
Figure 3. Scatter plot showing the most serious outcome in all studies, and for studies within each stage. Diamonds shows the results of random effects meta-analysis.
Loading..
Loading..
Figure 4. Random effects meta-analysis for all studies.
This plot shows pooled effects,
see the specific outcome analyses for individual outcomes.
Analysis validating pooled outcomes for
COVID-19 can be found below.
Effect extraction is pre-specified, using the most serious outcome reported.
For details see the appendix.
Loading..
Loading..
Figure 5. Random effects meta-analysis for mortality results.
Loading..
Figure 6. Random effects meta-analysis for ventilation.
Loading..
Figure 7. Random effects meta-analysis for ICU admission.
Loading..
Figure 8. Random effects meta-analysis for hospitalization.
Loading..
Figure 9. Random effects meta-analysis for progression.
Loading..
Figure 10. Random effects meta-analysis for recovery.
Loading..
Figure 11. Random effects meta-analysis for cases.
Loading..
Figure 12. Random effects meta-analysis for viral clearance.
Figure 13 shows a comparison of results for RCTs and non-RCT studies.
Figure 14, 15, and 16
show forest plots for random effects meta-analysis of
all Randomized Controlled Trials, RCT mortality results, and RCT hospitalization results.
RCT results are included in Table 1 and Table 2.
Loading..
Figure 13. Results for RCTs and non-RCT studies.
Loading..
Figure 14. Random effects meta-analysis for all Randomized Controlled Trials.
This plot shows pooled effects,
see the specific outcome analyses for individual outcomes.
Analysis validating pooled outcomes for
COVID-19 can be found below.
Effect extraction is pre-specified, using the most serious outcome reported.
For details see the appendix.
Loading..
Figure 15. Random effects meta-analysis for RCT mortality results.
Loading..
Figure 16. Random effects meta-analysis for RCT hospitalization results.
RCTs help to make study groups more similar and can provide a higher level of
evidence, however they are subject to many biases29, and
analysis of double-blind RCTs has identified extreme levels of bias30.
For COVID-19, the overhead may delay treatment, dramatically compromising
efficacy; they may encourage monotherapy for simplicity at the cost of
efficacy which may rely on combined or synergistic effects; the participants
that sign up may not reflect real world usage or the population that benefits
most in terms of age, comorbidities, severity of illness, or other factors;
standard of care may be compromised and unable to evolve quickly based on
emerging research for new diseases; errors may be made in randomization and
medication delivery; and investigators may have hidden agendas or vested
interests influencing design, operation, analysis, reporting, and the
potential for fraud. All of these biases have been observed with COVID-19
RCTs. There is no guarantee that a specific RCT provides a higher level of
evidence.
RCTs are expensive and many RCTs are funded
by pharmaceutical companies or interests closely aligned with pharmaceutical
companies. For COVID-19, this creates an incentive to show efficacy for
patented commercial products, and an incentive to show a lack of efficacy for
inexpensive treatments. The bias is expected to be significant, for example
Als-Nielsen et al. analyzed 370 RCTs from Cochrane reviews, showing that
trials funded by for-profit organizations were 5 times more likely to
recommend the experimental drug compared with those funded by nonprofit
organizations. For COVID-19, some major philanthropic organizations are
largely funded by investments with extreme conflicts of interest for and
against specific COVID-19 interventions.
High quality RCTs for novel acute diseases are more challenging, with
increased ethical issues due to the urgency of treatment, increased risk due
to enrollment delays, and more difficult design with a rapidly evolving
evidence base. For COVID-19, the most common site of initial infection is the
upper respiratory tract. Immediate treatment is likely to be most successful
and may prevent or slow progression to other parts of the body. For a
non-prophylaxis RCT, it makes sense to provide treatment in advance and
instruct patients to use it immediately on symptoms, just as some governments
have done by providing medication kits in advance. Unfortunately, no RCTs have
been done in this way. Every treatment RCT to date involves delayed treatment.
Among the 112 treatments we have analyzed,
65% of RCTs involve very late treatment 5+ days after
onset. No non-prophylaxis COVID-19 RCTs match the potential real-world use of
early treatments. They may more accurately represent results for treatments
that require visiting a medical facility, e.g., those requiring intravenous
administration.
RCTs have a bias against finding an effect for interventions
that are widely available — patients that believe they need the
intervention are more likely to decline participation and take the
intervention. RCTs for azithromycin are more likely to enroll low-risk
participants that do not need treatment to recover, making the results less
applicable to clinical practice. This bias is likely to be greater for widely
known treatments, and may be greater when the risk of a serious outcome is
overstated. This bias does not apply to the typical pharmaceutical trial of a
new drug that is otherwise unavailable.
For COVID-19, observational study results do not systematically
differ from RCTs, RR 1.00 [0.92‑1.08]
across 112 treatments32.
Evidence shows that observational studies
can also provide reliable results. Concato et al. found that well-designed
observational studies do not systematically overestimate the magnitude of the
effects of treatment compared to RCTs. Anglemyer et al. analyzed reviews
comparing RCTs to observational studies and found little evidence for
significant differences in effect estimates.
We performed a similar analysis across the 112 treatments
we cover, showing no significant difference in the results of RCTs compared to
observational studies, RR 1.00 [0.92‑1.08]. Similar results are found for all low-cost treatments, RR
1.02 [0.92‑1.12]. High-cost treatments
show a non-significant trend towards RCTs showing greater efficacy,
RR 0.92 [0.82‑1.03].
Details can be found in the
supplementary data.
Lee et al. showed that only 14% of the guidelines of the Infectious
Diseases Society of America were based on RCTs. Evaluation of studies relies
on an understanding of the study and potential biases. Limitations in an RCT
can outweigh the benefits, for example excessive dosages, excessive treatment
delays, or remote survey bias may have a greater effect on results. Ethical
issues may also prevent running RCTs for known effective treatments. For more
on issues with RCTs see36,37.
Currently, 48 of the treatments we analyze show statistically significant efficacy or harm, defined as ≥10% decreased risk or >0% increased risk from ≥3 studies. Of these, 60% have been confirmed in RCTs, with a mean delay of 7.1 months (68% with 8.2 months delay for low-cost treatments). The remaining treatments either have no RCTs, or the point estimate is consistent.
We need to evaluate each trial on its own merits. RCTs for a
given medication and disease may be more reliable, however they may also be
less reliable. For off-patent medications, very high conflict of interest
trials may be more likely to be RCTs, and more likely to be large trials that
dominate meta analyses.
To avoid bias in the selection of studies, we analyze all
non-retracted studies. Here we show the results after excluding
studies with major issues likely to alter results, non-standard studies, and
studies where very minimal detail is currently available. Our bias evaluation
is based on analysis of each study and identifying when there is a significant
chance that limitations will substantially change the outcome of the study. We
believe this can be more valuable than checklist-based approaches such as
Cochrane GRADE, which can be easily influenced by potential bias, may ignore
or underemphasize serious issues not captured in the checklists, and may
overemphasize issues unlikely to alter outcomes in specific cases (for example
certain specifics of randomization with a very large effect size and
well-matched baseline characteristics).
The studies excluded are as below.
Figure 17 shows a forest plot for random
effects meta-analysis of all studies after exclusions.
Kuderer, substantial unadjusted confounding by indication likely.
Loading..
Figure 17. Random effects meta-analysis for all studies after exclusions.
This plot shows pooled effects,
see the specific outcome analyses for individual outcomes.
Analysis validating pooled outcomes for
COVID-19 can be found below.
Effect extraction is pre-specified, using the most serious outcome reported.
For details see the appendix.
Heterogeneity in COVID-19 studies arises from many factors including:
The time between infection or the onset of symptoms and
treatment may critically affect how well a treatment works. For example an
antiviral may be very effective when used early but may not be effective in
late stage disease, and may even be harmful. Oseltamivir, for example, is
generally only considered effective for influenza when used within 0-36 or
0-48 hours39,40. Baloxavir marboxil studies for influenza
also show that treatment delay is critical — Ikematsu et al. report
an 86% reduction in cases for post-exposure prophylaxis, Hayden et al.
show a 33 hour reduction in the time to alleviation of symptoms for treatment
within 24 hours and a reduction of 13 hours for treatment within 24-48 hours,
and Kumar et al. report only 2.5 hours improvement for inpatient
treatment.
| Treatment delay | Result |
| Post-exposure prophylaxis | 86% fewer cases41 |
| <24 hours | -33 hours symptoms42 |
| 24-48 hours | -13 hours symptoms42 |
| Inpatients | -2.5 hours to improvement43 |
Figure 18 shows a mixed-effects meta-regression for efficacy
as a function of treatment delay in COVID-19 studies from 112 treatments, showing
that efficacy declines rapidly with treatment delay. Early treatment is
critical for COVID-19.
Loading..
Figure 18. Early treatment is more effective. Meta-regression showing efficacy as a function of treatment delay in COVID-19 studies from 112 treatments.
Details of the patient population including age and comorbidities may
critically affect how well a treatment works. For example, many COVID-19
studies with relatively young low-comorbidity patients show all patients
recovering quickly with or without treatment. In such cases, there is little
room for an effective treatment to improve results, for example as in
López-Medina et al.
Efficacy may depend critically on the distribution of
SARS-CoV-2 variants encountered by patients. Risk varies significantly across
variants45, for example the Gamma variant shows significantly
different characteristics46-49. Different
mechanisms of action may be more or less effective depending on variants, for
example the degree to which TMPRSS2 contributes to viral entry can differ
across variants50,51.
Effectiveness may depend strongly on the dosage and treatment regimen.
The quality of medications may vary significantly between
manufacturers and production batches, which may significantly affect efficacy
and safety. Williams et al. analyze ivermectin from 11 different sources,
showing highly variable antiparasitic efficacy across different manufacturers.
Xu et al. analyze a treatment from two different manufacturers, showing 9
different impurities, with significantly different concentrations for each
manufacturer.
Across all
studies there is a strong association between different outcomes, for example
improved recovery is strongly associated with lower mortality. However,
efficacy may differ depending on the effect measured, for example a treatment
may be more effective against secondary complications and have minimal effect
on viral clearance.
The
distribution of studies will alter the outcome of a meta analysis. Consider a
simplified example where everything is equal except for the treatment delay,
and effectiveness decreases to zero or below with increasing delay. If there
are many studies using very late treatment, the outcome may be negative, even
though early treatment is very effective.
All meta analyses combine heterogeneous studies, varying in population,
variants, and potentially all factors above, and therefore may obscure
efficacy by including studies where treatment is less effective. Generally, we
expect the estimated effect size from meta analysis to be less than that for
the optimal case.
Looking at all studies is valuable for providing an overview of all research,
important to avoid cherry-picking, and informative when a positive result is
found despite combining less-optimal situations. However, the resulting
estimate does not apply to specific cases such as
early treatment in high-risk populations.
While we present results for all studies, we also present treatment time and
individual outcome analyses, which may be more informative for specific use
cases.
For COVID-19, delay in clinical results translates into
additional death and morbidity, as well as additional economic and societal
damage. Combining the results of studies reporting different outcomes is
required.
There may be no mortality in a trial with low-risk patients,
however a reduction in severity or improved viral clearance may translate
into lower mortality in a high-risk population.
Different studies may report lower severity, improved recovery, and lower mortality,
and the significance may be very high when combining the results.
"The studies reported different outcomes" is not a good reason for
disregarding results.
Pooling the results of studies reporting different outcomes allows us to use
more of the available information. Logically we should, and do, use additional
information when evaluating treatments—for example dose-response and
treatment delay-response relationships provide additional evidence of efficacy
that is considered when reviewing the evidence for a treatment.
We present both specific outcome and pooled analyses.
In order to combine the results of studies reporting different outcomes we use
the most serious outcome reported in each study, based on the thesis that
improvement in the most serious outcome provides comparable measures of
efficacy for a treatment. A critical advantage of this approach is
simplicity and transparency.
There are many other ways to combine evidence for different outcomes, along
with additional evidence such as dose-response relationships, however these
increase complexity.
Trials with high-risk patients may be restricted due to ethics for treatments
that are known or expected to be effective, and they increase difficulty for
recruiting. Using less severe outcomes as a proxy for more serious outcomes
allows faster and safer collection of evidence.
For many COVID-19 treatments, a reduction in mortality logically
follows from a reduction in hospitalization, which follows from a reduction in
symptomatic cases, which follows from a reduction in PCR positivity. We can
directly test this for COVID-19.
Analysis of the the association between different outcomes across studies from
all 112
treatments we cover confirms the validity of pooled outcome analysis for COVID-19.
Figure 19 shows that lower hospitalization is very strongly associated
with lower mortality (p < 0.000000000001).
Similarly, Figure 20 shows that improved recovery is very strongly associated
with lower mortality (p < 0.000000000001).
Considering the extremes, Singh et al. show an association between viral clearance and
hospitalization or death, with p = 0.003 after excluding one large
outlier from a mutagenic treatment, and based on 44 RCTs including 52,384
patients.
Figure 21 shows that improved viral clearance is strongly associated
with fewer serious outcomes. The association is very similar to
Singh et al., with higher confidence due to the larger number of
studies. As with Singh et al., the confidence increases
when excluding the outlier treatment, from p = 0.00000032 to p = 0.000000011.
Loading..
Figure 19. Lower hospitalization is associated with lower mortality, supporting pooled outcome analysis.
Loading..
Figure 20. Improved recovery is associated with lower mortality, supporting pooled outcome analysis.
Loading..
Figure 19. Improved viral clearance is associated with fewer serious outcomes, supporting pooled outcome analysis.
Currently, 48 of the treatments we analyze show statistically significant efficacy or harm, defined as ≥10% decreased risk or >0% increased risk from ≥3 studies. 89% of these have been confirmed with one or more specific outcomes, with a mean delay of 5.0 months. When restricting to RCTs only, 56% of treatments showing statistically significant efficacy/harm with pooled effects have been confirmed with one or more specific outcomes, with a mean delay of 6.4 months.
Figure 22 shows when treatments were found effective during the
pandemic. Pooled outcomes often resulted in earlier detection of efficacy.
Loading..
Loading..
Figure 22. The time when studies showed that
treatments were effective, defined as statistically significant improvement of
≥10% from ≥3 studies.
Pooled results typically show efficacy earlier than specific
outcome results. Results from all studies often shows efficacy much earlier
than when restricting to RCTs.
Results reflect conditions as used in trials to date, these depend on the
population treated, treatment delay, and treatment regimen.
Pooled analysis could hide efficacy, for example a treatment that is
beneficial for late stage patients but has no effect on viral clearance may
show no efficacy if most studies only examine viral clearance. In practice, it
is rare for a non-antiviral treatment to report viral clearance and to not
report clinical outcomes; and in practice other sources of heterogeneity such
as difference in treatment delay is more likely to hide efficacy.
Analysis validates the use of pooled effects and shows significantly faster
detection of efficacy on average.
However, as with all meta analyses, it is important to review the different
studies included. We also present individual outcome analyses, which may be
more informative for specific use cases.
Publishing is often biased
towards positive results, however evidence suggests that there may be a negative bias for
inexpensive treatments for COVID-19. Both negative and positive results are
very important for COVID-19, media in many countries prioritizes negative
results for inexpensive treatments (inverting the typical incentive for
scientists that value media recognition), and there are many reports of
difficulty publishing positive results67-70.
For azithromycin, there is currently not
enough data to evaluate publication bias with high confidence.
One method to evaluate bias is to compare prospective vs.
retrospective studies. Prospective studies are more likely to be published
regardless of the result, while retrospective studies are more likely to
exhibit bias. For example, researchers may perform preliminary analysis with
minimal effort and the results may influence their decision to continue.
Retrospective studies also provide more opportunities for the specifics of
data extraction and adjustments to influence results.
Figure 23 shows a scatter plot of
results for prospective and retrospective studies.
54% of retrospective studies
report a statistically significant positive effect for
one or more outcomes, compared to
14% of prospective studies, consistent with a bias toward publishing positive results.
The median effect size for
retrospective studies is 15% improvement,
compared to 3% for prospective
studies, suggesting a potential bias towards publishing results showing higher efficacy.
Loading..
Figure 23. Prospective vs. retrospective studies.
The diamonds show the results of random effects meta-analysis.
Funnel
plots have traditionally been used for analyzing publication bias. This is
invalid for COVID-19 acute treatment trials — the underlying assumptions
are invalid, which we can demonstrate with a simple example. Consider a set of
hypothetical perfect trials with no bias. Figure 24 plot A
shows a funnel plot for a simulation of 80 perfect trials, with random group
sizes, and each patient's outcome randomly sampled (10% control event
probability, and a 30% effect size for treatment). Analysis shows no asymmetry
(p > 0.05). In plot B, we add a single typical variation in COVID-19 treatment
trials — treatment delay. Consider that efficacy varies from 90% for
treatment within 24 hours, reducing to 10% when treatment is delayed 3 days.
In plot B, each trial's treatment delay is randomly selected. Analysis now
shows highly significant asymmetry, p < 0.0001, with six variants of
Egger's test all showing p < 0.0571-78.
Note that these tests fail even though treatment delay is uniformly
distributed. In reality treatment delay is more complex — each trial has
a different distribution of delays across patients, and the distribution
across trials may be biased (e.g., late treatment trials may be more common).
Similarly, many other variations in trials may produce asymmetry, including
dose, administration, duration of treatment, differences in SOC,
comorbidities, age, variants, and bias in design, implementation, analysis,
and reporting.
Figure 24. Example funnel plot analysis for simulated perfect trials.
Pharmaceutical drug
trials often have conflicts of interest whereby sponsors or trial staff have a
financial interest in the outcome being positive. Azithromycin for COVID-19
lacks this because it is off-patent, has multiple manufacturers, and is very low cost.
In contrast, most COVID-19 azithromycin trials have been run by
physicians on the front lines with the primary goal of finding the best
methods to save human lives and minimize the collateral damage caused by
COVID-19. While pharmaceutical companies are careful to run trials under
optimal conditions (for example, restricting patients to those most likely to
benefit, only including patients that can be treated soon after onset when
necessary, and ensuring accurate dosing), not all azithromycin trials
represent the optimal conditions for efficacy.
Summary statistics from
meta analysis necessarily lose information. As with all meta analyses, studies
are heterogeneous, with differences
in treatment delay, treatment regimen, patient demographics, variants,
conflicts of interest, standard of care, and other factors. We provide analyses for specific
outcomes and by treatment delay, and we aim to identify key characteristics in
the forest plots and summaries. Results should be viewed in the context of
study characteristics.
Some analyses classify treatment based on early or late
administration, as done here, while others distinguish between mild, moderate,
and severe cases. Viral load does not indicate degree of symptoms — for
example patients may have a high viral load while being asymptomatic. With
regard to treatments that have antiviral properties, timing of treatment is
critical — late administration may be less helpful regardless of
severity.
Details of treatment delay per patient is often not available.
For example, a study may treat 90% of patients relatively early, but the
events driving the outcome may come from 10% of patients treated very late.
Our 5 day cutoff for early treatment may be too conservative, 5 days may be too late in many cases.
Comparison across treatments is confounded by differences in
the studies performed, for example dose, variants, and conflicts of interest.
Trials with conflicts of interest may use designs better suited to the
preferred outcome.
In some cases, the most serious outcome has very few events,
resulting in lower confidence results being used in pooled analysis, however
the method is simpler and more transparent. This is less critical as the
number of studies increases. Restriction to outcomes with sufficient power may
be beneficial in pooled analysis and improve accuracy when there are few
studies, however we maintain our pre-specified method to avoid any
retrospective changes.
Studies show that combinations of treatments can be highly
synergistic and may result in many times greater efficacy than individual
treatments alone54-65.
Therefore standard of care may be critical and benefits may diminish or
disappear if standard of care does not include certain treatments.
This real-time analysis is constantly updated based on
submissions. Accuracy benefits from widespread review and submission of
updates and corrections from reviewers. Less popular treatments may receive
fewer reviews.
No treatment or intervention is 100% available and
effective for all current and future variants. Efficacy may vary significantly
with different variants and within different populations. All treatments have
potential side effects. Propensity to experience side effects may be predicted
in advance by qualified physicians. We do not provide medical advice. Before
taking any medication, consult a qualified physician who can compare all
options, provide personalized advice, and provide details of risks and
benefits based on individual medical history and situations.
Enyeji et al. present a review covering azithromycin for COVID-19.
SARS-CoV-2 infection and replication involves a complex
interplay of 50+ host and viral proteins and other
factors16-22, providing many therapeutic
targets.
Over 8,000 compounds have been predicted to reduce COVID-19
risk23, either by directly
minimizing infection or replication, by supporting immune system function, or
by minimizing secondary complications.
Figure 25 shows an overview of the results for azithromycin
in the context of multiple COVID-19 treatments, and Figure 26 shows a plot
of efficacy vs. cost for COVID-19 treatments.
Loading..
Figure 25.
Scatter plot showing results within the context of multiple COVID-19 treatments.
Diamonds shows the results of random effects meta-analysis.
0.5% of 8,000+ proposed treatments show efficacy80.
Loading..
Loading..
Figure 26. Efficacy vs. cost for COVID-19 treatments.
Significantly lower risk is seen for mortality and hospitalization. 8 studies from 8 independent teams in 6 countries show significant
benefit.
Meta analysis using the most serious outcome reported shows
16% [5‑27%] lower risk. Results are similar for higher quality studies and worse for Randomized Controlled Trials.
RCT 7,763 hospitalized COVID-19 patients showing no significant differences with very late (75% on oxygen at baseline) azithromycin treatment. Only 91% of treatment patients received azithromycin and 17% of control patients received azithromycin or other macrolides.
Retrospective 848 ICU patients in Saudi Arabia, showing higher mortality with azithromycin in unadjusted results.
RCT 60 hospitalized COVID-19 patients evaluating the efficacy and safety of adding oral N-acetylcysteine (NAC) at 600mg three times daily to standard antiviral treatment regimens. There was lower mortality for patients that received azithromycin, without statistical significance.
RCT 1,388 outpatients in the UK showing no significant benefit with azithromycin. There was no significant difference in time to first reported recovery or risk of hospitalization or death by 28 days with azithromycin compared to usual care alone. Only 31% of participants had PCR-confirmed SARS-CoV-2 infection.
Late stage RCT of 667 hospitalized patients with up to 14 days of symptoms at enrollment and receiving up to 4 liters per minute supplemental oxygen, not finding a significant effect after 15 days.
Authors note: "the trial cannot definitively rule out either a substantial benefit of the trial drugs or a substantial harm", sample sizes are too small.
The paper uses the terms mild and moderate, however all patients had serious enough disease to be hospitalized, and 14% were actually randomized in the ICU.
The trial had significant protocol deviations and unusually low medication adherence.
Authors note: "our aim was to exclude patients already receiving longer and potentially therapeutic doses of the study treatments" in explanation for why the study protocol was changed to exclude patients with previous use of the medications >24hrs. Analyzing these patients rather than excluding them may have revealed effectiveness with early use as shown in other studies.
The trial initially required enrollment within 48 hours of admission and was changed to remove this requirement, this change is likely to reduce effectiveness because enrollment was moved later, compared to the time the disease became serious enough for hospitalization. Total HCQ dosage 5.6g.
A correction for 17 errors has been published81.
Authors note: "the trial cannot definitively rule out either a substantial benefit of the trial drugs or a substantial harm", sample sizes are too small.
The paper uses the terms mild and moderate, however all patients had serious enough disease to be hospitalized, and 14% were actually randomized in the ICU.
The trial had significant protocol deviations and unusually low medication adherence.
Authors note: "our aim was to exclude patients already receiving longer and potentially therapeutic doses of the study treatments" in explanation for why the study protocol was changed to exclude patients with previous use of the medications >24hrs. Analyzing these patients rather than excluding them may have revealed effectiveness with early use as shown in other studies.
The trial initially required enrollment within 48 hours of admission and was changed to remove this requirement, this change is likely to reduce effectiveness because enrollment was moved later, compared to the time the disease became serious enough for hospitalization. Total HCQ dosage 5.6g.
A correction for 17 errors has been published81.
Retrospective 723 hospitalized COVID-19 patients in Italy showing no significant difference with azithromycin treatment.
Retrospective 31,260 COVID-19 cases and 125,039 matched controls, showing lower risk of COVID-19 with previous azithromycin use.
Small RCT comparing the addition of AZ for very late stage patients on ventilation or oxygen. One notable result is that even within this extremely late stage population, results suggest increased efficacy with the addition of AZ for patients with earlier use of AZ/HCQ, OR 0.71 [0.25-2.03] (Figure S4).
Patients received 8g of HCQ over 10 days, approaching the high levels used in the RECOVERY trial (9.2g over 10 days), showing significantly more adverse events than typical trials. 50% of patients were on mechanical ventilation at baseline.
More than the increase in mortality at day 29 occurred on day 0, and more than 3x the increase occurred by day 2.
Patients received 8g of HCQ over 10 days, approaching the high levels used in the RECOVERY trial (9.2g over 10 days), showing significantly more adverse events than typical trials. 50% of patients were on mechanical ventilation at baseline.
More than the increase in mortality at day 29 occurred on day 0, and more than 3x the increase occurred by day 2.
RCT 298 mild-to-moderate COVID-19 outpatients in the UK showing no significant difference in hospitalization or death with late azithromycin treatment. Treatment was delayed an average of 6 days from onset.
7 vs. 3 hospitalizations occurred by day 1 in the treatment vs. control groups in this open label trial (Figure 2).
7 vs. 3 hospitalizations occurred by day 1 in the treatment vs. control groups in this open label trial (Figure 2).
Retrospective database analysis showing no significant differences with azithromycin use.
Retrospective 928 cancer patients, showing higher mortality with HCQ+AZ. The relative risks of different treatments suggest significant confounding by indication. Authors note that HCQ+AZ might not be the cause of increased mortality, but instead was given to patients with more severe COVID-19.
Retrospective 15,968 COVID-19 hospitalized patients in Spain, showing lower mortality with existing use of several medications including metformin, HCQ, azithromycin, aspirin, vitamin D, vitamin C, and budesonide. Since only hospitalized patients are included, results do not reflect different probabilities of hospitalization across treatments.
Independent analysis of the IHU-Mediterranean data82 with 30,423 COVID-19 patients showing significantly lower risk of ICU admission or death with early treatment of hydroxychloroquine plus azithromycin (HCQ-AZ), and with azithromycin, both compared to no treatment.
Retrospective 672 COVID-19 patients in Zimbabwe, showing lower mortality with azithromycin treatment.
Retrospective study of 917,198 hospitalized COVID-19 cases covered by the Iran Health Insurance Organization over 26 months showing that antithrombotics, corticosteroids, and antivirals reduced mortality while diuretics, antibiotics, and antidiabetics increased it. Confounding makes some results very unreliable. For example, diuretics like furosemide are often used to treat fluid overload, which is more likely in ICU or advanced disease requiring aggressive fluid resuscitation. Hospitalization length has increased risk of significant confounding, for example longer hospitalization increases the chance of receiving a medication, and death may result in shorter hospitalization. Mortality results may be more reliable.
Confounding by indication is likely to be significant for many medications. Authors adjustments have very limited severity information (admission type refers to ward vs. ER department on initial arrival). We can estimate the impact of confounding from typical usage patterns, the prescription frequency, and attenuation or increase of risk for ICU vs. all patients.
Confounding by indication is likely to be significant for many medications. Authors adjustments have very limited severity information (admission type refers to ward vs. ER department on initial arrival). We can estimate the impact of confounding from typical usage patterns, the prescription frequency, and attenuation or increase of risk for ICU vs. all patients.
Low risk patient RCT for HCQ+AZ and HCQ vs. control, not showing any significant differences.
Authors note that the results are not applicable to higher risk patients, that positive PCR may simply reflect detection of inactive (non-infectious) viral remnants, that an alternative dosage regimen may be more effective, and that medication adherence was unknown.
HCQ dosing was 600mg/day for 1 week, therapeutic levels may not be reached for several days. There were no deaths or serious adverse events.
Viral load was already very high at baseline.
Authors note that the results are not applicable to higher risk patients, that positive PCR may simply reflect detection of inactive (non-infectious) viral remnants, that an alternative dosage regimen may be more effective, and that medication adherence was unknown.
HCQ dosing was 600mg/day for 1 week, therapeutic levels may not be reached for several days. There were no deaths or serious adverse events.
Viral load was already very high at baseline.
Retrospective study of 367 hematology patients with COVID-19 in Spain. Among 216 patients with very severe COVID-19, there was significantly lower mortality with azithromycin treatment. Mortality was also lower with HCQ, but without statistical significance.
Randomized controlled trial of 111 hospitalized COVID-19 patients in Iran showing significantly shorter hospital stay, higher oxygen saturation, and lower respiratory rate at discharge with azithromycin plus hydroxychloroquine and lopinavir/ritonavir compared to hydroxychloroquine and lopinavir/ritonavir alone. There were no significant differences in ICU admission, intubation, or mortality, although there was a trend towards lower ICU admission with azithromycin (3.6% vs. 12.7%, p = 0.07). Patients with prior cardiac disease were excluded. The study is limited by the small sample size and open-label design.
Retrospective 7,158 hospitalized COVID-19 patients in the USA analyzing famotidine treatment, showing no significant difference in mortality with associated azithromycin treatment.
Retrospective 3,462 hospitalized COVID-19 patients across 13 states in Nigiera, showing lower mortality with AZ. Authors note the worse results with a combination of CQ/HCQ and AZ, compared to either alone, may be related to the side effects becoming more significant for late stage patients.
We perform ongoing searches of PubMed, medRxiv, Europe PMC,
ClinicalTrials.gov, The Cochrane Library, Google Scholar, Research
Square, ScienceDirect, Oxford University Press, the reference lists of other
studies and meta-analyses, and submissions to the site c19early.org.
Search terms are azithromycin and COVID-19 or SARS-CoV-2. Automated searches are performed twice daily, with all matches reviewed for inclusion.
All studies regarding the use of azithromycin for COVID-19 that report
a comparison with a control group are included in the main analysis.
Sensitivity analysis is performed, excluding studies with major issues,
epidemiological studies, and studies with minimal available information.
This is a living analysis and is updated regularly.
We extracted effect sizes and associated data from all studies.
If studies report multiple kinds of effects then the most serious
outcome is used in pooled analysis, while other outcomes are included in the
outcome specific analyses. For example, if effects for mortality and cases are
both reported, the effect for mortality is used, this may be different to the
effect that a study focused on.
If symptomatic
results are reported at multiple times, we used the latest time, for example
if mortality results are provided at 14 days and 28 days, the results at 28
days have preference. Mortality alone is preferred over combined outcomes.
Outcomes with zero events in both arms are not used, the next most serious
outcome with one or more events is used. For example, in low-risk populations
with no mortality, a reduction in mortality with treatment is not possible,
however a reduction in hospitalization, for example, is still valuable.
Clinical outcomes are considered more important than viral test status. When
basically all patients recover in both treatment and control groups,
preference for viral clearance and recovery is given to results mid-recovery
where available. After most or all patients have recovered there is little or
no room for an effective treatment to do better, however faster recovery is
valuable.
If only individual symptom data is available, the most serious symptom has
priority, for example difficulty breathing or low SpO2 is more
important than cough.
When results provide an odds ratio, we compute the relative risk when
possible, or convert to a relative risk according to83.
Reported confidence intervals and p-values were used when available,
using adjusted values when provided. If multiple types of adjustments are
reported propensity score matching and multivariable regression has preference
over propensity score matching or weighting, which has preference over
multivariable regression. Adjusted results have preference over unadjusted
results for a more serious outcome when the adjustments significantly alter
results. When needed, conversion between reported p-values and
confidence intervals followed Altman, Altman (B), and Fisher's exact
test was used to calculate p-values for event data. If continuity
correction for zero values is required, we use the reciprocal of the opposite
arm with the sum of the correction factors equal to 186.
Results are expressed with RR < 1.0 favoring treatment, and using the risk of
a negative outcome when applicable (for example, the risk of death rather than
the risk of survival). If studies only report relative continuous values such
as relative times, the ratio of the time for the treatment group versus the
time for the control group is used. Calculations are done in Python
(3.13.1) with
scipy (1.14.1), pythonmeta (1.26), numpy (1.26.4), statsmodels (0.14.4), and plotly (5.24.1).
Forest plots are computed using PythonMeta87
with the DerSimonian and Laird random effects model (the fixed effect
assumption is not plausible in this case) and inverse variance weighting.
Results are presented with 95% confidence intervals. Heterogeneity among studies was
assessed using the I2 statistic.
Mixed-effects meta-regression results are computed with R (4.4.0) using the metafor
(4.6-0) and rms (6.8-0) packages, and using the most serious sufficiently powered outcome.
For all statistical tests, a p-value less than 0.05 was considered statistically significant.
Grobid 0.8.0 is used to parse PDF documents.
We have classified studies as early treatment if most patients
are not already at a severe stage at the time of treatment (for example based
on oxygen status or lung involvement), and treatment started within 5 days of
the onset of symptoms. If studies contain a mix of early treatment and late
treatment patients, we consider the treatment time of patients contributing
most to the events (for example, consider a study where most patients are
treated early but late treatment patients are included, and all mortality
events were observed with late treatment patients).
We note that a shorter time may be preferable. Antivirals are typically only
considered effective when used within a shorter timeframe, for example 0-36 or
0-48 hours for oseltamivir, with longer delays not being effective39,40.
We received no funding, this research is done in our spare
time. We have no affiliations with any pharmaceutical companies or political
parties.
A summary of study results is below. Please submit
updates and corrections at https://c19early.org/azmeta.html.
Effect extraction follows pre-specified rules as detailed above
and gives priority to more serious outcomes.
For pooled analyses, the first (most serious) outcome is used, which may
differ from the effect a paper focuses on.
Other outcomes are used in outcome specific analyses.
| Lounnas, 2/29/2024, retrospective, France, peer-reviewed, 6 authors, study period March 2020 - December 2021. | risk of death/ICU, 27.3% lower, OR 0.73, p < 0.001, adjusted per study, propensity score matching, multivariable, RR approximated with OR. |
| Madamombe, 3/21/2023, retrospective, Zimbabwe, peer-reviewed, 9 authors, study period April 2020 - April 2022. | risk of death, 40.0% lower, OR 0.60, p = 0.03, adjusted per study, multivariable, RR approximated with OR. |
| Omrani, 11/20/2020, Double Blind Randomized Controlled Trial, placebo-controlled, Qatar, peer-reviewed, 19 authors, study period 13 April, 2020 - 1 August, 2020, Q-PROTECT trial. | risk of hospitalization, 33.3% higher, RR 1.33, p = 1.00, treatment 4 of 152 (2.6%), control 3 of 152 (2.0%), HCQ+AZ vs. HCQ. |
| progression to pneumonia, 66.7% lower, RR 0.33, p = 0.62, treatment 1 of 152 (0.7%), control 3 of 152 (2.0%), NNT 76, HCQ+AZ vs. HCQ. | |
| risk of symptomatic at day 21, 100% higher, RR 2.00, p = 0.50, treatment 6 of 152 (3.9%), control 3 of 152 (2.0%), HCQ+AZ vs. HCQ. | |
| relative improvement in Ct, 4.2% worse, RR 1.04, p = 0.63, treatment 152, control 152, HCQ+AZ vs. HCQ, day 6. |
Effect extraction follows pre-specified rules as detailed above
and gives priority to more serious outcomes.
For pooled analyses, the first (most serious) outcome is used, which may
differ from the effect a paper focuses on.
Other outcomes are used in outcome specific analyses.
| Abaleke, 2/28/2021, Randomized Controlled Trial, United Kingdom, peer-reviewed, mean age 65.0, 5370 authors, study period 7 April, 2020 - 27 November, 2020, average treatment delay 8.0 days, trial NCT04381936 (history) (RECOVERY). | risk of death, 3.0% lower, RR 0.97, p = 0.50, treatment 561 of 2,430 (23.1%), control 1,162 of 4,881 (23.8%), NNT 139, day 28, primary outcome. |
| risk of death, 3.0% lower, RR 0.97, p = 0.52, treatment 496 of 2,430 (20.4%), control 1,028 of 4,881 (21.1%), NNT 154. | |
| risk of mechanical ventilation, 8.0% lower, RR 0.92, p = 0.29, treatment 211 of 2,430 (8.7%), control 461 of 4,881 (9.4%), NNT 131. | |
| AlQadheeb, 5/10/2023, retrospective, Saudi Arabia, peer-reviewed, mean age 55.8, 9 authors, study period March 2020 - August 2021. | risk of death, 22.2% higher, RR 1.22, p = 0.08, treatment 467 of 775 (60.3%), control 36 of 73 (49.3%). |
| Atefi, 11/20/2023, retrospective, Iran, peer-reviewed, 10 authors, trial IRCT20200623047897N1. | risk of death, 85.1% lower, RR 0.15, p = 0.31, treatment 0 of 18 (0.0%), control 4 of 42 (9.5%), NNT 10, relative risk is not 0 because of continuity correction due to zero events (with reciprocal of the contrasting arm). |
| Butler, 3/31/2021, Randomized Controlled Trial, United Kingdom, peer-reviewed, 18 authors, study period 22 May, 2020 - 30 November, 2020, trial ISRCTN86534580 (PRINCIPLE). | risk of mechanical ventilation, 49.6% lower, RR 0.50, p = 0.47, treatment 2 of 496 (0.4%), control 5 of 625 (0.8%), NNT 252. |
| risk of ICU admission, 24.2% lower, RR 0.76, p = 1.00, treatment 3 of 495 (0.6%), control 5 of 625 (0.8%), NNT 516. | |
| risk of oxygen therapy, 16.2% lower, RR 0.84, p = 0.69, treatment 10 of 497 (2.0%), control 15 of 625 (2.4%), NNT 258. | |
| risk of hospitalization, 8.5% lower, RR 0.91, p = 0.87, treatment 16 of 500 (3.2%), control 22 of 629 (3.5%), NNT 336, concurrent randomization, day 28, Table S2. | |
| risk of no recovery, 7.4% lower, HR 0.93, p = 0.23, treatment 500, control 823, inverted to make HR<1 favor treatment, first reported recovery. | |
| Cavalcanti, 7/23/2020, Randomized Controlled Trial, Brazil, peer-reviewed, baseline oxygen required 41.8%, 35 authors, study period 29 March, 2020 - 18 May, 2020, trial NCT04322123 (history) (COALITION I). | risk of death, 57.0% lower, HR 0.43, p = 0.17, treatment 5 of 172 (2.9%), control 7 of 159 (4.4%), adjusted per study, HCQ+AZ vs. HCQ. |
| risk of death, 44.5% lower, RR 0.55, p = 0.49, treatment 3 of 172 (1.7%), control 5 of 159 (3.1%), NNT 71, HCQ+AZ vs. HCQ, day 15. | |
| risk of mechanical ventilation, 54.0% higher, HR 1.54, p = 0.28, treatment 19 of 172 (11.0%), control 12 of 159 (7.5%), adjusted per study, HCQ+AZ vs. HCQ. | |
| risk of 7-point scale, 18.0% lower, OR 0.82, p = 0.49, treatment 172, control 159, adjusted per study, HCQ+AZ vs. HCQ, RR approximated with OR. | |
| Donida, 1/31/2024, retrospective, Italy, peer-reviewed, 4 authors, study period 21 February, 2020 - 31 May, 2020. | risk of death, 7.0% lower, HR 0.93, p = 0.79, treatment 180 of 548 (32.8%), control 101 of 175 (57.7%), NNT 4.0, adjusted per study, multivariable, Cox proportional hazards. |
| Furtado, 9/4/2020, Randomized Controlled Trial, Brazil, peer-reviewed, 33 authors, average treatment delay 8.0 days, trial NCT04321278 (history) (COALITION II). | risk of death, 8.0% higher, HR 1.08, p = 0.63, treatment 90 of 214 (42.1%), control 73 of 183 (39.9%), day 29. |
| risk of death, 2.6% higher, RR 1.03, p = 0.91, treatment 66 of 214 (30.8%), control 55 of 183 (30.1%), day 15. | |
| 6 point scale, 30.1% lower, OR 0.70, p = 0.08, treatment 214, control 183, inverted to make OR<1 favor treatment, day 29, RR approximated with OR. | |
| 6 point scale, 26.5% lower, OR 0.74, p = 0.10, treatment 214, control 183, inverted to make OR<1 favor treatment, day 15, primary outcome, RR approximated with OR. | |
| Hinks, 10/31/2021, Randomized Controlled Trial, United Kingdom, peer-reviewed, mean age 45.0, 22 authors, study period 3 June, 2020 - 29 January, 2021, average treatment delay 6.02 days, trial NCT04381962 (history) (ATOMIC2). | risk of death, 1.4% higher, RR 1.01, p = 1.00, treatment 1 of 145 (0.7%), control 1 of 147 (0.7%). |
| risk of death/hospitalization, 1.0% lower, HR 0.99, p = 0.99, treatment 15 of 145 (10.3%), control 17 of 147 (11.6%), NNT 82, Cox proportional hazards, model 3. | |
| risk of death/hospitalization, 8.0% lower, RR 0.92, p = 0.82, treatment 15 of 145 (10.3%), control 17 of 147 (11.6%), NNT 82, adjusted per study, odds ratio converted to relative risk, model 3. | |
| progression to pneumonia, 80.3% lower, RR 0.20, p = 0.24, treatment 0 of 119 (0.0%), control 2 of 114 (1.8%), NNT 57, relative risk is not 0 because of continuity correction due to zero events (with reciprocal of the contrasting arm). | |
| Kuderer, 5/28/2020, retrospective, USA, peer-reviewed, 73 authors, excluded in exclusion analyses: substantial unadjusted confounding by indication likely. | risk of death, 26.8% higher, RR 1.27, p = 0.46, treatment 12 of 93 (12.9%), control 41 of 486 (8.4%), odds ratio converted to relative risk, AZ. |
| risk of death, 152.0% higher, RR 2.52, p < 0.001, treatment 45 of 181 (24.9%), control 41 of 486 (8.4%), odds ratio converted to relative risk, HCQ+AZ. | |
| Mehrizi, 12/18/2023, retrospective, Iran, peer-reviewed, 10 authors, study period 1 February, 2020 - 20 March, 2022. | risk of death, 32.0% lower, OR 0.68, p < 0.001, RR approximated with OR. |
| Sekhavati, 10/31/2020, Randomized Controlled Trial, Iran, peer-reviewed, 18 authors, study period 24 April, 2020 - 8 May, 2020. | risk of death, 66.9% lower, RR 0.33, p = 0.50, treatment 0 of 56 (0.0%), control 1 of 55 (1.8%), NNT 55, relative risk is not 0 because of continuity correction due to zero events (with reciprocal of the contrasting arm). |
| risk of mechanical ventilation, 85.8% lower, RR 0.14, p = 0.12, treatment 0 of 56 (0.0%), control 3 of 55 (5.5%), NNT 18, relative risk is not 0 because of continuity correction due to zero events (with reciprocal of the contrasting arm). | |
| risk of ICU admission, 71.9% lower, RR 0.28, p = 0.09, treatment 2 of 56 (3.6%), control 7 of 55 (12.7%), NNT 11. | |
| hospitalization time, 22.7% lower, relative time 0.77, p = 0.02, treatment 56, control 55. | |
| Yeramaneni, 2/28/2021, retrospective, USA, peer-reviewed, 6 authors, study period 11 February, 2020 - 8 May, 2020. | risk of death, 7.0% lower, OR 0.93, p = 0.84, treatment 4,003, control 3,155, adjusted per study, multivariable, day 30, RR approximated with OR. |
| Yilgwan, 5/11/2023, retrospective, Nigeria, peer-reviewed, 12 authors, study period 25 February, 2020 - 30 August, 2021. | risk of death, 67.0% lower, OR 0.33, p < 0.001, treatment 1,619, control 1,843, RR approximated with OR. |
Effect extraction follows pre-specified rules as detailed above
and gives priority to more serious outcomes.
For pooled analyses, the first (most serious) outcome is used, which may
differ from the effect a paper focuses on.
Other outcomes are used in outcome specific analyses.
| Dugot, 3/15/2023, retrospective, Israel, peer-reviewed, 7 authors, study period 1 March, 2020 - 31 December, 2020. | risk of case, 11.8% lower, OR 0.88, p < 0.001, treatment 1,297 of 31,260 (4.1%) cases, 5,919 of 125,039 (4.7%) controls, NNT 47, adjusted per study, case control OR, multivariable. |
| Huh, 12/19/2020, retrospective, database analysis, South Korea, peer-reviewed, 8 authors. | risk of progression, 53.6% higher, RR 1.54, p = 0.33, treatment 3 of 6 (50.0%), control 875 of 2,799 (31.3%), adjusted per study, odds ratio converted to relative risk, multivariable. |
| risk of case, 42.0% lower, OR 0.58, p = 0.10, treatment 11 of 7,341 (0.1%) cases, 103 of 36,705 (0.3%) controls, NNT 14, adjusted per study, case control OR, multivariable. | |
| Loucera, 8/16/2022, retrospective, Spain, peer-reviewed, 8 authors, study period January 2020 - November 2020. | risk of death, 15.0% lower, HR 0.85, p = 0.005, treatment 2,465, control 13,503, Cox proportional hazards, day 30. |
| Piñana, 8/25/2020, retrospective, Spain, peer-reviewed, median age 64.0, 46 authors, study period 1 March, 2020 - 15 May, 2020. | risk of death, 58.0% lower, OR 0.42, p = 0.02, adjusted per study, multivariable, RR approximated with OR. |
Viral infection and replication involves attachment, entry, uncoating and release, genome replication and transcription, translation and protein processing, assembly and budding, and release. Each step can be disrupted by therapeutics.
Rong et al., Persistence of spike protein at the skull-meninges-brain axis may contribute to the neurological sequelae of COVID-19, Cell Host & Microbe, doi:10.1016/j.chom.2024.11.007.
Yang et al., SARS-CoV-2 infection causes dopaminergic neuron senescence, Cell Stem Cell, doi:10.1016/j.stem.2023.12.012.
Scardua-Silva et al., Microstructural brain abnormalities, fatigue, and cognitive dysfunction after mild COVID-19, Scientific Reports, doi:10.1038/s41598-024-52005-7.
Hampshire et al., Cognition and Memory after Covid-19 in a Large Community Sample, New England Journal of Medicine, doi:10.1056/NEJMoa2311330.
Duloquin et al., Is COVID-19 Infection a Multiorganic Disease? Focus on Extrapulmonary Involvement of SARS-CoV-2, Journal of Clinical Medicine, doi:10.3390/jcm13051397.
Sodagar et al., Pathological Features and Neuroinflammatory Mechanisms of SARS-CoV-2 in the Brain and Potential Therapeutic Approaches, Biomolecules, doi:10.3390/biom12070971.
Sagar et al., COVID-19-associated cerebral microbleeds in the general population, Brain Communications, doi:10.1093/braincomms/fcae127.
Verma et al., Persistent Neurological Deficits in Mouse PASC Reveal Antiviral Drug Limitations, bioRxiv, doi:10.1101/2024.06.02.596989.
Panagea et al., Neurocognitive Impairment in Long COVID: A Systematic Review, Archives of Clinical Neuropsychology, doi:10.1093/arclin/acae042.
Ariza et al., COVID-19: Unveiling the Neuropsychiatric Maze—From Acute to Long-Term Manifestations, Biomedicines, doi:10.3390/biomedicines12061147.
Vashisht et al., Neurological Complications of COVID-19: Unraveling the Pathophysiological Underpinnings and Therapeutic Implications, Viruses, doi:10.3390/v16081183.
Ahmad et al., Neurological Complications and Outcomes in Critically Ill Patients With COVID-19: Results From International Neurological Study Group From the COVID-19 Critical Care Consortium, The Neurohospitalist, doi:10.1177/19418744241292487.
Eberhardt et al., SARS-CoV-2 infection triggers pro-atherogenic inflammatory responses in human coronary vessels, Nature Cardiovascular Research, doi:10.1038/s44161-023-00336-5.
Van Tin et al., Spike Protein of SARS-CoV-2 Activates Cardiac Fibrogenesis through NLRP3 Inflammasomes and NF-κB Signaling, Cells, doi:10.3390/cells13161331.
Borka Balas et al., COVID-19 and Cardiac Implications—Still a Mystery in Clinical Practice, Reviews in Cardiovascular Medicine, doi:10.31083/j.rcm2405125.
Malone et al., Structures and functions of coronavirus replication–transcription complexes and their relevance for SARS-CoV-2 drug design, Nature Reviews Molecular Cell Biology, doi:10.1038/s41580-021-00432-z.
Murigneux et al., Proteomic analysis of SARS-CoV-2 particles unveils a key role of G3BP proteins in viral assembly, Nature Communications, doi:10.1038/s41467-024-44958-0.
Lv et al., Host proviral and antiviral factors for SARS-CoV-2, Virus Genes, doi:10.1007/s11262-021-01869-2.
Lui et al., Nsp1 facilitates SARS-CoV-2 replication through calcineurin-NFAT signaling, Virology, doi:10.1128/mbio.00392-24.
Niarakis et al., Drug-target identification in COVID-19 disease mechanisms using computational systems biology approaches, Frontiers in Immunology, doi:10.3389/fimmu.2023.1282859.
Katiyar et al., SARS-CoV-2 Assembly: Gaining Infectivity and Beyond, Viruses, doi:10.3390/v16111648.
Wu et al., Decoding the genome of SARS-CoV-2: a pathway to drug development through translation inhibition, RNA Biology, doi:10.1080/15476286.2024.2433830.
Haque et al., Exploring potential therapeutic candidates against COVID-19: a molecular docking study, Discover Molecules, doi:10.1007/s44345-024-00005-5.
Agamah et al., Network-based multi-omics-disease-drug associations reveal drug repurposing candidates for COVID-19 disease phases, ScienceOpen, doi:10.58647/DRUGARXIV.PR000010.v1.
Rowland Yeo et al., Impact of Disease on Plasma and Lung Exposure of Chloroquine, Hydroxychloroquine and Azithromycin: Application of PBPK Modeling, Clinical Pharmacology & Therapeutics, doi:10.1002/cpt.1955.
Melis et al., The metabolic footprint of Vero E6 cells highlights the key metabolic routes associated with SARS-CoV-2 infection and response to drug combinations, Scientific Reports, doi:10.1038/s41598-024-57726-3.
Purwati et al., An in vitro study of dual drug combinations of anti-viral agents, antibiotics, and/or hydroxychloroquine against the SARS-CoV-2 virus isolated from hospitalized patients in Surabaya, Indonesia, PLOS One, doi:10.1371/journal.pone.0252302.
Jadad et al., Randomized Controlled Trials: Questions, Answers, and Musings, Second Edition, doi:10.1002/9780470691922.
Gøtzsche, P., Bias in double-blind trials, Doctoral Thesis, University of Copenhagen, www.scientificfreedom.dk/2023/05/16/bias-in-double-blind-trials-doctoral-thesis/.
Als-Nielsen et al., Association of Funding and Conclusions in Randomized Drug Trials, JAMA, doi:10.1001/jama.290.7.921.
Anglemyer et al., Healthcare outcomes assessed with observational study designs compared with those assessed in randomized trials, Cochrane Database of Systematic Reviews 2014, Issue 4, doi:10.1002/14651858.MR000034.pub2.
Lee et al., Analysis of Overall Level of Evidence Behind Infectious Diseases Society of America Practice Guidelines, Arch Intern Med., 2011, 171:1, 18-22, doi:10.1001/archinternmed.2010.482.
Deaton et al., Understanding and misunderstanding randomized controlled trials, Social Science & Medicine, 210, doi:10.1016/j.socscimed.2017.12.005.
Nichol et al., Challenging issues in randomised controlled trials, Injury, 2010, doi: 10.1016/j.injury.2010.03.033, www.injuryjournal.com/article/S0020-1383(10)00233-0/fulltext.
Kuderer et al., Clinical impact of COVID-19 on patients with cancer (CCC19): a cohort study, Lancet, June 20, 2020, doi:10.1016/S0140-6736(20)31187-9.
Treanor et al., Efficacy and Safety of the Oral Neuraminidase Inhibitor Oseltamivir in Treating Acute Influenza: A Randomized Controlled Trial, JAMA, 2000, 283:8, 1016-1024, doi:10.1001/jama.283.8.1016.
McLean et al., Impact of Late Oseltamivir Treatment on Influenza Symptoms in the Outpatient Setting: Results of a Randomized Trial, Open Forum Infect. Dis. September 2015, 2:3, doi:10.1093/ofid/ofv100.
Ikematsu et al., Baloxavir Marboxil for Prophylaxis against Influenza in Household Contacts, New England Journal of Medicine, doi:10.1056/NEJMoa1915341.
Hayden et al., Baloxavir Marboxil for Uncomplicated Influenza in Adults and Adolescents, New England Journal of Medicine, doi:10.1056/NEJMoa1716197.
Kumar et al., Combining baloxavir marboxil with standard-of-care neuraminidase inhibitor in patients hospitalised with severe influenza (FLAGSTONE): a randomised, parallel-group, double-blind, placebo-controlled, superiority trial, The Lancet Infectious Diseases, doi:10.1016/S1473-3099(21)00469-2.
López-Medina et al., Effect of Ivermectin on Time to Resolution of Symptoms Among Adults With Mild COVID-19: A Randomized Clinical Trial, JAMA, doi:10.1001/jama.2021.3071.
Korves et al., SARS-CoV-2 Genetic Variants and Patient Factors Associated with Hospitalization Risk, medRxiv, doi:10.1101/2024.03.08.24303818.
Faria et al., Genomics and epidemiology of the P.1 SARS-CoV-2 lineage in Manaus, Brazil, Science, doi:10.1126/science.abh2644.
Nonaka et al., SARS-CoV-2 variant of concern P.1 (Gamma) infection in young and middle-aged patients admitted to the intensive care units of a single hospital in Salvador, Northeast Brazil, February 2021, International Journal of Infectious Diseases, doi:10.1016/j.ijid.2021.08.003.
Karita et al., Trajectory of viral load in a prospective population-based cohort with incident SARS-CoV-2 G614 infection, medRxiv, doi:10.1101/2021.08.27.21262754.
Zavascki et al., Advanced ventilatory support and mortality in hospitalized patients with COVID-19 caused by Gamma (P.1) variant of concern compared to other lineages: cohort study at a reference center in Brazil, Research Square, doi:10.21203/rs.3.rs-910467/v1.
Willett et al., The hyper-transmissible SARS-CoV-2 Omicron variant exhibits significant antigenic change, vaccine escape and a switch in cell entry mechanism, medRxiv, doi:10.1101/2022.01.03.21268111.
Peacock et al., The SARS-CoV-2 variant, Omicron, shows rapid replication in human primary nasal epithelial cultures and efficiently uses the endosomal route of entry, bioRxiv, doi:10.1101/2021.12.31.474653.
Williams, T., Not All Ivermectin Is Created Equal: Comparing The Quality of 11 Different Ivermectin Sources, Do Your Own Research, doyourownresearch.substack.com/p/not-all-ivermectin-is-created-equal.
Xu et al., A study of impurities in the repurposed COVID-19 drug hydroxychloroquine sulfate by UHPLC-Q/TOF-MS and LC-SPE-NMR, Rapid Communications in Mass Spectrometry, doi:10.1002/rcm.9358.
Jitobaom et al., Favipiravir and Ivermectin Showed in Vitro Synergistic Antiviral Activity against SARS-CoV-2, Research Square, doi:10.21203/rs.3.rs-941811/v1.
Jitobaom (B) et al., Synergistic anti-SARS-CoV-2 activity of repurposed anti-parasitic drug combinations, BMC Pharmacology and Toxicology, doi:10.1186/s40360-022-00580-8.
Jeffreys et al., Remdesivir-ivermectin combination displays synergistic interaction with improved in vitro activity against SARS-CoV-2, International Journal of Antimicrobial Agents, doi:10.1016/j.ijantimicag.2022.106542.
Ostrov et al., Highly Specific Sigma Receptor Ligands Exhibit Anti-Viral Properties in SARS-CoV-2 Infected Cells, Pathogens, doi:10.3390/pathogens10111514.
Alsaidi et al., Griffithsin and Carrageenan Combination Results in Antiviral Synergy against SARS-CoV-1 and 2 in a Pseudoviral Model, Marine Drugs, doi:10.3390/md19080418.
Andreani et al., In vitro testing of combined hydroxychloroquine and azithromycin on SARS-CoV-2 shows synergistic effect, Microbial Pathogenesis, doi:10.1016/j.micpath.2020.104228.
De Forni et al., Synergistic drug combinations designed to fully suppress SARS-CoV-2 in the lung of COVID-19 patients, PLoS ONE, doi:10.1371/journal.pone.0276751.
Wan et al., Synergistic inhibition effects of andrographolide and baicalin on coronavirus mechanisms by downregulation of ACE2 protein level, Scientific Reports, doi:10.1038/s41598-024-54722-5.
Said et al., The effect of Nigella sativa and vitamin D3 supplementation on the clinical outcome in COVID-19 patients: A randomized controlled clinical trial, Frontiers in Pharmacology, doi:10.3389/fphar.2022.1011522.
Fiaschi et al., In Vitro Combinatorial Activity of Direct Acting Antivirals and Monoclonal Antibodies against the Ancestral B.1 and BQ.1.1 SARS-CoV-2 Viral Variants, Viruses, doi:10.3390/v16020168.
Xing et al., Published anti-SARS-CoV-2 in vitro hits share common mechanisms of action that synergize with antivirals, Briefings in Bioinformatics, doi:10.1093/bib/bbab249.
Thairu et al., A Comparison of Ivermectin and Non Ivermectin Based Regimen for COVID-19 in Abuja: Effects on Virus Clearance, Days-to-discharge and Mortality, Journal of Pharmaceutical Research International, doi:10.9734/jpri/2022/v34i44A36328.
Singh et al., The relationship between viral clearance rates and disease progression in early symptomatic COVID-19: a systematic review and meta-regression analysis, Journal of Antimicrobial Chemotherapy, doi:10.1093/jac/dkae045.
Boulware, D., Comments regarding paper rejection, twitter.com/boulware_dr/status/1311331372884205570.
Rothstein, H., Publication Bias in Meta-Analysis: Prevention, Assessment and Adjustments, www.wiley.com/en-ae/Publication+Bias+in+Meta+Analysis:+Prevention,+Assessment+and+Adjustments-p-9780470870143.
Stanley et al., Meta-regression approximations to reduce publication selection bias, Research Synthesis Methods, doi:10.1002/jrsm.1095.
Rücker et al., Arcsine test for publication bias in meta-analyses with binary outcomes, Statistics in Medicine, doi:10.1002/sim.2971.
Peters, J., Comparison of Two Methods to Detect Publication Bias in Meta-analysis, JAMA, doi:10.1001/jama.295.6.676.
Moreno et al., Assessment of regression-based methods to adjust for publication bias through a comprehensive simulation study, BMC Medical Research Methodology, doi:10.1186/1471-2288-9-2.
Macaskill et al., A comparison of methods to detect publication bias in meta-analysis, Statistics in Medicine, doi:10.1002/sim.698.
Egger et al., Bias in meta-analysis detected by a simple, graphical test, BMJ, doi:10.1136/bmj.315.7109.629.
Harbord et al., A modified test for small-study effects in meta-analyses of controlled trials with binary endpoints, Statistics in Medicine, doi:10.1002/sim.2380.
Enyeji et al., Effective Treatment of COVID-19 Infection with Repurposed Drugs: Case Reports, Viral Immunology, doi:10.1089/vim.2024.0034.
Brouqui et al., Outcomes after early treatment with hydroxychloroquine and azithromycin: An analysis of a database of 30,423 COVID-19 patients, New Microbes and New Infections, doi:10.1016/j.nmni.2023.101188.
Zhang et al., What's the relative risk? A method of correcting the odds ratio in cohort studies of common outcomes, JAMA, 80:19, 1690, doi:10.1001/jama.280.19.1690.
Altman (B) et al., How to obtain the confidence interval from a P value, BMJ, doi:10.1136/bmj.d2090.
Sweeting et al., What to add to nothing? Use and avoidance of continuity corrections in meta-analysis of sparse data, Statistics in Medicine, doi:10.1002/sim.1761.
Lounnas et al., An Independent Analysis of a Retrospective Cohort of 30,423 Covid-19 Patients Treated at IHU-Mediterranean in Marseille, France: Part 1, Efficacy of early Treatment with Hydroxychloroquine and Azithromycin, Archives of Microbiology & Immunology, doi:10.26502/ami.936500153.
Madamombe et al., Factors associated with COVID-19 fatality among patients admitted in Mashonaland West Province, Zimbabwe 2020-2022: a secondary data analysis, Pan African Medical Journal, doi:10.11604/pamj.2023.44.142.37858.
Omrani et al., Randomized double-blinded placebo-controlled trial of hydroxychloroquine with or without azithromycin for virologic cure of non-severe Covid-19, eClinicalMedicine, doi:10.1016/j.eclinm.2020.100645.
Abaleke et al., Azithromycin in patients admitted to hospital with COVID-19 (RECOVERY): a randomised, controlled, open-label, platform trial, The Lancet, doi:10.1016/S0140-6736(21)00149-5.
AlQadheeb et al., Impact of common comorbidities on antimicrobial consumption and mortality amongst critically ill COVID-19 patients: A retrospective two center study in Saudi Arabia, Clinical Infection in Practice, doi:10.1016/j.clinpr.2023.100229.
Atefi et al., Evaluation of the efficacy and safety of oral N‐acetylcysteine in patients with COVID‐19 receiving the routine antiviral and hydroxychloroquine protocol: A randomized controlled clinical trial, Immunity, Inflammation and Disease, doi:10.1002/iid3.1083.
Butler et al., Azithromycin for community treatment of suspected COVID-19 in people at increased risk of an adverse clinical course in the UK (PRINCIPLE): a randomised, controlled, open-label, adaptive platform trial, The Lancet, doi:10.1016/S0140-6736(21)00461-X.
Cavalcanti et al., Hydroxychloroquine with or without Azithromycin in Mild-to-Moderate Covid-19, NEJM, doi:10.1056/NEJMoa2019014.
Donida et al., First COVID-19 wave in the province of Bergamo, Italy: epidemiological and clinical characteristics, outcome and management of the first hospitalized patients, BMC Infectious Diseases, doi:10.1186/s12879-024-09034-4.
Furtado et al., Azithromycin in addition to standard of care versus standard of care alone in the treatment of patients admitted to the hospital with severe COVID-19 in Brazil (COALITION II): a randomised clinical trial, The Lancet, doi:10.1016/S0140-6736(20)31862-6.
Hinks et al., Azithromycin versus standard care in patients with mild-to-moderate COVID-19 (ATOMIC2): an open-label, randomised trial, The Lancet Respiratory Medicine, doi:10.1016/S2213-2600(21)00263-0.
Mehrizi et al., Drug prescription patterns and their association with mortality and hospitalization duration in COVID-19 patients: insights from big data, Frontiers in Public Health, doi:10.3389/fpubh.2023.1280434.
Sekhavati et al., Safety and effectiveness of azithromycin in patients with COVID-19: An open-label randomised trial, International Journal of Antimicrobial Agents, doi:10.1016/j.ijantimicag.2020.106143.
Yeramaneni et al., Famotidine Use Is Not Associated With 30-day Mortality: A Coarsened Exact Match Study in 7158 Hospitalized Patients With Coronavirus Disease 2019 From a Large Healthcare System, Gastroenterology, doi:10.1053/j.gastro.2020.10.011.
Yilgwan et al., Clinical profile and Predictors of Outcomes of Hospitalized Patients with Laboratory-Confirmed Severe Acute Respiratory Syndrome Coronavirus 2 in Nigeria: A Retrospective Analysis of 13 High Burden States in Nigeria, Nigerian Medical Journal, 64:2, nigerianmedjournal.org/index.php/nmj/article/view/174.
Dugot et al., The Association between Previous Antibiotic Consumption and SARS-CoV-2 Infection: A Population-Based Case-Control Study, Antibiotics, doi:10.3390/antibiotics12030587.
Huh et al., Association of prescribed medications with the risk of COVID-19 infection and severity among adults in South Korea, International Journal of Infectious Diseases, doi:10.1016/j.ijid.2020.12.041.
Please send us corrections, updates, or comments.
c19early involves the extraction of 100,000+ datapoints from
thousands of papers. Community updates
help ensure high accuracy.
Treatments and other interventions are complementary.
All practical, effective, and safe
means should be used based on risk/benefit analysis.
No treatment or intervention is 100% available and effective for all current
and future variants.
We do not provide medical advice. Before taking any medication,
consult a qualified physician who can provide personalized advice and details
of risks and benefits based on your medical history and situation. FLCCC and WCH
provide treatment protocols.