covid19
COVID-19: does metformin affect outcomes?
Takeaway
- In this large study, metformin did not influence susceptibility or mortality in COVID-19.
Why this matters
- Recent observational studies suggesting reduced COVID-19 mortality with metformin have included mostly hospitalized patients, with limited adjustment for confounders.
Study design
- UK primary care data set of participants with diabetes using metformin (MF+) and not using metformin (MF−) (n=10,183 for each) along with other glucose-lowering medications.
- Participants were propensity score-matched for demographic and behavioral risk factors, diabetes duration, complications, comorbidities, and other potential confounders.
- Funding:None.
Key results
- Suspected/confirmed COVID-19 found in 415 MF+ vs 188 MF−.
- Crude incidence per 1000 person-years: 20.4 MF+ vs 26.9 MF−.
- With propensity score matching, 172 MF+ vs 186 MF− had confirmed/suspected COVID-19.
- Incidence per 1000 person-years: 24.7 MF+ vs 26.8 MF−.
- For MF+ vs MF−, aHRs (95% CIs) for primary care consultations:
- 0.85 (95% CI, 0.67-1.08) for suspected/confirmed COVID-19.
- 0.80 (95% CI, 0.49-1.30) for confirmed COVID-19.
- All-cause/COVID-19 deaths: 403/31 of MF+ vs 275/21 MF−.
- Crude incidence per 1000 person-years: 19.7 MF+ vs 39.0 MF−.
- Following propensity score matching, 17/214 of MF+ and 20/266 MF− deaths were COVID-19-related.
- No significant difference in all-cause mortality (aHR, 0.89; 95% CI, 0.74-1.07) or COVID-19-related death (aHR, 0.87; 95% CI, 0.34-2.20).
Limitations
- Reliance on coding.
- Unknown adherence.
- Low event rates.
- No hospitalization data.
Only healthcare professionals with a Univadis account have access to this article.
You have reached your limit of complementary articles
Free Sign Up Available exclusively to healthcare professionals Register to read more