Study population and setting
This study describes the interim results of a prospective cohort study that examined the vaccine effectiveness of two messenger RNA (mRNA) vaccines – BNT162b2 (Pfizer-BioNTech) and mRNA-1273 (Moderna) – in real-world settings. The study collected data across 8 locations in the US between December 14, 2020 and March 13, 2021 on 3,950 health care personnel, first responders, and other essential and frontline workers (e.g. hospitality, delivery, retail, teachers) without prior laboratory confirmed SARS-CoV-2 infection. Participants self-collected a nasal swab weekly regardless of symptoms, and self-collected additional swabs if experiencing symptoms consistent with COVID-19; all swabs were tested for SARS-CoV-2 via RT-PCR. COVID-19 vaccination status was documented via self-report, telephone interviews, vaccine card upload at all sites, and medical records at some sites. Participants additionally self-reported COVID-19 associated symptoms weekly via text messages, email, or medical record reports.
Summary of Main Findings
Of the 3,950 participants (62% female, 86% white, 72% < 50 years old, 69% without self-reported chronic conditions), 205 were diagnosed with SARS-CoV-2 infection during the study period (11% asymptomatic). The incidence rates of SARS-CoV-2 infection were 1.38 per 1000 person-days among unvaccinated individuals, 0.19 per 1000 person-days among partially immunized individuals (≥14 days after first dose but before second dose), and 0.04 per 1000 person-days among fully immunized individuals (≥14 days after second dose), corresponding to site-adjusted vaccine effectiveness estimates of 80% (95% CI: 59-90%) and 90% (95% CI: 68-97%) among partially immunized individuals and fully immunized individuals, respectively. Vaccine effectiveness estimates individually adjusting for sex, age, ethnicity and occupation were similar to the main results. Notably, 33 PCR-confirmed infections occurring with 13 days of first or second dose of vaccination (67,483 person-days) were excluded from analyses.
This was a prospective and moderately sized cohort study. Individuals self-collected nasal swabs to test for SARS-CoV-2 infection weekly, regardless of symptom status which enabled capture of asymptomatic infection and assessment of the impact of vaccine on overall infection, rather than just symptomatic illness.
33 PCR-confirmed infections that occurred within 13 days of either vaccine dose were excluded, and it was not clear why the 13 days following second dose were considered a period of indeterminate immune status rather than partial immunization. Depending on how many of the 33 excluded infections occurred within 13 days of the second dose, the estimate of vaccine effectiveness among partially immunized individuals may be overestimated. Individuals with prior laboratory confirmed SARS-CoV-2 infection were excluded, so vaccine effectiveness estimates are unlikely to be generalizable to that group. Nasal swabs were self-collected, and symptoms were self-reported, which may have influenced the sensitivity of the test (especially if vaccination reduces viral shedding) or lead to misclassification. Only 3 infections occurred among people in the fully immunized group and 8 among people in the partially immunized group, which limited the ability to adjust for other potential confounders and led to wide confidence intervals.
This is one of the first real-world examinations of vaccine effectiveness for the Pfizer-BioNTech and Moderna mRNA vaccines.