Interim efficacy results were available and are reported for two of the four ongoing trials (from the UK and Brazil) based on cases occurring within approximately 4 months of follow-up in 11 636 participants, the majority of whom were aged 18–55 years (10 218 [87·8%] participants), white (9625 [82·7%] participants), and female (7045 [60·5%] participants). No COVID-19-related hospital admissions occurred in ChAdOx1 nCoV-19 recipients, whereas ten (two of which were severe) occurred in the control groups. Vaccine efficacy for the prespecified primary analysis (combining dose groups) against the primary endpoint of COVID-19 occurring more than 14 days after the second dose was 70·4% (95·8% CI 54·8 to 80·6; 30 [0·5%] of 5807 participants in the ChAdOx1 nCoV-19 group vs 101 [1·7%] of 5829 participants in the control group). Surprisingly, however, efficacy was substantially lower in the SD/SD cohort (62·1% [95% CI 41·0 to 75·7]; 27 [0·6%] of 4440 vs 71 [1·6%] of 4455) than in the LD/SD cohort (90·0% [67·4 to 97·0]; three [0·2%] of 1367 vs 30 [2·2%] of 1374), which remained after accounting for differences in age and time between doses. Efficacy was similar when evaluated starting at 21 days after the first standard dose (192 cases), suggesting there is at least short-term protection with one dose. Although efficacy was lower (58·9% [1·0 to 82·9]) against asymptomatic infection in the LD/SD cohort (and unfortunately only 3·8% [−72·4 to 46·3] in the SD/SD group), the results nonetheless provide some hope that COVID-19 vaccines might be able to interrupt some asymptomatic transmission, although fewer data (69 cases among 6638 participants) were available with this outcome and more data are needed to confirm. Only 1418 (12·1%) of those assessed for efficacy were older than 55 years (none of whom were in the LD/SD cohort), meaning that from the interim analysis of these trials, we cannot yet infer efficacy in older adults, who are the group at greatest risk of severe COVID-19 outcomes.
https://www.thelancet.com/journals/l...623-4/fulltext