Active vaccine safety surveillance: Experience from a prospective cohort event monitoring study of COVID-19 vaccines in Kenya

IntroductionActive vaccine safety surveillance (VSS) can complement passive VSS while overcoming the inherent limitations of spontaneous reporting systems. However, active VSS is rarely implemented in sub-Saharan Africa. We undertook active VSS of COVID-19 vaccines in Kenya. MethodsWe conducted a post-authorization cohort event monitoring study in Kilifi County, Kenya. Eligible individuals who had received any dose of any COVID-19 vaccine brand were followed up weekly over 13 weeks for adverse events, including hospitalization. A subset of participants was followed up daily for one week for solicited systemic reactogenicity events (chills, fatigue, fever, headache, joint pain, malaise, muscle aches, nausea). Follow up was done through telephone calls and/or short message service (SMS) with data captured electronically. Reports of adverse events following immunization (AEFI) were submitted as per national guidelines. We aimed to enroll 10,000 participants into the cohort and 1,000 participants into the reactogenicity sub-study. The daily prevalence of reactogenicity events was compared to the 3-day pre-vaccine average prevalence using McNemars test. The association of baseline characteristics with any systemic reactogenicity event was assessed using logistic regression. ResultsBetween 28th September 2022 and 30th June 2023, a total of 2,440 participants were enrolled into the cohort, out of which 1,000 systematically sampled participants were included in the reactogenicity sub-study. Most of the participants were aged 17-39 years (1683; 69.0%) and were female (1895; 77.7%); 535 (28.2%) female participants were pregnant. Only 34.4% of 814 participants contacted by SMS responded appropriately. In the reactogenicity sub-study, 595 (62.2% of 956 completing follow-up) participants reported [&ge;]1 systemic reactogenicity events. The three most frequently reported were fatigue (422; 44.1%), headache (370; 38.7%), and malaise (346; 36.2%). The proportion of severe systemic reactogenicity events ranged from 2.3% (22) for nausea to 5.0% (48) for malaise. Except for headache, the prevalence of systemic reactogenicity events was significantly higher in the first two days post-vaccination than in the three days prior to vaccination (p-values <0.05). The odds of reporting solicited systemic events were higher among non-pregnant women (adjusted odds ratio [aOR] 1.81; 95% CI 1.28-2.55) and pregnant women (aOR 1.69; 1.03-2.78) than among men. They were also higher among individuals receiving Johnson & Johnson (aOR 2.05; 1.40-3.00) or Moderna (aOR 4.19; 2.34-7.51) vaccine than among Pfizer vaccine recipients. The prevalence of pregnancy complications was 2.6% (95% CI 1.4-3.5%) against a background prevalence of 3-49%. ConclusionSystemic reactogenicity events following COVID-19 vaccination were non-severe and transient. There was no evidence of an elevated risk of pregnancy-related complications. The utility and feasibility of future active VSS studies could be improved by addressing context-specific challenges to enrollment and SMS-based follow-up.