Waning of COVID-19 vaccine effectiveness: individual and public health risk.

High coverage rates of vaccination against COVID-19 are envisaged to end the pandemic. However, waning of vaccine-induced protection is a growing concern that has been fostered by data on vaccine effectiveness against the currently circulating omicron SARS-CoV-2 variant of concern (VOC).

A systematic review and meta-regression by Daniel R Feikin and colleagues in The Lancet provides robust evidence of waning vaccine effectiveness over time. The authors identified 18 studies matching their inclusion criteria, of which three were randomised controlled trials (RCTs). Studies with participants of any age were included in the main analysis, with nine of 18 studies also including adolescents (aged ≥12 years). Information on sex or ethnicity distribution per study was not provided. Evidence from the meta-regression suggested a decrease in protection against SARS-CoV-2 infection by 21·0% (95% CI 13·9–29·8; on the basis of evidence from six studies) over a 6-month period from full vaccination across all ages and for all investigated vaccine types (Pfizer–BioNTech Comirnaty, Moderna-mRNA-1273, Janssen-Ad26.COV2.S, and AstraZeneca-Vaxzevria). Vaccine effectiveness against severe disease decreased by 10·0% (95% CI 6·1–15·4; on the basis of evidence from five studies); however, vaccine effectiveness against severe disease remained higher than 70% for 6 months. Subgroup analysis of studies with older adults (as defined per study, but with a minimum age of 50 years) showed no statistically significant difference when compared with analyses of all ages. Variant-specific time analysis supported that reduced vaccine effectiveness does not only relate to alternating effectiveness against specific variants, but that waning immunity is probable. The authors raised concerns about the serious risk of bias caused by confounding of several non-RCTs that were included. Primary data were adjusted for incomplete and different sets of covariates across studies, and considerable heterogeneity was detected. The pooled analysis might therefore be considered controversial, and the magnitude of estimated effects should be interpreted with caution. Nevertheless, all but one of the identified studies (a phase 3 RCT on mRNA-1273 with completed follow-up of the masked phase, median 5·3 months, before the emergence of VOCs) reported waning vaccine effectiveness over time.

The findings of Feikin and colleagues relate to the effect of waning immunity after full vaccination, without booster doses. Furthermore, they are restricted to evidence before the emergence of omicron. A report on population-based surveillance data from the UK illustrated waning of protection against symptomatic disease after two-dose and three-dose vaccination schedules, which was also observed when infections by the omicron variant began. The decline of protection in the UK was more distinctive for omicron than for delta. With an mRNA-based booster dose (Pfizer-BNT162b2 or mRNA-1273), vaccine effectiveness against omicron reached more than 60% 2 weeks after the booster dose. Approximately 4 months after the booster, a decline in protection was noted. Similar to the findings of Feikin and colleagues, a reduction in vaccine effectiveness against severe disease (ie, hospitalisation) was observed after full vaccination; however, this reduction was less great than that observed after symptomatic infection. The protective effect against hospitalisation after omicron infection could be restored up to 90% with an mRNA vaccine booster; a decrease to 75% 3–4 months after the booster was noted. US data from a test-negative study design support high vaccine effectiveness against omicron-related hospital admission after three doses (89%, 95% CI 84–92).

The need for repeated booster vaccination is widely discussed, with some countries starting to offer a fourth dose. Preliminary data from Israel suggest an increased protective effect against infection (risk reduced by a factor of 2·0, 95% CI 2·0–2·1) and severe illness (risk reduced by a factor of 4·3, 2·4–7·6) 12 or more days after dose four when compared with people who received three doses. Optimal vaccination strategies are being sought, and heterologous vaccination schedules, optimal time interval between doses, or variant-adapted vaccines are being discussed. The overall aim of vaccination against COVID-19 is to prevent severe disease and deaths. The prevention of severe disease is strongly related to the maintenance of a functioning health-care system, and thus combines the individual and public health risk of the COVID-19 pandemic. Therefore, the goal of public health measures is also to limit the spread of the virus and to interrupt transmission chains. Vaccination also has an effect on transmission rates; however the magnitude of effect changed in the light of arising VOCs. A study from Denmark investigated household transmissions of the omicron and delta variant. The secondary attack rate was approximately 10% lower in households with fully vaccinated primary delta cases and 20% lower in households with booster-vaccinated primary delta cases, when compared with unvaccinated primary cases. Although a vaccine-induced reduction of transmission under omicron was also observed (but to a lesser extent), the findings underline that the continuing emergence of new VOCs poses a threat to reducing the spread of SARS-CoV-2. Without sufficient vaccine coverage and equitable access to booster vaccination, waning of vaccine effectiveness represents both an individual and public health risk.

We declare no competing interests.