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

expert reaction to QCovid tool used to identify which groups are at highest risk of hospitalisation and death after being vaccinated

A study, published in The BMJ, reports a new tool to identify which groups are at highest risk of hospitalisation and death after being vaccinated against COVID-19.

This Roundup accompanied an SMC Briefing 


Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:

“In my view, this is an excellent study in statistical terms, and it provides useful information and a potentially very useful prediction tool to identify those most at risk of hospitalisation or death after vaccination against Covid-19. The statistical prediction tool was developed from linked datasets giving information on almost 7 million vaccinated adults. Importantly, the predictions from the system were validated using data from an independent set of over 600,000 other patients. (The validation dataset did come from the same overall database, but used data from a completely different set of GP practices.) It’s crucial to carry out a validation of this kind, to check that the tool gives good results generally, and not only on the original dataset on which it was developed – and the validation results are very satisfactory.

“There are inevitably a few limitations of the study, and these are made clear in the research report. An overall point is that this is an observational study, so it can’t clearly establish that the risk factors that it found actually cause the increases in risk of hospitalisation and death to which they are linked. But that’s pretty well irrelevant in the context of a predictive tool like this – the idea is to identify patients at higher risk, regardless of whether the higher risk is actually caused by the identified risk factors or is the result of some more complicated pattern of cause and effect. The researchers mention that the follow-up period, during which Covid-related hospitalisations and deaths were identified, had to be relatively short. That’s inevitable in developing a risk prediction system that needs to be used as soon as possible during a still-developing vaccination programme.

“One good aspect emerging from the data, though it’s a problem for the development of the risk prediction system, is that very few hospitalisations or deaths were observed for people who had been vaccinated twice, fourteen days or more after their second dose. In the study, over 5 million people had two vaccine doses, but out of the 5 million, there were only 71 hospital admissions for Covid-19 and 81 Covid-related deaths during the follow-up. There would doubtless have been some more deaths and hospitalisations in fully vaccinated people if the follow-up could have been longer, but the low numbers do indicate to a considerable extent how good the vaccines are at preventing serious illness and death in people who have had two doses. The problem for the development of the system is that, really, the key number for this sort of development is not how many patients there are in all, but how many of the outcome events occurred. And 71 and 81 events are not high numbers of outcome events for a study such as this. (The extreme case would be if nobody at all was hospitalised in a study predicting the risk of hospitalisation. Obviously it couldn’t tell you anything at all about risk factors for hospitalisation, even if it used data on 10 million patients.) But the risk prediction model also uses data from the patients who had had only one dose, where there were considerably more Covid-related hospitalisation and deaths, sadly, and there’s no statistical indication from the analysis that the pattern of risk factors was very different after two doses than after one – apart from the fact that having had both doses reduces the risk of Covid-related hospitalisation and death to around a fifth of what it would be after just one dose. That’s yet another indication of how important it is to have both vaccine doses.

“The press release does say that the researchers are “from the University of Oxford”. That’s really the only misleading thing in it, in my view. It’s true that two of the researchers, including the lead author and another who played a major role in the study, are from Oxford. But, as the rest of the release makes clear, many others from many other institutions were involved. Altogether there are 18 authors of the report, from 12 different organisations across the UK. This is truly a broad-based piece of research.”


Prof Penny Ward, Independent Pharmaceutical Physician, Visiting Professor in Pharmaceutical Medicine at King’s College London, said:

Is this good quality work and are the conclusions backed up by the data?

“This paper from the Oxford group extends their prior work, conducted using digitised health information, on factors leading to an increased risk of severe outcomes from COVID, this time focusing on breakthrough infections occurring among the vaccinated population. We already know that, while highly effective in preventing infection and illness, vaccination is not 100% effective in preventing either in the general population. Risk of a more severe outcome are higher in certain patient populations, and this analysis demonstrates that some of these risk factors remain in the event of a breakthrough infection occurring despite vaccination. In particular, the assessment highlights increasing age, immunocompromise (presumably weakening vaccine response) and a number of pre-existing medical conditions as conferring a higher risk of a more severe outcome if the individual becomes infected despite having been fully vaccinated. The study does not compare risk of infection/illness among vaccinated and unvaccinated persons, but we know from other work that, for the most part, vaccination continues to offer a high level of protection notwithstanding other risk factors, with the possible exception of the more severely immunocompromised.

What are the strengths and limitations of the work and how do they effect the interpretation?

“The principle limitation of this work is that healthcare datasets do not include information on whether or not individuals have been directly exposed to contact with a covid infected person, so the contribution of closeness of exposure to breakthrough infection is not available. However one presumes that previous work demonstrating the importance of relatively prolonged close contact with an infected case in transmission still holds.

How worried should we be about these findings? Are they surprising or quite expected in terms of which groups are at increased risk?

“The data show, unsurprisingly, that factors shown previously to increase the risk of severe disease if infected are similar among vaccinated subjects becoming infected despite vaccination. These factors enable identification of a group of subjects that may require additional treatment to prevent infection i.e. post exposure prophylaxis with monoclonal antibody if they come into contact with a case, or early treatment with a monoclonal antibody treatment to ameliorate disease should they become infected. This information can be used to evolve a target population for access to these additional treatments in these circumstances.

Can you help put this increased risk into context for these groups e.g. compared to if they had not been vaccinated? Are they still provided protection from the vaccine?

“As noted before, this study does not compare unvaccinated to vaccinated persons, but we already know that most of these high risk groups are protected against infection and illness by completing the vaccine course. Most may also be afforded additional protection from the forthcoming booster doses being instituted now.

Are there any direct policy implications of this work e.g. on shielding/ booster doses etc? And how does that fit with current plans in the UK?

“Most of those at higher risk of an adverse outcome may be expected to increase their antibody response when given a booster dose. However, some populations, particularly those with immunecompromise, may remain at risk despite vaccination, and might be well advised to consider ways to reduce possible exposure to infection. Those that come into contact with infection would potentially benefit from the available monoclonal antibody treatments becoming available, to prevent infection post exposure, or, if unwell, to early treatment to prevent more severe disease.”


Dr Peter English, Retired Consultant in Communicable Disease Control, Former Editor of Vaccines in Practice, Immediate past Chair of the BMA Public Health Medicine Committee, said:

“There have been several papers looking at risk, in order to better target interventions to prevent Covid-19 in occupational settings,1-3 and for the wider population.

“This paper describes a tool developed to help patients, the doctors treating them to identify more accurately their own particular risk of hospitalisation or death from Covid-19, on the basis of a range of factors that have been identified as affecting risk. Similar tools – such as the well-established QRisk tool for calculating cardiovascular risk, have been in use for years.4 (Some of the authors of today’s paper, were involved in developing the QRisk tool.)

“A “research only” version of the calculator can be found at – please note that you should only use it to get a feel for the tool, not to undertake a “real” risk-assessment, as this version states: “This implementation of the QCovid risk calculator is NOT intended for use supporting or informing clinical decision-making. It is ONLY to be used for academic research, peer review and validation purposes, and it must NOT be used with data or information relating to any individual.”5

“As time goes by, more data are accumulated, and more hypotheses are generated and tested, such risk calculators can be updated and improved. Evaluations of previous versions of this QCovid tool have been published.6

“The value of such tools is to identify people who may benefit from additional measures to protect them from exposure, or to reduce their risk of disease after exposure. We cannot give everybody who is exposed antivirals or monoclonal antibody therapy to prevent the disease developing to the serious autoimmune phase, for example; but we could consider such treatments for some; and the tool may also assist policy-makers in decisions about, for example, whom to prioritise for earlier, more frequent vaccination, or vaccination with new anti-variant vaccines. The tool might help some people to make better-informed decisions about what level of “shielding” is appropriate for them. Will they, for example, feel more comfortable using public transport, or having visitors in their house, if they know their risk score more accurately?

“The research described in this paper identified a number of risk factors for severe Covid-19 outcomes despite being vaccinated: Down’s syndrome, kidney transplantation, sickle cell disease, care home residency, chemotherapy, recent bone marrow transplantation or a solid organ transplantation ever, HIV/AIDS, dementia, Parkinson’s disease, neurological conditions, and liver cirrhosis.

“As academic papers go, I think this one is not too challenging for a lay reader, and the press release is accurate and fair.

“As more data become available, I would expect the tool to be updated to incorporate new knowledge, such as:

  • Information on the impact of new variants.
  • Vaccines have only been used since late December 2020 and we know relatively little about the duration of protection; we have only recently extended the primary course to three doses for some at higher risk, we have only just introduced routine booster doses for people aged 50+ or with underlying conditions. Our surveillance programmes will provide a lot more information over time.
  • We still know relatively little about longer-term sequelae of Covid-19 infection. As more is learned, it is possible that future iterations of the calculator will incorporate these.
  • The impact of other policy interventions, such as non-pharmaceutical interventions, might become clearer.”


  1. Williams K, Cherrie JW, Dobbie J, Agius RM. The Development of a Covid-19 Control Measures Risk Matrix for Occupational Hygiene Protective Measures. Annals of Work Exposures and Health 2021. (
  2. Jankowski j, Davies A, English P, Friedman E, McKeown H, Rao M, et al. Risk Stratification for Healthcare workers during the CoViD-19 Pandemic; using demographics, co-morbid disease and clinical domain in order to assign clinical duties (Version 2). medRxiv 2020:2020.05.05.20091967. ( or
  3. Strain WD, Jankowski J, Davies A, English PM, Friedman E, McKeown H, et al. Development of an Objective Risk Stratification Tool to facilitate workplace assessments of healthcare workers when dealing with the CoViD-19 pandemic. BMA 2020; Updated 03 Jul 2020; Accessed: 2020 (03 Jul): ( or via
  4. Hippisley-Cox J, Coupland C, Vinogradova Y, Robson J, Minhas R, Sheikh A, et al. Predicting cardiovascular risk in England and Wales: prospective derivation and validation of QRISK2. BMJ 2008;336(7659):1475-1482. (
  5. Oxford Computer Consultants. QCovid® risk calculator. 2021; Updated v0.2.3 (undated); Accessed: 2021 (17 Sep): (
  6. Nafilyan V, Humberstone B, Mehta N, Diamond I, Coupland C, Lorenzi L, et al. An external validation of the QCovid risk prediction algorithm for risk of mortality from COVID-19 in adults: a national validation cohort study in England. Lancet Digit Health 2021;3(7):e425-e433  PMID: 34049834. (



‘Risk prediction of covid-19 related death and hospital admission in adults after covid-19 vaccination: national prospective cohort study’ by Julia Hippisley-Cox et al was published in The BMJ at 23:30 UK time on Friday 17 September.

DOI: 10.1136/bmj.n2244

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


Declared interests

Prof Kevin McConway: “I am a Trustee of the SMC and a member of its Advisory Committee.  I am also a member of the Public Data Advisory Group, which provides expert advice to the Cabinet Office on aspects of public understanding of data during the pandemic.  My quote above is in my capacity as an independent professional statistician.”

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).”

Dr Peter English:  Dr English is on the editorial board of Vaccines Today: an unpaid, voluntary, position. While he is also a member of the BMA’s Public Health Medicine Committee, this comment is made in a personal capacity. Dr English sometimes receives honoraria for acting as a consultant to various vaccine manufacturers, most recently to Seqirus.None received.

in this section

filter RoundUps by year

search by tag