AstraZeneca COVID Vaccine: Efficacy, Side Effects & Status

The AstraZeneca COVID-19 vaccine, known formally as AZD1222 or by its brand name Vaxzevria, demonstrated roughly 70–74% efficacy against symptomatic COVID-19 in large clinical trials and played a major role in global vaccination campaigns before the company voluntarily withdrew it from the market in 2024. Built on a chimpanzee adenovirus platform developed at the University of Oxford, it was one of the first vaccines authorized for emergency use worldwide, and its relatively simple cold-chain requirements made it especially important in low- and middle-income countries. Its story, though, is more complicated than a single efficacy number suggests, involving variant-driven shifts in real-world protection, a rare but alarming blood-clotting syndrome, and an eventual commercial exit that still shapes how vaccine science moves forward.

How the Vaccine Worked

AZD1222 belonged to a class of vaccines that use a harmless virus, modified so it cannot replicate, to deliver genetic instructions for the SARS-CoV-2 spike protein into human cells. In this case, the carrier virus was a chimpanzee adenovirus called ChAdOx1. Once inside your cells, the genetic payload prompted them to produce the spike protein on their surface, which trained the immune system to recognize and attack the real coronavirus if it showed up later. Several other COVID-19 vaccines used the same general approach with different adenoviruses, including the Johnson & Johnson (Janssen) vaccine and Russia’s Sputnik V.¹

The reason Oxford chose a chimpanzee adenovirus rather than a common human one was practical: many people have already been exposed to human adenoviruses and carry antibodies against them. Those pre-existing antibodies can intercept a vaccine’s viral vector before it delivers its payload, weakening the immune response. Using a chimpanzee adenovirus sidesteps most of that problem, since human immune systems rarely encounter it naturally.

Efficacy in Clinical Trials

The initial clinical picture of AZD1222 came from pooled data across randomized controlled trials in the UK, Brazil, and South Africa. An interim analysis published in The Lancet reported overall vaccine efficacy of about 70% against symptomatic COVID-19 when combining two dosing regimens. Among participants who received two standard doses, efficacy was around 62%. An unexpected finding was that a lower first dose followed by a standard second dose yielded efficacy of roughly 90%, though this subgroup was smaller and the confidence interval was wide.² Importantly, none of the hospitalized COVID-19 cases in those trials occurred in the vaccine group; all ten hospitalizations, including two severe cases and one death, were in people who received the placebo.

A larger Phase 3 trial conducted primarily in the United States, Chile, and Peru later reported a higher headline number. Overall estimated efficacy there was 74%, and among adults aged 65 and older, efficacy reached about 84%.³ The difference between the earlier and later trial numbers likely reflected variations in dosing intervals, population demographics, and the strains circulating at the time. Taken together, the trials indicated that two doses of AZD1222 provided solid protection against the original SARS-CoV-2 virus and early variants, with strong protection against severe disease and hospitalization.

Real-World Performance Against Delta and Omicron

Clinical trial efficacy, measured under controlled conditions, is one thing. How a vaccine performs in the messy real world, against evolving variants, is another. The Delta variant was the first major test. A large English study found that two doses of AZD1222 were about 67% effective against symptomatic Delta infection, compared with roughly 75% against the earlier Alpha variant.⁴ That drop was noticeable but not devastating, and protection against hospitalization and death remained considerably higher than protection against any symptomatic infection.

Omicron was a different story. By 20 weeks after a two-dose course of AZD1222, researchers in England found essentially no measurable protection against symptomatic Omicron infection. This was not unique to AstraZeneca; protection from all vaccine types waned faster against Omicron than against Delta. The solution was boosters. Among people who had received two AstraZeneca doses, a Pfizer (BNT162b2) booster restored effectiveness to about 62% in the first two to four weeks, though it faded again to around 40% by ten weeks. A Moderna (mRNA-1273) booster performed somewhat better, pushing effectiveness to about 70% initially and holding at roughly 61% through five to nine weeks.⁵

When it came to severe outcomes like hospitalization, the picture was more reassuring. Protection against hospitalization held up better than protection against any symptomatic illness, even with Omicron, and waning was less pronounced for the most severe outcomes.⁶ In practical terms, even when the vaccine could no longer reliably prevent you from catching a milder Omicron infection, it still reduced the chance of ending up in the hospital.

Common Side Effects

If you received the AstraZeneca vaccine, you likely felt it. Reactogenicity, the term for expected short-term side effects, was widespread and typically hit hardest after the first dose. Pain at the injection site was the most common local reaction, reported by roughly two-thirds to as many as 90% of recipients depending on the study population.⁷⁸ Swelling and redness at the injection site were also common.

Systemic side effects, the ones that make your whole body feel rough, included fever, headache, muscle pain, chills, and fatigue. In a large Bangladeshi observational study, fever was reported by about 85% of recipients and headache by about 82%.⁸ These symptoms typically appeared within a day or two of vaccination and resolved within about three days.⁹ One consistent pattern was that the first dose triggered more symptoms on average than the second. In the Iranian cross-sectional study, recipients reported an average of about six symptoms after dose one but only about three after dose two.⁹ This is the opposite of the mRNA vaccines, where the second dose was typically the rougher one, and it reflected different mechanisms of immune activation between the two vaccine types.

The Blood-Clotting Risk That Changed Everything

Within months of AZD1222’s rollout, reports emerged of a rare but serious condition that would eventually be called vaccine-induced immune thrombotic thrombocytopenia, or VITT. It involved unusual blood clots, sometimes in the brain’s venous sinuses or in abdominal veins, paired with a paradoxical drop in platelet count. Early surveillance data from the EudraVigilance database captured 28 thromboembolic reports out of more than 54,000 adverse reaction reports for the vaccine, with three fatalities among those early cases.¹⁰

Researchers proposed a mechanism in which the adenovirus vector interacts with a protein called platelet factor 4, or PF4. That interaction could trigger the production of antibodies against PF4, which in turn activated platelets and set off the clotting cascade.¹¹ The condition resembled a well-known drug reaction to heparin (heparin-induced thrombocytopenia) but occurred without any heparin exposure. VITT was genuinely dangerous when it struck, but the absolute risk was very low. Population-level estimates generally placed it in the range of a few cases per hundred thousand doses. Countries responded differently: some restricted the vaccine to older age groups (where the risk-benefit calculation favored vaccination more strongly), some paused use temporarily, and some continued administering it without restriction in settings where COVID-19 itself posed a far greater thrombotic risk.

VITT was the single most consequential factor in the vaccine’s eventual withdrawal. Although the condition was treatable once recognized, its emergence eroded public confidence and gave regulators in high-income countries reason to favor mRNA alternatives that did not carry the same signal.

The Guillain-Barré Signal

Beyond VITT, a neurological safety signal also drew scrutiny. Guillain-Barré syndrome (GBS), a condition where the immune system attacks peripheral nerves, was reported at a modestly elevated rate following AZD1222 vaccination. A systematic review drawing on English national immunization data found an incidence of roughly 0.6 cases per 100,000 first doses for all COVID-19 vaccines combined. No increased risk was linked to the Pfizer vaccine specifically, but AstraZeneca vaccination was associated with about double the baseline risk of GBS.¹² A separate analysis of case reports found a statistically significant association between AstraZeneca and the most common form of GBS.¹³ Another study placed the relative risk even higher, at roughly three times baseline.¹⁴

Context matters here. The risk of developing GBS after a COVID-19 infection itself was substantially higher than the risk from the vaccine. In the English data, a positive COVID-19 test was associated with more than five times the baseline risk of GBS, far exceeding the roughly twofold risk seen with AstraZeneca vaccination.¹² GBS is also associated with several other common infections and vaccines; the signal was not entirely unexpected for an adenovirus-based product, but it added to the growing list of reasons regulators scrutinized AZD1222 more closely than the mRNA vaccines.

Mix-and-Match Boosting

One of the more useful findings to come out of the AstraZeneca era was that mixing vaccine platforms, receiving an AstraZeneca primary course followed by an mRNA booster, often produced a stronger immune response than sticking with the same vaccine throughout. Research consistently showed that heterologous (mixed) schedules generated higher antibody levels and stronger cellular immune responses without significantly increasing side effects.¹⁵

A randomized trial in Indonesia found that a Pfizer booster after an AstraZeneca primary series produced significantly higher antibody and neutralizing antibody levels than a homologous AstraZeneca booster, for both half and full booster doses.¹⁶ A Thai study examining multiple vaccine combinations found that the AstraZeneca-then-Pfizer sequence was the most effective at sustaining antibody levels over 90 days in people without diabetes, with anti-spike antibody levels remaining above 3,000 units per milliliter at that point.¹⁷ That same study noted that people with type 2 diabetes had somewhat weaker antibody responses, though their neutralizing antibody levels with the mixed AstraZeneca-Pfizer regimen were comparable to those of people without diabetes.

These findings had direct policy implications. Many countries that had rolled out AstraZeneca as their primary vaccine pivoted to mRNA boosters, and the mix-and-match data gave them confidence that switching platforms mid-course was not just acceptable but potentially advantageous.

T-Cell Immunity and Long-Term Protection

Much of the public discussion about vaccine effectiveness focused on antibody levels, which are easier to measure and tend to correlate with protection against infection. But T cells, the other major arm of adaptive immunity, are at least as important for long-term protection against severe disease. Research comparing different vaccination regimens found that all of them, including AstraZeneca-based schedules, induced broad and long-lasting spike-specific T cell immunity. These T cells were diverse, capable of producing multiple immune signaling molecules, and remained stable over time. Booster doses did not significantly further enhance the T cell response, suggesting the initial vaccination had already established a durable memory.¹⁸ The T cell response also showed cross-recognition of Omicron subvariants, which helps explain why vaccinated people remained protected against severe disease even when antibody-mediated protection against infection waned sharply.

Protection Against Long COVID

A question many people had, and still have, is whether vaccination reduced the risk of long-lasting post-COVID symptoms. A multi-country staggered cohort study using data from the UK, Spain, and Estonia found that even a first dose of any COVID-19 vaccine roughly halved the risk of developing long COVID symptoms compared with being unvaccinated. The protective effect was seen across all the datasets examined, with hazard ratios ranging from about 0.48 to 0.71 depending on the country. When AstraZeneca and Pfizer were compared head-to-head, the data suggested Pfizer had a slightly stronger protective effect against long COVID, though the difference was modest and the confidence intervals overlapped in some of the datasets.¹⁹

A Logistical Advantage in Global Vaccination

One dimension of the AstraZeneca vaccine that rarely gets enough credit is its practical suitability for mass deployment in resource-limited settings. The vaccine could be stored at standard refrigerator temperatures (2–8°C), unlike the Pfizer vaccine, which initially required ultra-cold storage at around minus 70°C. This made a huge difference for countries without extensive cold-chain infrastructure. Research even suggested the vaccine was more resilient than its official storage specifications required: a study found that doses stored at 21°C for 18 hours before administration showed comparable safety and efficacy to properly refrigerated doses.²⁰ That kind of thermal tolerance could have reduced vaccine waste in tropical regions where maintaining a cold chain during transport is a constant challenge.

AstraZeneca also committed early on to supplying the vaccine at cost during the pandemic, and it was manufactured under license by the Serum Institute of India (under the name Covishield), which became the world’s largest producer by volume. Billions of doses were distributed globally through the COVAX initiative. For many countries in Africa, South Asia, and Latin America, AZD1222 was the first COVID-19 vaccine available at any scale.

Pregnancy and Maternal Antibody Transfer

Data on COVID-19 vaccination during pregnancy accumulated gradually, and most early guidance prioritized mRNA vaccines. However, research showed that pregnant women who received COVID-19 vaccines, including adenoviral vector types, developed immune responses and produced antibodies that transferred across the placenta to the fetus. A review of available evidence concluded that maternal vaccination led to detectable transplacental antibody transfer, though the transfer rate and duration of those antibodies in newborns needed further study.²¹ In practice, most countries eventually recommended mRNA vaccines as the preferred option during pregnancy, partly because more safety data accumulated for those products and partly because of the VITT concern, but millions of pregnant women worldwide received AstraZeneca without widespread signals of harm to the pregnancy itself.

Why AstraZeneca Withdrew and What Comes Next

In May 2024, AstraZeneca voluntarily withdrew the marketing authorization for Vaxzevria in the European Union and other markets. The company framed this as a commercial decision: demand had fallen to essentially zero as updated mRNA vaccines became the standard for boosters, and the original formulation was designed against the ancestral SARS-CoV-2 strain, which was no longer circulating. The withdrawal came amid ongoing litigation in several countries from individuals who developed VITT, and the timing was inevitably read through that lens, but the commercial explanation was straightforward. There was no longer a market for the product.

The adenoviral vector platform itself, however, is far from finished. Researchers are working on next-generation designs that address the limitations exposed during the COVID-19 campaign. Pre-existing immunity to the vector backbone remains a challenge: people who received an adenovirus-based COVID-19 vaccine now carry antibodies against that vector, which could blunt the effectiveness of future vaccines using the same approach. Strategies under development include chimeric capsids engineered to dodge those antibodies, self-amplifying RNA payloads carried by viral vectors, and the use of insect-specific viruses that human immune systems have never encountered.²²²³ The goal is a platform that can be rapidly adapted to new pathogens while avoiding both the rare clotting risks and the immunological ceiling that comes from reusing a familiar vector. The lessons from AZD1222, both the successes and the setbacks, are embedded in every one of those design choices.