Summary of findings 1. Colchicine plus standard care compared to standard care (plus/minus placebo) for hospitalised patients with COVID‐19 and moderate to severe disease.

Colchicine plus standard care compared to standard care (plus/minus placebo) for hospitalised patients with COVID‐19 and moderate to severe disease
Patient or population: people with COVID‐19 and moderate to severe disease Settings: hospitalised Intervention: colchicine plus standard care Comparator: standard care (plus/minus placebo)
Outcomes	Absolute effects from study(ies)* (95% CI)			No of participants (studies)	Certainty of the evidence (GRADE)	Plain language summary	Comments
	Risk with placebo or standard care alone	Risk with colchicine	Relative risk [risk difference; 95% CI]
All‐cause mortality assessed up to day 28	207 per 1000	207 per 1000 (193 to 224)	RR 1.00 (95% CI 0.93 to 1.08)a [0 more per 1000; 14 fewer to 17 more]	11,445 (2 studies)	⊕⊕⊕⊖ Moderateb	Colchicine probably results in little to no difference in all‐cause mortality up to 28 days.	Additionally, one study reported all‐cause mortality at hospital discharge for 75 participants (Lopes 2021). One study analysed also reported all‐cause mortality over time (time‐to‐event) for 11,340 participants (Horby 2021), which similarly showed little to no effect on mortality (HR 1.01, 95% CI 0.93 to 1.10).
Worsening of clinical status: participants with clinical deterioration, defined as new need for invasive mechanical ventilation or death up to day 28	244 per 1000	249 per 1000 (234 to 266)	RR 1.02 (95% CI 0.96 to 1.09)c [4 more per 1000, 95% CI 10 fewer to 22 more]	10916 (2 studies)	⊕⊕⊕⊖ Moderateb	Colchicine probably has little to no impact on new need for invasive mechanical ventilation or death.	—
Improvement of clinical status:  participants discharged alive up to day 28 without clinical deterioration or death	704 of 1000	697 of 1000 (676 to 711)	RR 0.99 (95% CI 0.96 to 1.01) [7 fewer per 1000, 95% CI 28 fewer to 7 more]	11,340 (1 study)	⊕⊕⊕⊖ Moderateb	Colchicine probably results in little to no difference in improvement of clinical status, if this is measured with the number of participants discharged alive up to day 28 without clinical deterioration or death.	One study reported participants discharged alive at the longest follow‐up and followed all participants until discharge (Lopes 2021) which similarly showed that colchicine may result in little to no difference in the improvement of clinical status assessed as participants discharged alive (RR 1.09, 95% CI 0.98 to 1.21)
Quality of life, including fatigue and neurological status at longest follow‐up available	We identified no studies reporting quality of life.			—	—	We do not know whether colchicine has any impact on quality of life.	—
Adverse events (follow‐up: until discharge)	389 per 1000	389 per 1000 (218 to 692)	RR 1.00 (95% CI 0.56 to 1.78) [0 fewer per 1000, 95% CI 171 fewer to 303 more]	72 (1 study)	⊕⊖⊖⊖ Very lowd,e,f	The evidence is very uncertain about the effect of colchicine on adverse events	—
Serious adverse events (follow‐up: until hospital discharge or a maximum of 21 days)	0 events observed	0 events observed	Not estimable	105 (1 study)	⊕⊖⊖⊖ Very lowe,g,h	The evidence is very uncertain about the effect of colchicine on adverse events	—
*The basis for the control group absolute risks from the study(ies) is mean risk across study(ies) unless otherwise stated in comments. The intervention absolute risk and difference is based on the risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; COVID‐19: coronavirus disease 2019; HR: hazard ratio; RR: risk ratio. GRADE Working User Group grades of evidence High quality: further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: we are very uncertain about the estimate.

^aSensitivity analysis results presented using a fixed‐effect model. This was because the random‐effects model meta‐analysis resulted in very wide confidence intervals (RR 0.72, 95% CI 0.21 to 2.44), due to more weight being given to one small study with few events (Deftereos 2020), and contributing only 105/11,445 participants included in the mortality analysis.
^bDowngraded one level for serious study limitations due to the unblinded study design.
^cSensitivity analysis results presented using a fixed‐effect model. This was because the random‐effects model meta‐analysis resulted in very wide confidence intervals (RR 0.53, 95% CI 0.09 to 3.15), due to more weight being given to one small study with few events (Deftereos 2020), and contributing only 105/10,916 participants included in the analysis of the worsening of clinical status.
^dDowngraded two levels for very serious imprecision due to only one study with very small number of participants and events.
^eDowngraded one level for other considerations due to selective reporting of adverse events across studies (e.g. severe treatment‐associated events only).
^fDowngraded one level for serious study limitations due to the high risk of bias because of the competing event 'death.'
^gDowngraded one level for serious study limitations due to the unblinded study design, and high risk of bias because of the competing event 'death.'
^hDowngraded two levels for very serious imprecision due to only one study and no events were observed.