select search filters
roundups & rapid reactions
before the headlines
Fiona fox's blog

expert reaction to preprint looking at the impact of the delta variant on vaccine effectiveness

A preprint, an unpublished non-peer reviewed study, looks at the impact of Delta on viral burden and vaccine effectiveness against new SARS-CoV-2 infections in the UK.

This Roundup accompanied an SMC Briefing.


Prof Sheila Bird, Formerly Programme Leader, MRC Biostatistics Unit, University of Cambridge, said:

“The ONS Infection Survey team has produced important evidence on vaccine effectiveness and cycle-threshold-value distribution (ct-value as proxy for viral load) for new positives in the Delta era.

“Although ONS Infection Survey’s incentivized response-rate is low (less than 20%), cardinal strengths that the authors rightly highlight include: a) survey-visits are routinely-scheduled, independently of symptoms, b) ONS Infection Survey has cross-checked self-reported immunization-status against England’s Immunization Management System to good effect, c) careful definition of positive-episode (duration of up to 120 days) and d) robust standard errors to account for multiple visits per participant.

“Eligible participants were aged 18+ years; but the vast majority of those aged 65+ years were already doubly-vaccinated by 16-May 2021 and relatively few remained unvaccinated. During the Delta era, which was defined as 17 May to 1 August 2021, some 359,000 participants {in 213,825 households} contributed nearly 812,000 study-visits and 3,123 first PCR-positives were identified. Secondary analyses therefore focus on those aged 18-64 years during the Delta era, who contributed some 297,000 study-visits (Table S1) but their number of first PCR-positives is surprisingly difficult to locate!

“Key messages from this secondary analysis are that double-vaccinated (+ at least 14 days) vaccine effectiveness (VE) against Delta-era infection was 82% (95% CI: 79% to 85%) for BNT162B2 {Pfizer-BioNTech} versus 67% (95% CI: 67% to 71%) for ChAdOx1 {Oxford-AstraZeneca}; and against  Delta-era infection with ct-value less than 30 was 86% (95% CI: 84%-88%) versus 69% (95% CI: 65%-73%). The significantly higher VE for BNT162B2 against Delta is close to the order of magnitude (20 percentage points) that the NIHR’s cancelled research-call for randomized comparison of dose-intervals had determined as policy-relevant.

“The ONS Infection Survey team has addressed dose-interval but the chosen cut-point was less than 9 weeks between 1st and 2nd dose whereas the dose-interval in randomized controlled trials of the MRNA vaccines (Pfizer-BioNTech and Moderna) was 3 or 4 weeks. A further complication is UK’s policy-shift on dose-interval (from 12 to 8 weeks) which occurred around the start of the authors’ definition of Delta-era.

“Cycle-threshold value, as analysed by the authors, is the lowest ct-value registered within a positive-episode (which may comprise several visits) and so, if anything, is nudged down to the ct-value that might be observed shortly before a person develops symptoms. The authors’ Figures 3B and 3C show the ct-value distribution for doubly-vaccinated new positive episodes, most of which occur in the Delta-era: reading off Figure 3C, there were of the order of 9,000 ct-values less than 30 but only around 3,000 would have been less than 18.3 (an extreme definition of “infectiousness” that has been promoted in support of INNOVA lateral flow tests) whereas only around 2,000 exceeded 25 (alternative cut-point for infectiousness that Liverpool mass screening interim report adopted).

“Importantly, the current analysis has also investigated (Table S4) within its double-vaccinated low ct-value PCR-positive sub-population how vaccine-type (PfizerBioNTech versus OxfordAstrazeneca) influenced ct-value mean: BNT162b2-escape ct-values were initially higher by 4.0 on average (95% CI: 1.5 to 6.6) but this advantage reduced (modelled as linear reduction) by 1.6 (95% CI: 0.6 to 2.5) per 30-days since double-protection was established. Further investigation along these pioneering  lines should be undertaken  – as a public health priority and without prevarication – within the mass of ct-values held by NHS Test & Trace for PCR-positives in the Delta-era. These data are linkable to Immunization Management Service to establish vaccine-type and dates. Although Pillar 2 PCR-testing is not independent of symptoms, it is surely worth putting into the public domain the intelligence that can be gleaned from NHS Test & Trace.

“Indeed, it is critical to do so as ONS Infection Survey team warn in their Discussion (p8) that “peak viral load now appears similar in infected vaccinated and unvaccinated individuals, with potential implications for onward transmission risk . . . Nevertheless, there may be implications for any policies that assume a low risk of onward transmission from vaccinated individuals (eg relating to self-isolation, travel), despite vaccines both still protecting against infection, thereby reducing transmission overall.”

“Since England’s change of policy this week on self-isolation comes within the above caution, I look to NHS Test & Trace to monitor the impact of the policy-change by putting its pertinent analyses into the public domain before August is out. This important paper gives a template to work from.”


Dr Simon Clarke, Associate Professor in Cellular Microbiology at the University of Reading, said:

“We’re seeing here the real-world data of how two vaccines are performing, rather than clinical trial data, and the data sets all show how the Delta variant has blunted the effectiveness of both the Pfizer and AstraZeneca jabs. 

“Of particular concern, the AstraZeneca vaccine’s effectiveness is reduced substantially by Delta and it appears to offer no more protection than what someone would get from having Covid-19 and building some natural immunity.

“The Pfizer jab provided greater initial protection than the AstraZeneca one, but then after around five months the level of immunity dropped to about the same level seen for both of the vaccines looked at. On this evidence, it certainly supports the case for third “booster” jabs for vulnerable individuals, as is now happening in Israel.  Moreover, these datasets scotch claims that protection provided by the UK’s vaccination programme is likely to be superior to Israel’s because of our longer window between first and second doses.”

“While vaccinated individuals are less likely to get Covid-19, we also see that it did not affect the amount of Delta variant which could be ‘shed’ from coughs and sneezes when someone does get infected. So although vaccines are still reducing cases and asymptomatic infection, we still need to take seriously the signs of potential infection when fully jabbed.”


Prof Stephen Evans, Professor of Pharmacoepidemiology, London School of Hygiene & Tropical Medicine, said:

“This is a good observational study on vaccine effectiveness. We need to be cautious in trying to draw conclusions about any specific biological explanations for the differences between the two vaccines (Pfizer BioNTech and Oxford AstraZeneca). The study was not designed to address this type of question.

“The really clear message has been set out by the researchers themselves- “‘However, whilst vaccinations reduce the chance of getting COVID-19, they do not eliminate it. More importantly, our data shows the potential for vaccinated individuals to still pass COVID-19 onto others, and the importance of testing and self-isolation to reduce transmission risk.”


Dr Penny Ward, Faculty of Pharmaceutical Medicine, Visiting Professor in Pharmaceutical Medicine, Kings College London, said:

“This investigation furthers our understanding of protection against COVID infections and symptomatic infection conferred by the various vaccines used in the UK and has important findings which may affect future policy. Firstly and as expected, the performance of the vaccines against the delta variant is not as great as observed previously during the alpha wave. No vaccine is completely protective against infection with the delta variant. This is not unexpected as sterilising immunity following vaccination is relatively uncommon and the purpose of vaccination is primarily to protect the individual against symptomatic, severe disease rather than from infection per se. Regrettably this paper does not include an analysis of protection against severe disease, but the low incidence of hospitalisation seen to date suggests that in this respect at least the vaccines are protecting individuals from developing severe COVID. We are not provided with information on the groups receiving the various vaccine types, and it is not immediately clear whether these are well balanced with respect to risk factors for severe disease. This information would be helpful in interpreting whether the observed differences in vaccine efficacy may be linked to patient factors rather than vaccine type given that protection is higher among younger subjects who were, potentially, more likely to have received the mRNA vaccines. Vaccine effectiveness appears to decline more rapidly after completion of a course of BNT162b2 vaccine than following ChAdOx01 vaccine, suggesting that a booster dose may be required to sustain effectiveness and indeed has just been recommended in the US. Although protection does not decline as rapidly following receipt of the ChAdOx01 vaccine, its lower overall protection suggests that a booster dose with a mRNA vaccine might enhance protection against this variant as winter approaches. As both protection against infection is reduced and the Ct values suggest viral load after infection may approach that in unvaccinated individuals there is potential for transmission of infection by vaccinated people. This suggests a need for continued focus on test, trace and isolate activities to reduce the level of spread across the community. All of this reiterates the pressing need for additional options for epidemic control within the community – this might include using monoclonal antibodies with activity against the delta strain early to prevent onward transmission of infection in households as well as early intervention with antiviral treatments both to further reduce risk of serious disease despite vaccination in an infected individual and to prevent onwards transmission to their close contacts. Perhaps we can look forward to a progress update from the Antiviral Taskforce soon.”


Prof Paul Hunter, Professor in Medicine, The Norwich School of Medicine, University of East Anglia, said:

“The study by Pouwels and colleagues is a very important contribution to our understanding of the relative effectiveness of the main vaccines in use in the UK and its comparison with the protective effect of a natural infection. Unlike most earlier studies this one looked at infections in a random sample of individuals rather than rely on people presenting for testing. As such it is likely to be less biased than these other studies. One previous study that has done this the REACT-1 study in its Round 13, but this did not compare the different vaccine types nor did it report on the protective effects of a natural infection. There is a wealth of information in the study by Pouwels and colleagues but for me the most important findings are

  1. All three vaccines currently in use in the UK are effective at reducing the risk of infection and the risk of symptomatic infection even from the Delta variant, especially after the second dose. It does look like AZ is less effective than Pfizer, though the long term difference is a bit more complex as the initial high effectiveness of Pfizer seems to diminish more quickly which is not what would have been expected Interestingly Moderna vaccine seems to be more effective against Delta after a single dose than either Pfizer or AZ as single dose or indeed after two doses AZ.
  2. The biological reason for why the mRNA vaccines seem to be more effective than AZ at preventing infection, is not certain at least to me. But early animal trials of AZ found that the vaccine “protected six monkeys from pneumonia, but the animals’ noses harboured as much virus as did those of unvaccinated monkeys” & unlike the mRNA vaccines which did reduce nasal shedding so the finding is perhaps not that surprising. I have previously discussed the difference between systemic and mucosal immunity and how vaccines administered systemically (by infection into the arm) may not be so good at mucosal infections. One possibility is that the AV vaccine uses the same vector for both first and second dose unlike the Russian sputnik vaccine which uses to different vectors (rad26-s+rad5-s) . The Sputnik vaccine does seem to have higher efficacy and it is possible that the reduced effectiveness of AZ is because of immunity against the adenoviral vector for the second dose reducing delivery However this hypothesis remains to be tested.
  3. The protective effect of a single natural infection is roughly similar to two dose immunization, with Pfizer being slightly more effective and AZ being less. However, given the fact that  many people with a previous natural infection will have had their infection quite a while ago, the apparent slightly lower effectiveness against delta for natural infection is likely to be an artifact.
  4. The paper does show that immunization after a natural infection offers quite a bit greater protection than immunization alone or presumably natural infection alone. Although not included in the paper it is likely that natural infection after previous immunization will also offer rather greater protection. We are already seeing cases of infection even after double doses and these are likely to rise over coming months.

“The main caveat to this paper is that it is in effect measuring vaccine effectiveness against asymptomatic or mildly symptomatic infection. There is now quite a lot of evidence that all vaccines are much better at reducing the risk of severe disease than they are at reducing the risk from infection. We now know that vaccination will not stop infection and transmission, although they do reduce the risk. The main value of immunization is in reducing the risk of severe disease and death and the evidence available shows that protection lasts longer against severe disease than against mild disease and all current UK vaccines are very good at this even against the Delta variant.  To me that is the most important value of immunizations.”


Dr Alexander Edwards, Associate Professor in Biomedical Technology, University of Reading, said:

“There are different ways to test how effective a vaccine is. Initially we relied on randomised clinical trials- these are great because you directly compare real vaccine with placebo, and can precisely measure protection. However, they don’t always tell you how useful the new medicine is “in the wild” because they can only test volunteer populations. The other way is to follow who actually gets infected after widespread vaccination. This study is powerful because it does exactly that- it shows how remarkably effective the vaccine is in the real world. What we really need to know is: how well do the vaccines delivered in the UK protect us during the current situation where high levels of virus are circulating including the new dominant delta variant. Of course, there are limitations to all study types- for this observation study, it’s not ideal to compare vaccinated vs unvaccinated people because there may be other differences between the groups- but that’s missing the point: the protection from the vaccines is so strong, the findings remain compelling.

“Overall this study is excellent as it shows that although delta is better at infecting vaccinated people than previous variants, the vaccines still work remarkably well. There are subtle differences- between different vaccine types (the mRNA types remain, as expected, measurably better than the adenovirus one), and some changes over time, but they all work brilliantly. Two doses remain significantly better than one.

“It does remain vital to remember that even if double jabbed, you can still get infected and pass the virus on.

“Whilst there are further details discussed in this major study – for example trying to analyse the highly variable PCR signal to explore if the amount of virus might be different between vaccinated or unvaccinated groups – these to a certain extent distract from the big headlines: vaccination protects against COVID-19 infection. Worldwide sharing of vaccine doses, and transparent information sharing to inform everyone of their safety, remains the most urgent response to this extremely harmful virus.”



‘Impact of Delta on viral burden and vaccine effectiveness against new SARS-CoV-2 infections in the UK’ by Koen B. Pouwels et al. is a preprint of work carried out by the University of Oxford, ONS and DHSC. It was under embargo until 00:01 UK time Thursday 19 August.



All our previous output on this subject can be seen at this weblink:



Declared interests

Prof Stephen Evans: “No conflicts of interest.  I am funded (one day per week) by LSHTM.  They get funding from various companies, including Astra Zeneca and GSK but I am not funded by them, I have no involvement in obtaining funding from them and I am not an investigator on any grants obtained from them.  I am the statistician to the ‘meta-Data Safety and Monitoring Board’ for CEPI.  I am paid for my attendance at those meetings and will be paid expenses for travel if that occurs.  I am a participant in the Oxford/Astra Zeneca trial, and on 13th January 2021 learnt I had received the active vaccine.”

Dr Penny Ward: “No COIs. I am semi-retired, but I am owner/Director of PWG Consulting (Biopharma) Ltd a consulting firm advising companies on drug and device development. Between December 2016 and July 2019 I served as Chief Medical Officer of Virion Biotherapeutics Ltd, a company developing antiviral treatments for respiratory viral diseases. Previous employee of Roche, makers of tocilizumab (anti IL6 antibody) and CMO of Novimmune, makers of empalumab (anti IFN gamma antibody). These are my personal views and do not reflect those of either institution.”

Dr Alexander Edwards: “No conflicts of interest.”

None others received.

in this section

filter RoundUps by year

search by tag