Background

The first vaccine for COVID-19 received Emergency Use Authorization by the US Food and Drug Administration (FDA) in December 2020.¹ As new SARS-CoV-2 variants emerge, evidence has continued to support the use and updating of boosters against the virus.² In August 2022, the United Kingdom first granted conditional approval for a bivalent booster targeting the BA.1 strain as well as the ancestral strain.³

This and subsequent approvals were based on nonrandomized, open-label trials of Omicron-containing boosters that indicated superior immunogenicity against the new variants compared with the original vaccine.⁴ To further test the relative safety and efficacy of different booster formulations, Lee and colleagues conducted a larger, randomized trial of both monovalent and bivalent Omicron-targeted boosters using active control groups and observer blinding.⁵

Methods

Across 28 sites in the UK, the researchers enrolled participants aged 16 and older who had received at least 2 initial doses of an approved SARS-CoV-2 vaccine series. In an initial segment of the study intervention, 366 participants received the BA.1-containing monovalent booster and 357 received the original mRNA-1273 vaccine formulation. In the following segment, 1418 participants received the BA.1 bivalent booster, while 1395 received the original mRNA-1273.⁵

Results

The primary study outcomes were safety and immunogenic response; the latter was indicated by changes in antibody titers 29 days after booster administration. Notably, no new safety concerns arose during the study. Immunogenic response to both monovalent and bivalent Omicron-containing boosters demonstrated superiority to the original vaccine and were noninferior against the ancestral strain of the virus. Exploratory analysis also suggested that incidence rates of infection with BA.2 (both bivalent and monovalent boosters) and BA.4 isolates (bivalent boosters only) were also lower than those obtained with the original mRNA booster. Neither of the updated boosters appeared to affect incidence of infection with the BA.5 variant.⁵

The superior immunogenicity of the newer boosters supported the hypothesis that matching booster formulations to dominant circulating SARS-CoV-2 variants increases the effectiveness of COVID-19 infection control strategies. The researchers acknowledged the need to monitor real-world effectiveness in future research on boosters for COVID-19.

Discussion

Onyema Ogbuagu, MBBCh

Adaora Okoli, MD, MPH

Francisco Machiavello Roman, MD

Onyema Ogbuagu, MBBCh, FACP, FIDSA, is an associate professor of medicine in infectious disease at the Yale School of Medicine, New Haven, Connecticut. He is also affiliated with the Yale Institute for Global Health and directs the Yale AIDS Clinical Trials Program. Adaora Okoli, MD, MPH, and Francisco Machiavello Roman, MD, are clinical fellows in infectious disease at the Yale School of Medicine. The 3 clinician-researchers discuss this research report.

Onyema Ogbuagu, MBBCh (OO): Dr Okoli and Dr Machiavello Roman, thank you for taking time during a busy day to participate in this Journal Club.

If you were flipping through a journal and you saw the title of this paper, what would make it an article you would want to look at? Is there anything about the topic that grabs your attention immediately?

Adaora Okoli, MD, MPH (AO): I think what grabbed my attention was that during the pandemic, when we had rapid evolution of the virus and different strains emerged, it was difficult to carry out a large clinical trial because of how quickly things changed. The fact that this study had a pretty sizable number of participants and was randomized and observer-blind makes it the gold standard we want vaccine trials to achieve.

OO: I agree. As COVID-19 hit, there were a lot of observational studies, single-arm studies, generally small studies. A lot of the data that we had, frankly, was relatively lower quality. Paying attention to the quality of data being generated on these topical issues is critical.

The fact that this was a randomized clinical trial got Dr Okoli’s attention. Dr Machiavello Roman, you were in the front lines as well. From what we know about immunity and what we have observed of COVID-19, why do we need boosters in the first place?

Francisco Machiavello Roman, MD (FMR): What we learned from COVID-19 is that we were never fully safe, even after the first wave of the pandemic. We knew from virology studies that mutations were underway. And as they have hit us, we have seen clusters of disease, as small epidemics within the major pandemic have occurred with the new variants. So boosters and variants of concern became the topic of the day within the infectious disease community. We wanted to make sure that if there was a new variant of concern with potential for an epidemic, we would be offering some degree of protection. In this sense, changing our vaccines to make sure that these new variants of concern were included made perfect sense for everyone within the community.

OO: I agree; these are excellent points. If I can summarize: we saw immunity wane both from natural infection and from vaccines, which was really what started the conversation around boosters. When we started to hear about breakthrough infections in people who had completed their primary series, it became pretty clear that something was going wrong. And then there were people who had hybrid immunity but still were infected even post-vaccination, or had even second and third reinfections. It was pretty clear that we were not winning the immunity game with COVID-19.

New variants continue to emerge with COVID-19, and we know part of what drives immune evasion is simply prior infection and vaccination. Viruses are smart and they try to evade a preexisting immunity, no matter how that was generated. We know there are still ongoing high levels of community transmission, so we still need to protect vulnerable individuals and we want to continue to mitigate adverse clinical outcomes like hospitalizations, death, and — let us not forget — long COVID.

We all agree there was a clear need to have updated COVID-19 shots in the face of changing viral epidemiology and ongoing transmission, and because there was evidence of patient benefit. With this in mind, as you read the title of the paper and look at the author list, you will notice some unique features of the article. The Lancet is a highly reputable journal, with a high impact factor, and it appeals to an international audience. Last fall, in the US, which viral strains were we targeting with bivalent vaccines?

AO: Those were BA.4 and BA.5.

OO: Exactly correct. So, you can see that in other parts of the world with BA.1, monovalent and then bivalent boosters were available. But we did not even have that in the US. Again, this is an international journal, so we always need to pay attention to those little nuances around applicability in our locale. Still, we continue to learn from strategies in other parts of the world.

When you look at the author list, there is a good mix of hospitals, clinics, and investigators within those settings in the UK, and you also have involvement of investigators from Moderna. You both may not know, but I have been here long enough: Rituparna Das, the senior investigator on this study, many years ago was an Infectious Disease Fellow at Yale, just like you. And she has gone on to be a senior director in their research program. Moderna is manufacturing these vaccines and they are involved in this clinical trial. When you have involvement of investigators who have a vested interest, what are generally your concerns in this context?

FMR: Bias is a major concern when we are examining positive study results. Another thing I would be looking out for would be making sure that the results we are seeing are relevant to the study’s questions. Sometimes, if researchers are investing a lot of money in a project, they want to see positive results. If that does not happen, sometimes they will present surrogate results or data that are peripheral to the core issue. So those are things I watch out for.

AO: Yes, supporting Dr Machiavello Roman’s point is that you may see more positive results when a pharmaceutical company is funding the research, and the language can sometimes be more glowing. You have to take a close look and just be sure that you are getting the right information.

OO: Great, I agree with all the points you have made. You have highlighted some biases that can occur, including with any drug sponsor. There is also publication bias, where — especially when there is a need for good information in the COVID-19 era — studies reporting positive results tend to get advanced to publication and negative studies tend to be relatively suppressed. That is a phenomenon to be aware of, especially when you have investigators who are part of the company or study sponsor.

Let us move on to consider the evidence base regarding boosters. We have been talking about well-powered, randomized, controlled trials. But we know that when it comes to defining immunogenicity with variants, many previous studies are animal studies or small-n studies. What were the gaps in knowledge around boosters, or in previous studies to assess boosters, that this clinical trial tried to address?

FMR: The paper noted in the introduction that what we knew previously about immunogenicity and safety in these boosters was essentially based on nonrandomized data and on observational information.⁴ So they aimed to do a “proper” randomized, controlled trial to see if earlier hypotheses about immunogenicity and safety held true or not.

AO: In addition, not only were most of the previous studies not randomized, but they were also open-label. That means that participants knew what they were receiving, and that tends to introduce bias.

OO: Great point, Dr Okoli. Going further, when your endpoints include safety, where subjects are reporting side effects, and an investigator must adjudicate whether those side effects are related to an investigational product, biases can definitely emerge if the investigator is not blinded appropriately.

Now, in a real-world situation, we do need rapid information in the COVID-19 era about the role of boosters, and it takes time to design, conduct, and report on high-quality clinical trials. So we have had to default to smaller studies in the face of emerging variants. For example, in this study, they looked at the monovalent and bivalent BA.1 boosters, but at the same time, the viral epidemiology was changing: there was BA.2, and I think they even started to log BA.4 cases.

We do not want to discount the animal studies because they do provide valuable information about, for example, the breadth of neutralization or variant-specific immune responses. In fact, some of those precursor studies have helped inform how these larger studies are conducted.

Let us dive deeper into the methodology of the study. If we organize the discussion in the PICO (Patients, Intervention, Comparison, Outcome) format, then what were the considerations around the study population?

FMR: This study included people living in the UK, aged 16 and older. The participants were required to have had prior vaccination; this study was very specific that they needed to have completed a vaccine series. The authors also made a comment that a booster could be either a third or fourth dose. But given the rate of immunization in the UK, the authors expected that most participants would be getting their fourth dose — meaning their second booster — during the intervention. Indeed, this would be the sine qua non requirement for admissibility into the study. There would be no sense administering a booster for somebody who had not had the vaccine series first.

AO: I would like to note that the authors do not really describe the population in terms of their comorbidities — how sick or healthy they were. We do know that no immunocompromised patients were included in this study. They needed also to have had no evidence of COVID-19 in the past 90 days prior to enrolling. Since most had the fourth booster, this says that the study population probably has relatively good access to healthcare. It does not necessarily mirror what you would find in other parts of the world.

The mean age of the participants was 57, across all subgroups and with similar standard deviations. I did notice that quite a large proportion of the participants were between 16 and 65 years old, and a smaller proportion were older than 65. But we also know that most of our patients who present with COVID-19 and also have comorbidities tend to be in the older age group.

Part 2 of the study had slightly more sex equity, while part 1 had a slightly greater female representation. But what really stood out for me was the fact that there was not a single Black person in part 1 of the study, and in part 2, it was 0.4% Black participants. Just 2.2% to 2.9% of the study participants in each subgroup were Asian, so we definitely see a lack of diversity in the study population. Mean body mass index (BMI) was between 27 and 28. I did not see much other detail regarding their comorbidities or health status.

OO: Right. The journal did have supplementary material that might not fit in the main tables but could answer some of those additional questions. You have highlighted important things because some of these factors are known to influence vaccine responses.⁶ I agree that while at least one-third of participants were over 65, and there was good gender representation, these demographics would not represent racial minorities in the US.

FMR: I was really surprised that they were able to find a group where 86% of the participants had no known prior COVID-19 infection. I do not think I could find a group in the US where 86% of the population have not had COVID-19. These participants did not have natural COVID-immunity from acquired infection and may have been people who had followed vaccine recommendations closely and who were extremely careful about being exposed to COVID-19.

OO: That is a good point. They did try to eliminate those who had COVID-19 within 3 months or so prior to study entry; they also acknowledged that they could have missed people with asymptomatic infection. But note that for an immunogenicity study, you also want to eliminate the statistical noise of COVID-19 infection because that can boost immune responses. So even though it is less representative of people who have had COVID-19, I would expect this to reduce bias in the immunogenicity results they reported.

Now, the intervention was monovalent and then bivalent boosters. If the point was to study variant matching, such as matching the vaccine with a variant, why was a monovalent not enough? Why do you think they needed to include the bivalent vaccine?

AO: My understanding is that they had originally started doing the monovalent, and when they realized that most people were starting to get the bivalent vaccine, they switched over. It might not have been their plan from the beginning.

OO: There is a fascinating theory about what they call antigenic seniority, which has at least been described, as I recall, in 1 study in China, where they examined historically a strain of influenza. They took sera from individuals and performed neutralization tests with both older and newer forms of influenza virus, and the bottom line was that they found a participant’s age when a strain first circulated was predictive of the strength of their antibody against that strain, suggesting we produce progressively fewer specific antibodies to each subsequent infection as we age.⁷ So as newer variants emerge and we continue to give people additional vaccinations, 1 concern is this: are they primed by the original strain, such that the immune responses to newer variants will be blunted? That has been an open question.

You could argue that the variants were already emerging, so we should just match the vaccine to the variant. Or you could argue the other way, that we still should retain some of the original vaccine so that we are maintaining fidelity with the original antigenic insult. I think that was part of the question. And some of the precursor data suggested a little bit more breadth of protection when using a bivalent booster, compared with a monovalent⁸; that, of course, helped lead to its availability for use in this study.

These researchers looked at people receiving either the monovalent or bivalent version of the BA.1-containing vaccine compared with the original Wuhan-strain vaccine, which I think was appropriate. Note that many of the participants had received a mix-and-match regimen of primary-series vaccines, and I believe the majority — at least for the primary series —had received viral vector-based vaccines. This is a very different mix of vaccines compared with what you would see in the US. But again, almost everyone had already received at least 1 mRNA booster, so I would say that kind of “evens out” any concerns around the mix of pre-entry treatments.

But here is another question you might have: the primary outcomes in this study were safety and immunogenicity — as an infectious disease physician, if you want to assess a new vaccine, are you satisfied with surrogate endpoints such as immunogenicity responses, as compared with true clinical outcomes?

FMR: I feel that the incidence and severity of infections would have been more desirable study outcomes. The authors mentioned that this was explored in the supplementary data, but we do not have that information clearly upfront in the main paper. I would say that immunogenicity is possibly not the strongest surrogate result, but it does at least give us a biochemical demonstration that the vaccine is active against the virus.

OO: There was a nice paper in Nature by Khoury in May 2021. That paper showed how the breadth of neutralizing antibody responses mapped to reported efficacy in different SARS-CoV-2 vaccine trials. They showed that levels of neutralizing antibodies were at least closely, almost linearly, correlated with reported efficacy.⁹ So yes, neutralizing antibodies are a surrogate endpoint, but I think enough evidence suggests that they are a decent correlate of protection against symptomatic COVID-19 disease. Also, we will keep in mind that another outcome was safety, comparing local and systemic side effects between the BA.1-containing vaccines.

Let us go over to the results. Of course, the most critical results are the immunogenicity data. For the monovalent vaccine, they measured neutralizing antibody titers against BA.1, and the geometric mean ratio of those numbers was about 1.68. In the Methods section, they stated definitions for noninferiority and for superiority. For noninferiority, it was a ratio greater than 0.67, and if it was greater than 1, it was superior.

Against BA.1, the ratio for the monovalent booster was 1.68, so this met the superiority criterion. Against the Wuhan strain, it met the criterion for noninferiority, although it was numerically lower. And when they looked at the bivalent vaccine, the ratio was 1.53 against the BA.1 strain, again meeting superiority. When it came to the Wuhan strain, it was noninferior, but in this case, the titers were similar, right? You did not see a lower response when it came to the Wuhan strain, so we are seeing that the bivalent booster seems at least to show more breadth in regard to strain matching.

Another result I thought was really fascinating was that on day 29, they examined — roughly speaking — seroconversion rates, which they described as seroresponse rate (SRR). They looked for an increase from below the detectable limit of antibodies to a 4-fold increase in titers —essentially, the seroconversion response. And here you start to see quite significant differences, where at day 29, against Omicron BA.1, about 83% of those who received the monovalent booster met that definition of a 4-fold rise in titer, vs only 56% who received the original mRNA-1273. Those differences roughly held also for those who received the bivalent booster compared with those who received the original strain. Putting this together, what do you each think about these immunogenicity data?

FMR: I would say that if they had done only a primary analysis of noninferiority and not done the secondary analysis for superiority, that would have been unsatisfactory. Because at least in principle, I would not be expecting the original vaccine to have significant activity against Omicron. After all, that is why they wanted to study these boosters. I think it is a good thing that they were very intentional about their secondary analysis — showing superiority — in terms of neutralization.

OO: Dr Okoli, what would you say is the clinical significance of a superior neutralizing response — or do we even know?

AO: I think it is hard to say. What it signifies is that the response is superior. If I had to choose between 1 or the other, I would choose the 1 that was superior. Yet they are both noninferior; that is, they would each be able to achieve the same or a closely similar result, so I feel the clinical importance of superiority is unclear.

FMR: I might frame it differently. You could say that 2 things are not inferior, and that 1 is not inferior to the other, but that does not mean that each 1 is necessarily good. You could have 2 “bad” therapeutics that are not inferior to each other, but neither is conferring protection. So in this sense, I think it is good to have the superiority analysis because it tells me that not only is it not worse, but it is actually better.

OO: Exactly, yes. Yet again, let us recognize that these antibody levels are a surrogate endpoint and that the study was not powered or designed to draw firm conclusions about efficacy.

As for safety results, serious events were almost 0, which is good for the public to know. They were reported to be comparable between monovalent and bivalent BA.1 booters. But if you look at Figure 2 on grading of local and systemic side effects and adverse reactions, the original mRNA-1273 vaccine had at least numerically higher rates of reactions, sometimes both in frequency and severity, than the bivalent BA.1 booster.

Let us talk about limitations and other concerns. Dr Okoli has raised the issue of participant demographics. Children were not included in this study, so the results do not apply to them. We talked about racial minorities, and this matters because we know that, for example, Black people have higher rates of HIV,¹⁰ which can affect immune responses. Immunocompromised patients and patients with obesity or another chronic medical condition, such as chronic kidney disease, were not represented. Again, those conditions alter immune responses. I think we can agree that this is a limitation of this study.

AO: In the US, we have a Black population of 13.6%,¹¹ compared with the UK, which has a Black population of about 4%.¹² But even that was not quite represented in this group. Yet, in the COVID-19 pandemic, we have seen that Black people have had higher rates of both infection and complications as well as higher severity.

OO: Dr Machiavello Roman, did you note other limitations or concerns?

FMR: I would like to echo the authors’ self-criticism, which was that they were essentially aiming at a moving target. If they were recruiting in February and publishing in November, then updates were emerging about new mutations within that period, and they were very honest about saying they would have wanted to include more participants and gain statistical power.

I know it is very, very difficult to do a study like this. But it is concerning when I think about the pace of real-world events, such as when we had our Omicron surges here in the US. I do not think a study of this kind would have given us readily available data for clinical practice in that situation.

OO: I agree. Following up in that context, I took away the message that both monovalent and bivalent booster forms are effective with the newer circulating strains. In other words, a very specific vaccine could still be relevant in the context of subsequent variants because of the relatedness to the strain the vaccine was developed to combat.

Now the argument for the bivalent booster is that you are keeping some breadth of protection beyond just BA.1. But that is breadth looking backward. If we are talking about Alpha and Beta and Delta — we may never see those again, and the Wuhan strain was gone a long time ago. So from a clinical standpoint, are you impressed by the breadth? Do we really need breadth? Or are you satisfied with a variant-specific booster that neutralizes what is currently circulating?

Or let us put it this way — moving into this fall, we are going to have a monovalent vaccine, right? That could have been informed by some of these considerations. Questioning whether we really need that “breadth,” or instead whether it is enough to only hit the contemporary circling variants hard, I think, is an open question. But this tempers my enthusiasm a little with regard to the bivalent vaccine.

AO: So if we were targeting the currently dominant circulating strain, which is XBB.1.5, then how quickly are we able to adapt if the dominant variant is different 6 months from now, and will we be able to just focus on targeting that specific subvariant? It is kind of hard to tell.

OO: I agree with you: those are some of the real-world issues. But then again, I think there is still value to what we have available in the moment. For example, during many of these studies, researchers are able to bank sera from some of the participants that they collected on day 29. This allows us to do ex vivo studies looking at neutralization of new variants, including those that did not even exist at the time of the previous study.

Is there 1 important message you take from these data? What is the clinical relevance for you as we go into what will be already a busy COVID-19 season? Will these data change or influence the conversations you have with your patients?

AO: I feel these data do help to show, like Dr Machiavello Roman was saying, that in terms of superiority, if I had to choose between the updated boosters and previous vaccines, certainly I would want the updates to be available. They are more able to target the strains of concern that are circulating at this time, especially for the kind of patients who have comorbidities and require that protection.

FMR: In my clinic, I work with patients who are awaiting a transplant. When I talk about COVID-19 vaccines with these patients, a number of whom are immunocompromised, 1 question that comes up often is safety. You can hear in the media that a vaccine is safe, but a randomized trial like this gives firmer evidence I can describe to those patients. I can feel comfortable telling them, “This is a safe vaccine and we should move forward with this.”

At the same time, it is important for me to keep those patients protected because if they get COVID-19 when their organs come in, they are going to be dropped from the list, so I am always communicating to them that we need to keep them updated with protection against the circulating strain. Now I have evidence that this circulating strain will be at least neutralized by the antibodies that they produce with this vaccine. Hopefully, that will help “sell” them on getting the vaccine and being better prepared for their transplant.

OO: I agree with both of you. To summarize and offer my reflections: in this study, participants were receiving a fourth or fifth dose of COVID-19 vaccine. At the same time, we have seen in our part of the world that uptake of COVID-19 boosters has been very low: less than 20% of US population received the updated booster, and there have been concerns moving forward about the appetite of the public to receive the shots in the future. I think data like these are critically important because they are demonstrating the level of safety. We can tell people that receiving more and more doses of COVID-19 vaccine does not appear to be associated with any significant harm, severe adverse events, or other concerning outcomes.

And not only are we hearing that people can tolerate more antigenic stimulation from the vaccines, but this work also further defines the benefits of boosters. It is saying that 4 is better than 3, and 5 is better than 4. In the face of waning immunity, boosters can at least do exactly what they are called: they can boost against the decrease in circulating antibodies. And we have seen that this correlates with clinical protection.

The last strong message I get here is that there seems to be an advantage to strain matching with these vaccines. We clearly see much more robust antibody responses with vaccines that match the circulating strains, so that should at least be a part of our approach to updating vaccines moving forward.

I want to thank both of you for a robust discussion today. I hope that it is stimulating for our readers and that some of our critical comments prepare others to better appraise the study that we have presented.

This article was edited for clarity and length.

References

1. FDA takes key action in fight against COVID-19 by issuing emergency use authorization for first COVID-19 vaccine. News release. US Food and Drug Administration. December 11, 2020. Accessed September 28, 2023. https://www.fda.gov/news-events/press-announcements/fda-takes-key-action-fight-against-covid-19-issuing-emergency-use-authorization-first-covid-19

2. Solante R, Alvarez-Moreno C, Burhan E, et al. Expert review of global real-world data on COVID-19 vaccine booster effectiveness and safety during the Omicron-dominant phase of the pandemic. Expert Rev Vaccines. 2023;22(1):1-16. doi:10.1080/14760584.2023.2143347

3. Medicines and Healthcare Products Regulatory Agency (MHRA) authorizes Moderna’s Omicron-containing bivalent booster in the UK. News release. Moderna, Inc. August 15, 2022. Accessed September 28, 2023. https://investors.modernatx.com/news/news-details/2022/Medicines-and-Healthcare-Products-Regulatory-Agency-MHRA-Authorizes-Modernas-Omicron-Containing-Bivalent-Booster-in-the-UK/default.aspx

4. Chalkias S, Harper C, Vrbicky K, et al. A bivalent Omicron-containing booster vaccine against Covid-19. N Engl J Med. 2022;387(14):1279-1291. doi:10.1056/NEJMoa2208343

5. Lee IT, Cosgrove CA, Moore P, et al. Omicron BA.1-containing mRNA-1273 boosters compared with the original COVID-19 vaccine in the UK: a randomised, observer-blind, active-controlled trial. Lancet Infect Dis. 2023;23(9):1007-1019. doi:10.1016/S1473-3099(23)00295-5

6. Zimmermann P, Curtis N. Factors that influence the immune response to vaccination. Clin Microbiol Rev. 2019;32(2):e00084-18. doi:10.1128/cmr.00084-18

7. Lessler J, Riley S, Read JM, et al. Evidence for antigenic seniority in influenza A (H3N2) antibody responses in southern China. PLOS Pathog. 2012;8(7):e1002802. doi:10.1371/journal.ppat.1002802

8. Scheaffer SM, Lee D, Whitener B, et al. Bivalent SARS-CoV-2 mRNA vaccines increase breadth of neutralization and protect against the BA.5 Omicron variant in mice. Nat Med. 2023;29(1):247-257. doi:10.1038/s41591-022-02092-8

9. Khoury DS, Cromer D, Reynaldi A, et al. Neutralizing antibody levels are highly predictive of immune protection from symptomatic SARS-CoV-2 infection. Nat Med. 2021;27(7):1205-1211. doi:10.1038/s41591-021-01377-8

10. US Department of Health and Human Services Office of Minority Health. HIV/AIDS and African Americans. Updated December 13, 2022. Accessed September 29, 2023. https://minorityhealth.hhs.gov/hivaids-and-african-americans

11. US Census Bureau. QuickFacts: United States. Accessed September 27, 2023. https://www.census.gov/quickfacts/fact/table/US/PST045222

12. United Kingdom Government Digital Service. Population of England and Wales. Updated April 4, 2023. Accessed September 27, 2023. https://www.ethnicity-facts-figures.service.gov.uk/uk-population-by-ethnicity/national-and-regional-populations/population-of-england-and-wales/latest

Posted by Haymarket’s Clinical Content Hub. The editorial staff of MPR had no role in this content’s preparation.

Reviewed October 2023