After the storm

The COVID-19 pandemic placed unprecedented demands on global healthcare systems, regulators and researchers. Vaccines were developed, approved and rolled out at exceptional speed.

While this should be seen as a success, the pace of development also brought into focus an essential component of pharmacovigilance: the need for robust, ongoing monitoring of vaccine safety in clinical practice.

Our recently study: Safety and utilisation of AZD1222 (ChAdOx1 nCoV-19) COVID-19 vaccine: a UK post-authorisation active surveillance study, provides one of the most comprehensive real-world safety evaluations of the AstraZeneca COVID-19 vaccine in the UK to date.

The study followed nearly 18,000 participants for up to 18 months after their first dose, collecting information about adverse events (AEs) at multiple time points.

Physicians assessed all participant-reported AEs and, for those classified as either serious or an adverse event of special interest (AESI), further information was sought from the participant’s GP. We observed no new safety signals and confirmed the most commonly reported AEs – headache and fatigue – were consistent with known side effects.

While we detected more cases for the AESIs anaphylaxis and anosmia (loss of smell) and/or ageusia (loss of taste) than expected, this was not altogether surprising since anosmia and ageusia are commonly associated with COVID-19 infection and anaphylaxis is listed as a potential side effect in the vaccine’s summary of product characteristics.

Large-scale, real-world surveillance studies are critical to vaccine development, implementation and uptake.

Not only do they ensure ongoing public trust, but they are also important in how we prepare for future pandemics. In this piece, we discuss the benefits and challenges of these studies and why they are needed long after a pandemic subsides.

While clinical trials remain the gold standard for determining efficacy and initial safety, they have practical limitations. They often involve highly controlled environments and a small, selected group of participants, excluding groups such as pregnant women, children, elderly people and patients on multiple medications.

As vaccines enter wider use, it’s essential to capture data from larger and more diverse groups across a range of health conditions, geographies and life circumstances. Active surveillance studies, such as described in this article, fill that gap.

Where enough people are included, these studies allow for the detection of rare but serious AEs, especially in subgroups that may have been under-represented in earlier trials.

Importantly, the design of our study allowed for real-time data sharing with the UK Medicines and Healthcare products Regulatory Agency (MHRA), thereby contributing to its signal detection activities and its continuous safety oversight during an ongoing public health emergency.

However, this real-time vigilance wasn’t just about AZD1222, it was about building an infrastructure capable of supporting public confidence in rapid medical responses in the future.

The COVID-19 pandemic exposed both the vulnerabilities and strengths of our global health infrastructure. One clear takeaway is the need for embedded surveillance systems that can be put into action quickly and at scale in response to emerging threats.

Our study highlights several critical components of pandemic preparedness:

• There need to be accelerated but safe approval processes for new medicines and vaccines, which are supported by strong post-authorisation monitoring. With regulatory agencies engaging with vaccine developers much earlier than usual, an accelerated regulatory approval pathway is possible, even likely, whereby the vaccine may be licensed as soon as efficacy is proven but with conditions of extensive post-marketing surveillance. Fundamentally, it is important that the speed of vaccine development does not come at the cost of vaccine safety and effectiveness
• There should be an ongoing focus on standardising case definitions and pre-prepared accurate background incidence rates for key AESIs, thus enabling more rapid identification of safety signals via Observed vs Expected analysis, a cornerstone of pharmacovigilance during a crisis
• Collaboration with general practitioners to obtain further clinical details for serious or complex cases is needed to ensure we capture real-world data. Active surveillance studies encourage participants to report, and subsequent data collection is structured. This creates a richer, more accurate data set, particularly for events that are subtle, subjective or have delayed onset.

While our study demonstrated the utility of the active surveillance methodology in a pandemic situation, this design had acknowledged limitations, many of which were related to the COVID-19 pandemic environment and can therefore be considerations for the next pandemic.

There may have been a reporting bias where participants who experienced AEs post-vaccination may have been more willing to join the study than those who did not.

This may have been heightened by the media attention and public awareness leading to inflated reporting of certain AEs such as anosmia and ageusia. In addition, participants could be more likely to report serious AEs.

Conversely, there may be underreporting by participants of very common or expected AEs such as injection site pain. The multiple follow-up time points were designed to collect short- and long-term outcomes and to reduce the risk of recall bias and loss to follow up by participants. However, attrition was evident at each time point, which may have resulted in underreporting across all AEs.

Finally, certain patient populations may be more likely to participate in studies leading to underrepresentation of some sub-groups such as ethnic minorities.

Real-world safety studies are more than regulatory formalities. They are public health imperatives. Transparent, high-quality data on long-term safety reinforces public trust, especially in times of uncertainty.

They demonstrate a continued commitment to protecting health beyond the initial emergency, offering reassurance that science remains at the service of society.

As we look to the future, the systems and lessons developed through studies like ours must not be shelved until the next crisis. Instead, they should be viewed as integral components of routine pandemic preparedness, ensuring that when the next emergency arises, we’re not just reacting, but ready.