Drug treatments for COVID-19

EVIDENCE AT A GLANCE

As of December 3, 2020, we analysed

43 trials 

41 669 patients

339 drug comparisons

11 outcomes

We identified an additional 45 trials, that will be included in the next update.

DETAILED NARRATIVE SUMMARIES

EFFECT ESTIMATES

See summary of estimates of effect for each treatment and outcome.

STUDY-LEVEL DATA

See data and risk of bias assessments for each study.

NETWORK ESTIMATES & CERTAINTY

LATEST MANUSCRIPTS

Read publications related to our drug treatment LNMA.

WHO, BMJ & MAGIC RECOMMENDATIONS

 

PAGE MENU

DETAILED NARRATIVE SUMMARIES

Efficacy & safety of all drugs

Safety of remdesivir, hydroxychloroquine and lopinavir-ritonavir

Mortality

Last updated: December 14, 2020

74 trials randomising 51, 038 patients

Corticosteroids likely reduce mortality. Recombinant Human Granulocyte Colony-Stimulating Factor and ivermectin may reduce mortality. There is no convincing evidence yet that any of the other treatments have a mortality benefit when compared with standard care or each other.

What to GRADE certainty ratings mean?​

  • Very low: true effect is probably markedly different from the estimated effect.

  • Low: true effect might be markedly different from the estimated effect.

  • Moderate: we believe true effect is probably close to the estimated effect. 

  • High: we have a lot of confidence that the true effect is similar to the estimated effect.

130 deaths per 1,000

in standard care group

Drugs that are among the most effective and have high or moderate certainty evidence probably reduce mortality.

Drugs that are among the most effective and have low or very low certainty evidence might reduce mortality.

Drugs that are not different than standard care and have high or moderate certainty evidence probably do not reduce on mortality.

Drugs that are not different than standard care and have low or very low certainty evidence may not reduce mortality.

Mechanical ventilation

Last updated: October 21, 2020

40 trials randomising 33, 727 patients

Corticosteroids likely reduce the need for mechanical ventilation. Recombinant Human Granulocyte Colony-Stimulating Factor may reduce the need for mechanical ventilation.

Time to symptom resolution

Last updated: October 21, 2020

32 trials randomising 4, 424

Remdesivir probably does not reduce the time to symptoms resolution. There is no convincing evidence that any of the other treatments reduce time to symptom resolution when compared with standard of care or each other.

Adverse events leading to discontinuation

Last updated: October 21, 2020

32 trials randomising 4, 698 patients

Remdesivir probably does not importantly increase the risk of adverse events compared with standard care. Hydroxychloroquine and tocilizumab may not importantly increase the risk of adverse events leading to discontinuation.

Remdesivir

Last updated: October 27, 2020

2 trials randomising 1, 281 patients:

  • 2 trials (1, 281 patients) reporting on acute kidney injury

  • 1 trial (1, 048) reporting on cognitive dysfunction

Remdesivir may have little or no effect on acute kidney injury or cognitive dysfunction/delirium when compared to placebo.

Hydroxychoroquine

Last updated: October 27, 2020

10 trials randomising 3, 663 patients:

  • 7 trials (3, 287 patients) reporting on cardiac toxicity

  • 6 trials (979 patients) reporting on diarrhoea

  • 7 trials (1, 429 patients) reporting on nausea and/or vomiting

  • 1 trial (234 patients) reporting on cognitive function/delirium

  • 2 trials (180 patients) reporting on fatigue

Hydroxychloroquine may increase the risk of cardiac toxicity and cognitive dysfunction/delirium when compared to standard care or placebo. Hydroxychloroquine probably increases the risk of diarrhoea, and nausea and/or vomiting when compared to standard care or placebo. The effect of hydroxychloroquine on fatigue is uncertain when compared to standard care or placebo.

Hydroxychoroquine with azithromycin

Last updated: October 27, 2020

1 trial randomising 667 patients

The effect of hydroxychloroquine with azithromycin on cardiac toxicity and nausea and/or vomiting is uncertain when compared to standard care or placebo.

Lopinavir/ritonavir

Last updated: October 27, 2020

4 trials randomising 370 patients:

  • 4 trials (370 patients) reporting on diarrhoea

  • 4 trials (370 patients) reporting on nausea and/or vomiting

  • 2 trials (259) patients reporting on acute kidney injury

  • 2 trials (254 patients) reporting on fatigue

Lopinavir/ritonavir may increase the risk of diarrhoea and nausea and/or vomiting when compared to standard care or

placebo.  The effect of lopinavir/ritonavir on acute kidney injury and fatigue is uncertain when compared to standard

care or placebo.

 
 
 
 
 
 
 
 
 

EFFECT ESTIMATES

Efficacy & safety of all drugs

Last updated: October 21, 2020

Legend
Most beneficial
Intermediate benefit
High/moderate certainty
Low/very low
certainty
Not different from standard care
Harmful

Safety of remdesivir, hydroxychloroquine and lopinavir-ritonavir

 

NETWORK ESTIMATES & CERTAINTY

Network plots

Direct, indirect, and network estimates, including certainty

 

LATEST MANUSCRIPTS

Read the most updated data, 
published in The BMJ.
Read about adverse effects of remdesivir, hydroxychloroquine, and lopinavir/ritonavir. Article published on medRxiv, submitted to BMJ Open.
 

© 2020 by COVID-19 Living Network Meta-Analysis