EXECUTIVE SUMMARY
Perhaps nothing has consumed our practices more than the effect of COVID-19 and the disputes with patients over vaccination. Despite the miracle of modern science that multiple vaccines have been developed in an amazingly short time after the onset of the pandemic, so many Americans have been hesitant or actively resistant to vaccination. Never before has the authority and scientific integrity of such prestigious and respected organizations as the World Health Organization, U.S. Food and Drug Administration, and Centers for Disease Control and Prevention been so critically questioned. The conjunction of the politicizing of the science and the powerful influence of social media has powerfully propelled potential misinformation.
- Concepts of the totality of the data and the quality of the data are discussed.
- Multiple COVID-19 myths are deconstructed.
- The challenges of the effectiveness of natural immunity vs. immunization are discussed. The Delta variant has evolved since the original development of the vaccines and, along with waning immunity, has continued to lead to re-infections.
- The recent onset of the new Omicron variant currently is under investigation to assess its relative infectiousness, severity, and responsiveness to current vaccines and therapeutics.
Introduction
The COVID-19 pandemic has killed more than 800,000 American citizens.1 The global pandemic has altered not only the medical profession, but also life in general. The virus has invaded every aspect of our lives, including our politics.
Everyone desires a path out of this pandemic, but how is this to be achieved? One possibility is to achieve herd immunity by using the new COVID-19 vaccines. The degree of vaccinations and post-infection cases needed to achieve herd immunity is a moving target because of several factors that are difficult to model and often changing.2 These same factors call into question whether achieving herd immunity should remain a long-term public health goal. Regardless of one’s stance on vaccines, they will not obviate the important role of masking, hand hygiene, and social distancing for this and other communicable diseases.
Currently, there are four major vaccines in use around the world: Pfizer/BioNTech (also known as BNT162b2 or Comirnaty), Moderna (also known as mRNA-1273 and Spikevax), Johnson & Johnson (Janssen COVID-19 vaccine, or J&J), and AstraZeneca (AZ or Vaxzevria). This review will focus on the two mRNA vaccines, Pfizer/BNT, which now has received full U.S. Food and Drug Administration (FDA) approval, and Moderna, which still is approved by the FDA under an emergency use authorization (EUA). We also will discuss, to a limited degree, the J&J and AZ vaccines.
Initially, vaccination rates were high, but lately there has been a significant deceleration.3 It seems we have reached a point at which almost anyone who desired a vaccine has received it.
At the time of this writing, the United States population is 57.2% fully vaccinated, with 68.7% of adults 18 years of age or older vaccinated. But a large portion of the population is deciding against the vaccine. Many factors have been suggested to help explain this phenomenon. One of the common terms used now is “vaccine hesitancy,” which describes a large group of people who, for various reasons, are unsure if they want to get vaccinated.4 In a study of 1,878 people, the vaccine-hesitant group made up 22% of the population. This group should be clearly separated from the group that is anti-vaccination.
There have been several different attempts to entice this vaccine-hesitant group and increase vaccination rates, including offering monetary or other benefits for getting vaccinated. Some companies (including healthcare systems) have moved to mandating vaccination as a term of employment. Some do not think mandates are the most effective way to motivate vaccine-hesitant persons to make the choice, but instead think laying out a logical approach to vaccination that addresses the most common concerns and misinformation is the best path forward.5 This review will address these concerns and misinformation.
The Current Psychological and Intellectual Milieu
Before discussing the data pertaining to the vaccines themselves, we should look at several background issues that significantly affect how information is received and judged by an individual.
We should realize we are living in a social media/internet, postmodern age. The flood of easily shareable information has led the director-general of the World Health Organization (WHO) to label this as an “infodemic,” or misinformation epidemic.6 In one study, researchers examined more than 112 million social media posts and found that 40% had unreliable sources and 42% were generated by bots.7 Another study showed that 50.5% of the fake stories about COVID-19 were spread via social media, totaling more than 4.5 billion views in a single month.8,9
Another issue is the classic Dunning-Kruger effect, which can be summarized in this way: The most incompetent people also are the ones who most often overestimate their own abilities, while the most competent people typically underestimate their own abilities.10 This leads to a “dual burden” in that those who were the most incompetent not only performed poorly but also lacked the self-awareness to realize this fact. However, one important finding of this study was that incompetent people could be trained to improve but first had to realize their deficit. This underscores the problem of health literacy. In America, about one in every three adults has low health literacy.11 In the United Kingdom, four in 10 adults have difficulty understanding health information designed for the general public and six in 10 have difficulties with health-related numbers and statistics.12,13 Educating against misinformation could be a useful strategy in the vaccine-hesitant group, especially if they are being unduly influenced by poor data.
Furthermore, there is the Hawthorne effect: The fact that you are being observed can affect your behavior.14 The importance of this effect applies both for people who have been vaccinated and those who are against vaccination, since both groups are keenly aware of how they will and can be viewed. Social media has increased the effect of this phenomenon. Many people have posted pictures of themselves getting vaccinated, while many others have stated clearly that they do not want to get vaccinated. This goes back to concepts, such as virtue signaling and tribalism, which are beyond the scope of this review but are important nonetheless to understand the current mindset. Educating this group will be less effective.
Finally, there is the “elephant and the rider” theory proposed by Jonathan Haidt in his book The Righteous Mind: Why Good People Are Divided by Politics and Religion.15 His hypothesis, based on extensive psychosocial research, uses the elephant and the rider analogy. The elephant is massive, makes us move down the path, and represents our biases, instant reactions, habits, and comfort level. The rider is small, rides the elephant, but has very little ability to control it and decide on the path to take, and represents our use of logic. Haidt argues that the little logic we do use mostly is for post-hoc rationalization of what was our initial reaction. This is why people will continue to believe in theories that have been discredited and will argue to protect their initial stance on topics regardless of how much new data are presented to them. Clearly, this affects the effort to correct misinformation.
The Vaccine Data
Two concepts should be discussed before reviewing the state of the current medical literature regarding vaccinations.
The first is the idea of the totality of the data. Many prefer the phrase “preponderance of the evidence.” Both of these concepts support that the best estimate of effect is closer to the summed total of all the evidence. We should not put all our logical weight behind a single study that is in opposition to the majority of the published data. Many will believe in (or not believe in) an idea based on a few data points, thus ignoring the vast volume of data to the contrary. It is best to put your trust not on the fringes or extremes, but where the most data fall.
Another point that needs to be addressed is the quality of the data. A well-accepted standard exists to grade the quality of a study and has been adopted by the American College of Cardiology and the American Heart Association.16 This is shown in Table 1.
Throughout the rest of this article, we will note consistently the size and quality of the different data sets.
Table 1. Quality of Evidence from Various Types of Published Papers | ||
Level of Evidence |
Descriptions |
Type of Data |
A |
|
Randomized |
B-R |
|
Randomized |
B-NR |
|
Nonrandomized |
C-LD |
|
Limited data |
C-EO |
|
Expert opinion |
Deconstructing the COVID-19 Vaccine Myths
Myth 1: “The vaccine was developed using aborted fetuses.”
A common point of contention about the utility of the vaccines comes not from a scientific argument, but from a religious one. Many people are reluctant to accept a vaccine that has been developed using HEK 293 cells because of their personal beliefs on abortion. The original HEK cell line was derived either from a spontaneous miscarriage or an aborted fetus in 1973 and has been immortalized for cell biology research for many years.17,18
It is important to remember that the mRNA vaccines (Pfizer/BNT and Moderna) were developed and produced differently from typical vaccines, such as the J&J and AZ vaccines. The mRNA vaccines were not developed in HEK cells, thus eliminating the real issue for some people regarding the use of aborted fetal cells in vaccine production.19 Having said that, the Catholic Church already has released a statement that they do not feel that taking any of the vaccines, regardless of manufacturing process, is a theological issue.20
Myth 2: “Many healthcare workers are against getting the vaccine.”
Confidence in the vaccine has been eroded by a few vocal doctors speaking out against it. This is a small minority that, unfortunately, has received a lot of “hits” through social media. When looking at the actual reported numbers of doctors receiving the vaccine, it is the vast majority.
In two studies, well more than 90% of doctors have received the vaccine. In a survey study of about 300 doctors by the American Medical Association (AMA), 96% of doctors already had been vaccinated (level of evidence C-LD).21
In another large study of 3,975 healthcare workers looking at infection rates in multiple hospitals and published in the New England Journal of Medicine, 94% of the doctors and 84% of the nurses already were vaccinated (level of evidence B-NR).22 Finally, in the first legal challenge to hospital vaccine mandates, 117 workers sued the Houston Methodist hospital for wrongful termination.23 Out of ~26,000 total hospital employees, this represents only ~0.43% of the workforce that refused the vaccine. The judge dismissed the case.
Most healthcare workers, especially doctors, are in favor of personally getting the vaccine. The vocal minority does not represent the larger medical community as a whole (level of evidence C-LD).
Myth 3: “Doctors are making a lot of money from the vaccine.”
Another myth is that doctors and scientists are making money from the COVID-19 vaccine. At the National Institutes of Health (NIH), unlike Congress, researchers cannot invest more than $15,000 with any individual company and researchers cannot own stock in a business that works with NIH.24 The limits actually were stricter, but Congress relaxed the monetary limits to try to retain talented researchers at the NIH.
Myth 4: “The vaccines have killed a lot of people.”
“Is the vaccine safe?” This probably is the area of greatest concern for both medical providers and patients. In medicine, we swear an oath to “do no harm.” Since it is impossible to predict each person’s individual reaction to any treatment, we try to minimize the risk/benefit ratio in each person to limit their exposure to adverse outcomes (risk) and maximize the potential benefits. The easiest and simplest way to examine this ratio is to focus on the risks, since benefits are much harder to predict. The major driver of concern about vaccine safety and misinformation is related to misunderstanding the role and function of the Vaccine Adverse Event Reporting System (VAERS) database.
There are multiple difficulties with understanding the VAERS database, which is why we will spend considerable space discussing it in detail. Although it is a large data set, its overall quality is low. Before we discuss the numerical data commonly cited and taken out of context, we should discuss why the overall quality of the data is low.
The first issue is how the data are collected. The data are entered by anyone who wants to enter an event and it is voluntary. These issues reduce the reliability and quality of the data significantly. Furthermore, since it is time-consuming and cumbersome, adverse events typically are viewed as grossly underreported.25 In this study, rashes associated with the measles, mumps, and rubella (MMR) vaccine were estimated to be less than 1%, but more adverse events were reported more frequently, which is not consistent with what typically is stated by anti-vaccine proponents. In a more recent evaluation of the VAERS database, the sensitivity to detect adverse events ranged from 13% to 76%, depending on the vaccine (seven were evaluated) and the type of potential adverse event, with more serious possible complications being reported more frequently.26
Then there is the issue of causation vs. association. Events that are related in a temporal fashion do not imply causation automatically. In medicine, we use prospective randomized controlled trials to try to establish causation as much as possible. All other studies only show an association that may point to causation, which is why lower-quality studies should be only hypothesis-generating.
According to the VAERS site, “A report to VAERS generally does not prove that the identified vaccine(s) caused the adverse event described. It only confirms that the reported event occurred sometime after vaccine was given. No proof that the event was caused by the vaccine is required in order for VAERS to accept the report. VAERS accepts all reports without judging whether the event was caused by the vaccine.”27 This caveat also is noted in the footnotes: “VAERS accepts reports of adverse events and reactions that occur following vaccination. Healthcare providers, vaccine manufacturers, and the public can submit reports to VAERS. While very important in monitoring vaccine safety, VAERS reports alone cannot be used to determine if a vaccine caused or contributed to an adverse event or illness. The reports may contain information that is incomplete, inaccurate, coincidental, or unverifiable. Most reports to VAERS are voluntary, which means they are subject to biases. This creates specific limitations on how the data can be used scientifically. Data from VAERS reports should always be interpreted with these limitations in mind.” Finally, these disclaimers are noted in the footnotes when searching the VAERS database: “Reports may include incomplete, inaccurate, coincidental, and unverified information” and, “The number of reports alone cannot be interpreted or used to reach conclusions about the existence, severity, frequency, or rates of problems associated with vaccines.”28
In prospective randomized controlled trials, not only is there a data safety monitoring board (DSMB), but there also is an adjudication committee. This is a group that reviews and confirms all endpoints. They look at the patient’s chart to determine if the event (death, heart attack, or side effect) should or should not be attributed to the intervention. VAERS does not allow that. The bias, or logic error, we have just discussed at length is called “attribution error.” Are we attributing the cause, or blame, to the appropriate (or true) reason? Larger studies and higher qualities of evidence give greater confidence in making the correct assessment.
The absolute number of adverse events reported to VAERS is concerning. This is why it has drawn so much attention from the general public and from people who are hesitant about receiving the vaccine. Although the absolute numbers are high, to better understand it we need to look at the relative numbers. This will allow us to better understand the proportional risks involved. Since VAERS does not have a formal control group, we will compare event rates to those in the general public and also to those in the COVID-19 population.
First, the numbers — the absolute numbers from which all others are derived. Absolute numbers are important, since they show the scope of the problem. Here are the absolute death rates that we will use for reference and as a background:
- There are 328,239,523 people in the United States.29
- There are 2,854,838 deaths per year (7,821 deaths per day) in the United States.30
- Each year, 655,041 persons die from heart disease.31
- Influenza and pneumonia account for 49,783 deaths per year.30
- Sudden cardiac death (SCD) accounts for more than half of all cardiac deaths, totaling 356,000 per year (975 per day).32
Here are the COVID-19 absolute numbers, as of Sept. 9, 2021:
- There have been 649,299 deaths as a result of COVID-19 in the United States since the start of the pandemic.33
- In the United States, 177,104,652 people, which represents 53.3% of the population, have been fully vaccinated.34
- There were 14,704 possible vaccine-related deaths reported by the end of August 2021.
- There were 2,315 people with “low platelets” and 509 people with “idiopathic thrombocytopenia” (ITP) for a total of 2,824 persons with abnormal platelets.
- There were 7,862 with people with “blood clots,” 2,834 people with “deep vein thrombosis” (DVT), and 849 people with “pulmonary emboli” for a total of 11,545 people with some type of blood clots.
- There were 8,536 people with “stroke.”
- There were 2,766 people with “myocarditis.”
- There were 3,876 people with “heart attack” and 2,156 people with “myocardial infarction,” for a total of 6,032.
- There were 1,402 “miscarriages.”
For reference, about 17,000 children go to the emergency room as a result of injuries related to televisions, but no one is demanding a halt to watching television to protect our children.35
The fact that identical diagnoses are listed under separate terms (such as “heart attack,” a layperson’s term, and “myocardial infarction”) point to the problems with the accuracy of the VAERS data set.
Now for the relative numbers: Ratios, rates, and percentages help us compare large and hard-to-understand numbers and, thus, their relative importance. This is important, since there is no control group in VAERS and so no other way to gauge the potential effect of the vaccines.
Deaths caused by COVID-19 are 44.2 times higher than deaths possibly linked to the vaccine. The effect of COVID-19 far exceeds the effect of potential vaccine side effects. But more than 14,000 potentially vaccine-related deaths is alarming and concerning.
Calculating the risks helps to put these numbers into perspective.
- Even if all these deaths were directly related to the vaccine, the death rate from the vaccine would be 0.008%.
- The risk of low platelets would be 0.0016%.
- The risk of any blood clot would be 0.0065%.
- The risk of stroke would be 0.0048%.
- The risk of myocarditis would be 0.002%.
- The risk of myocardial infarction would be 0.003%.
- The risk of miscarriage could not be calculated from the VAERS database.36
So, although the absolute numbers are troubling, the relative rates of these occurrences are extremely rare. Another way to put these numbers into perspective is to compare them to the de novo rate that these occur in the general population and also to compare how often they occur with COVID-19 infections.
Table 2 shows the relative rates of these adverse events. What can be seen is that, even if the VAERS data were taken to be completely accurate, the event rates are similar to the rates in the general public. Moreover, the event rates pale in comparison to those in people who actually contract COVID-19, again demonstrating the protective effect of the vaccine.
Table 2. Event Rates Comparisons and Level of Evidence | |||
Crude Annual Rate in the General Population (Level of Evidence: B-NR) |
Possible Vaccine-Related Event Rates by VAERS (Level of Evidence: C-LD) |
Event Rates with COVID-19 Infection (Level of Evidence: B-NR and C-LD) | |
Myocarditis37,38 |
0.02% |
0.002% |
0.31% to 2.3% |
Low platelets39 |
0.0015% to 0.0039% |
0.0016% |
Rare, case reports only |
Nonfatal stroke40,41 |
0.24% |
0.0048% |
1.4% |
DVT/pulmonary embolus/all blood clots42,43 |
0.12% to 0.27% |
0.0065% |
4.4% to 46% |
Nonfatal heart attack/myocardial infarction40,44 |
0.32% |
0.003% |
20% |
Death45 |
0.87% |
0.008% |
1.6% |
Miscarriage46,47 |
11% to 22% |
NA |
0.5% |
VAERS: Vaccine Adverse Event Reporting System; DVT: deep vein thrombosis | |||
A more formal analysis was done by looking at 6.2 million persons from the VAERS database during the first six months of the vaccination program.48 There was no signal of increased harm across multiple clinical complications, including Guillain-Barré, central cerebral venous sinus thrombosis, myocarditis, ITP, Bell’s palsy, stroke, or myocardial infarction. At this point, the concerns raised about the vaccines based on the VAERS data set are unwarranted. The level of evidence for this study is C-LD.
At present, the rate, or proportion, of these events happening in vaccinated people is similar or lower than in the general population, and clearly much lower than in those who incur an actual COVID-19 infection.
The direct risks of the vaccines are low and the potential benefit in avoiding COVID-19 infection and subsequent clinical sequelae are high.
As mentioned earlier, VAERS itself warns that side effects that occur temporally related to vaccination do not prove causation. To illustrate this point, we are going to do a gedankenexperiment — German for “thought experiment” — popularized by Albert Einstein.
Cardiovascular (CV) disease is the most common cause of death in the United States.31 The largest proportion of CV deaths is sudden cardiac death.32 SCD often is a clinical diagnosis, since autopsies are rare. It is the typical stated cause of death when someone is found dead after being seen in a stable state of health shortly before. Many deaths are assigned to SCD regardless of knowing the actual cause of death because of the low number of autopsies. There are many other possibilities that could lead to sudden death: massive pulmonary embolus, massive internal bleeding, ruptured aortic aneurysm, acute intracranial hemorrhage, sepsis, or choking. No matter what the actual etiology for death was, it typically is listed as SCD. Since there are 356,000 deaths attributed to SCD each year, COVID-19 vaccine-related deaths could fit into that number 24.2 times. So, all possible COVID-19 vaccine-related deaths could easily be explained within the context of SCD, but that statistically is very unlikely.
Consider a case study on John Doe. Doe has known hypertension, hyperlipidemia, diabetes type II, and obesity, but no documented CV disease. He had slept badly; ate butter on his toast, with bacon as a side; had a fight with his wife before going to work; was in a traffic jam and was late to the office; ate from a vending machine at lunch; and walked up several flights of stairs to get to a meeting. Then, while presenting at the meeting, he dropped dead. Most will think he had a massive myocardial infarction or had a small myocardial infarction with ventricular fibrillation. It is unclear what the final etiology of his death was, but it most likely will be labeled (appropriately so) as SCD.
Repeat the same scenario, but this time factor in that Doe got the COVID-19 vaccine two days ago. What caused his death? We still do not know for sure, but many will suspect the vaccine.
Suppose we repeat the two scenarios discussed earlier, but in both cases an autopsy was performed. The autopsy showed a large, fresh, acute myocardial infarction. In the first case, without the vaccine, we would say that he had a poor lifestyle, had all the risk factors for CV disease, was acutely stressed, and probably had blood pressure elevation. His blood was in a hypercoagulable state that led to his acute infarct.
Repeat the second example with Doe getting the vaccine; what would you say? Would it be the same explanation noted earlier or should we attribute it to the vaccine?
A very large-scale, real-world study of 1,736,832 persons in Israel demonstrated that the risk profile of the Pfizer/BNT vaccine was similar to that in unvaccinated persons and was considerably more favorable and safer than in those with COVID-19 infection.49 Compared to the unvaccinated group, vaccination led to less acute kidney injury and intracranial hemorrhage, but higher rates of herpes infection and myocarditis compared to unvaccinated and uninfected persons. However, vaccination offered significant protection for acute kidney injury, arrhythmias, DVT, intracranial hemorrhage, myocardial infarction, myocarditis, pericarditis, and pulmonary embolism when compared against unvaccinated and infected persons (level of evidence B-NR).
Overall, the mRNA vaccines are very safe in the clinical trials, in real-world experience so far, and even within the often-mentioned VAERS database.
Myth 5: “The vaccine will make me sterile.”
A major concern, especially for the younger population, is the effect of COVID-19 and the vaccines on reproductive function.
It appears that COVID-19 rarely is vertically transmitted, with a rate of 3.91%.50 The rate of fetal death was 1.28%. These rates came from a systematic review of the literature and only included 230 pregnant women. Newborns were tested for SARS-CoV-2 by throat swabs, and all positive cases were delivered by cesarean. Even though confirmatory testing was done, the overall level of evidence is C-LD.
In another systematic review, 324 pregnancies were studied.47 In this review, the vertical transmission rate was 1.94%. There were four miscarriages (two were a set of twins), for a crude miscarriage rate of 1.23%. Also, a subset of mothers’ maternal milk (22 patients) and vaginal secretions (six patients) were tested for SARS-CoV-2 and all were found to be negative. Similar to the other study, the level of evidence is C-LD.
At this time, the COVID-19 vaccines appear to be safe in pregnancy. In an analysis of the v-safe and VAERS systems, there was no significant difference in maternal or perinatal outcomes in vaccinated patients compared to pre-pandemic reported incidence rates.36 The study included 35,691 patients, but because of all the limitations already discussed regarding the VAERS database, the level of evidence is C-LD.
In a larger, real-world study of eight large healthcare systems, more than 105,446 pregnancies were investigated.51 The vaccination rate was low, at 6% to 7.8%, depending on which mRNA vaccine was administered. Of those who had been vaccinated, there was no difference in spontaneous abortion rates within 28 days of vaccination compared to non-vaccinated pregnancies (level of evidence B-NR).
Ovarian function appears not to be affected by either COVID-19 infection itself or the Pfzier/BNT vaccine.52 This was a small, case control study of 32 patients undergoing in vitro fertilization (IVF) therapy. Both post-COVID-19 and vaccinated patients had anti-COVID immunoglobulin G (IgG) antibodies in the follicular fluid. The researchers did not check for spike protein in the follicular fluid. The level of evidence for this study is C-LD.
Currently, there is only a single study on male reproduction post-vaccination.53 In a small prospective study of 45 men, sperm counts increased after mRNA vaccination. The authors stated that, although the numbers were statistically higher post-vaccination, that they all were still in the normal range. Finally, there was no spike protein found in the semen post-vaccination. The level of evidence for this study is B-NR.
At this time, the vaccines appear not to affect reproductive function or miscarriages in clinical experience, the VAERS database, and in small, mechanistic studies of both male and female reproductive function.
Myth 6: “The vaccines don’t protect you.”
The Pfizer/BNT vaccine was evaluated in a large, prospective, randomized, double-blinded trial of 43,548 adults. The efficacy was 95%. There were only two deaths in the treatment arm and four in the placebo arm. The death rate for the vaccine arm was 0.009% and the death rate in the placebo arm was 0.018%, which was not a significant difference.54 These death rates are what one would expect for the general population. The adjudication committee did not attribute any of the deaths to the vaccine or the placebo. The level of evidence for this study is A. This vaccine now has received full FDA approval.
The Moderna vaccine underwent a similar evaluation in a separate large, prospective, randomized, double-blinded trial of 30,420 adults.55 The efficacy was 94.1%. There were two deaths in the vaccine arm (one from suicide) and three deaths in the placebo arm. As in the Pfizer/BNT study, these death rates were similar to what would be expected in the general public and were not thought to be related to the vaccine or placebo. The level of evidence for this study is A.
The vaccines appear to reduce transmission of the virus as well. Although it may seem obvious, since the major role of a vaccination program is to reduce societal spread, transmission of COVID-19 post-vaccine has not been well studied because the vaccines only are a recent development. To date, the only study to address this issue looked at healthcare workers.56 In this Scottish study of 194,362 persons in 92,470 households, 144,525 were healthcare workers. The rate of seroconversion dropped from 9.4/100 person-years before vaccination to 2.98/100 person-years after full vaccination. The hazard ratio for infection was 0.46 for fully vaccinated healthcare workers. The effect most likely is larger, since the authors pointed out that not all infections were transmitted by the healthcare worker to the family but could have arisen from other social contacts. The level of evidence for this study is B-NR.
The vaccines are effective in high-risk, real-world situations. In a large, prospective, cohort study of 3,974 healthcare workers (one of the highest risk groups based on age and exposure), the vaccines were 91% effective in preventing infections. Vaccinated people who had breakthrough infections did derive some protection; they had 40% lower viral load. Also, the risk of febrile illness was reduced by 58%, and their length of illness was 2.3 days shorter. Finally, there were no deaths (level of evidence B-NR).22
Also, in this study of all the healthcare workers who seroconverted, 76.5% were unvaccinated. We are seeing a similar picture across the country now, in which COVID-19 now is grossly a disease of the unvaccinated (level of evidence B-NR).
There are other supportive pieces of evidence for the effectiveness of the vaccines. Both the case and death curves have fallen as vaccinations have risen.57 This type of data strictly is correlative and does not prove causation (level of evidence C-LD).
Two statistical modeling studies have been published and look at the potential effect on lives and costs with vaccinations. In the first study, the authors estimated that vaccinations may have saved up to 140,000 lives in the United States as of May 2021 (level of evidence C-LD).58 In the second study by the Kaiser Family Foundation using an estimated cost of $20,000 per COVID-19 admission, they predicted that vaccinations could have prevented 280,000 hospitalizations between June and August 2021, which would total $5.7 billion in medical costs (level of evidence C-LD).59
Myth 7: “The Delta variant makes the vaccines useless.”
The Delta variant has led to a resurgence in COVID-19 infections and has made many people question the long-term durability of the vaccines.
The Mayo Clinic study has been circulated widely among the public and demonstrated a significant drop in vaccine efficacy in the Pfizer/BNT vaccine. The researchers looked at 25,589 matched patients from Minnesota. Although the Pfizer/BNT vaccine had significantly lower effectiveness in preventing infection against the Delta variant, it still protected the patients from hospitalization. It is important to note that this study was a pre-publication release and has not yet undergone peer review. But because of its very public reporting, we have decided to discuss it and several other pre-publication studies like it in this review (level of evidence B-NR). (See Table 3.)
Table 3. Mayo Clinic Experience with the Delta Variant60 | ||||
Vaccine |
Efficacy | |||
Infection with Alpha |
Hospitalization with Alpha |
Infection with Delta |
Hospitalization with Delta | |
Moderna |
86% |
91.6% |
76% |
81% |
Pfizer/BNT |
75% |
85% |
42% |
75% |
Another study that already has been circulated widely (and also has not undergone peer review) was done in Israel, which has the highest vaccination rate in the world at 78%. It showed a time-dependent decrease in effectiveness in the Pfizer/BNT vaccine, such that there was a 53% increase in breakthrough infections with the Delta variant in people who had the vaccine early in their country’s roll-out. The authors suggested that a booster may solve this issue, but that remains to be seen (level of evidence C-LD).61
Another point about Israel: Much has been made of the fact that most of the COVID-19 cases in Israel now are in the vaccinated group. There are two potential reasons for this: 1) waning vaccine efficacy, as postulated earlier; but 2) Since Israel has the highest rate of vaccinations in the world, with the vaccinated population outnumbering the unvaccinated by three to one, there is a much larger pool of vaccinated patients to draw from. To think otherwise is to not understand proportions and is another error in logic.
A factor in the Mayo Clinic and Israeli experience may be a difference between the Pfizer/BNT and Moderna vaccines. The Pfizer/BNT vaccine appears to produce a lower antibody response compared to the Moderna vaccine in persons older than 50 years of age.62 This could be related to the Moderna vaccine having 100 mcg of mRNA compared to 30 mcg in the Pfizer/BNT vaccine. This was a small, mechanistic study of 167 patients, but it may explain the profound drop-off seen in both of these studies, especially in Israel, which used the Pfizer/BNT vaccine (level of evidence C-LD).
This also has been replicated in another, larger study of 1,647 healthcare workers.63 Persons who received the Moderna vaccine had a greater antibody response than those who received the Pfizer/BNT vaccine when measured six to 10 weeks after full vaccination. The Moderna vaccine, in comparison, also elicited a larger antibody response in patients who previously were infected with COVID-19. There also was an age-dependent decline in antibody response to both vaccines (level of evidence B-NR).
The Centers for Disease Control and Prevention (CDC) estimates that all vaccines combined have dropped in effectiveness to prevent infection from 91% to 66% over time.64 This was a study of 4,135 participants of the HEROES-RECOVER cohort of healthcare workers. This change over time also corresponds to the increase in the Delta variant. The waning effectiveness of the vaccine could represent repeated exposures vs. a drop in antibody production (waning protection) vs. another explanation (level of evidence B-NR). Larger studies, discussed later, seem to indicate better long-term performance.
The British experience demonstrated that the Pfizer/BNT vaccine had only a slight decrease in effectiveness. It was 93.7% effective against Alpha and still 88% effective against the Delta variant.65 This study was a case control study that included 96,371 controls, 7,313 patients with the Alpha variant, and 4,043 patients with the Delta variant. The difference between their experience and that of the Mayo Clinic and Israel may be related to the much longer timeframe between doses in the United Kingdom and may be another piece of evidence supporting booster shots (level of evidence B-NR).
In observational studies, it appears immunity wanes over time. This may be from a combination of changes in masking requirements and an increase in Delta variant concentration, as well as decreased humoral response. In California, the University of California San Diego Health System achieved a full vaccination rate of 76% by March 2021.66 There was a significant increase in cases of COVID-19 in July 2021 among the workforce. When analyzing the attack rate in July, it was found that the attack rate for workers who completed vaccination in January or February 2021 was 6.7 per 1,000, compared to 3.7 per 1,000 for those who completed the vaccine in March through May. The attack rate for unvaccinated employees was 16.4 per 1,000 in July. The vaccine effectiveness had been approximately 95% at avoiding infection through June and, in July, it dropped to 65.5%. Symptoms were present in 83.8% of the vaccinated workers and in 88.9% of the unvaccinated workers. Of the 227 workers who tested positive, no deaths were reported, and there was only one hospitalization of an unvaccinated employee. Coincidentally, California’s mask mandate ended June 15 and B.1.617.2 (Delta) emerged in mid-April.
In California, Los Angeles County reported 43,127 COVID-19 cases from May to July 2021. Of the cases, 71.4% were in unvaccinated persons.67 During this timeframe, the Delta variant became the dominant strain and was responsible for more than 87% of all cases. Furthermore, vaccinated patients were more stable clinically. All comparisons were statistically significant (level of evidence B-NR). (See Table 4.)
Table 4. Los Angeles County Experience with Vaccinated and Unvaccinated Individuals | ||
Vaccinated |
Unvaccinated | |
Cases |
10,895 |
30,801 |
Percent of Total Cases |
25.3 |
71.4 |
Hospitalized |
3.2% |
7.6% |
Intensive Care Unit |
0.5% |
1.5% |
Mechanical Ventilation |
0.2% |
0.5% |
Percent Delta Variant |
91.2% |
87.1% |
A study of New York suggests that effectiveness to avoid infection dropped from 92% in May 2021 to 78% in July.68 However, the effectiveness to avoid hospitalization remained constant at 95% from May to July.
A wide-scale study of 13 jurisdictions in the United States demonstrated a drop in effectiveness of the vaccines after Delta became the predominate strain, but the vaccines still were highly effective in preventing hospitalizations and death compared to unvaccinated individuals (level of evidence B-NR).69 (See Table 5.)
Table 5. Sampling of the U.S. Experience with the Delta Variant | |||
Infection Rate Ratio |
Hospitalization Rate Ratio |
Death Rate Ratio | |
Pre-Delta |
11.1 |
13.3 |
16.6 |
Delta Variant Dominant |
4.6 |
10.4 |
11.3 |
In a nine-state study of 32,867 medical encounters during the surge of the Delta variant in June through August 2021, vaccination reduced infection by 83.6%.70 It also reduced hospitalizations by 86% and emergency department/urgent care visits by 82%. One significant finding in this analysis was that there was a statistically significant reduction in effectiveness in persons older than 75 years of age, 76% compared to 89%, but the overall effectiveness still was high (level of evidence B-NR).
The good news is that both mRNA vaccines protect people from infection, illness, hospitalization, and death. There does seem to be a waning of effectiveness that may be related to the Delta virus or to other factors. As more studies are published, the vaccines appear to perform much better than the first pre-publication reports of the Mayo Clinic and Israeli experiences.
There now have been multiple published peer-reviewed studies that show that the overall effectiveness of the vaccines has been affected as of late, but it is less clear if the reason is the Delta variant, waning antibody production, changes in masking and social distancing, or a combination of all these variables. However, it has been demonstrated consistently that vaccination still provides robust protection against severity of disease, hospitalization, and death, even if a person becomes infected.
Myth 8: “Natural immunity is durable, total, and better than vaccination.”
In a study looking at the presence of IgG antibodies in hospital workers in the Johns Hopkins Health System, there was durability beyond 250 days.71 Seventy-two percent of workers (eight out of 11) tested beyond 250 days still had IgG antibodies present. The rate of decay was estimated to be 7% per month (level of evidence C-LD). A study from Israel that is not yet published or peer-reviewed compared rates of protection against the Delta variant between Pfizer-vaccinated individuals vs. previously infected individuals.72 Researchers compared rates of infection between June 1, 2021, and Aug. 14, 2021, for people who had infection vs. vaccination prior to Feb. 28, 2021. When looking at more than 32,000 people, 257 were infected in the follow-up period. A total of 238 were vaccinated and 19 had prior infection. This represents a 13-fold increased likelihood of breakthrough infection vs. reinfection. There were eight hospitalizations in the vaccination group and one in the previous infection group. It also was confirmed that individuals infected previously who had received one dose of vaccine were reinfected less often (20 patients) than those who had not been vaccinated on top of previous infection (37 patients) in a group of 28,000 patients. Once again, this may point toward waning vaccine efficacy and the potential need for a booster (level of evidence B-NR).
In a small study of 72 persons who were infected with COVID-19, 36% did not have IgG or IgA responses. (level of evidence C-LD).73 More than one-third of infected individuals did not mount an antibody response. This group most likely would benefit from vaccination. An untested strategy could be to test for antibody production in a serial fashion and then vaccinate when or if the levels decrease significantly.
A recent study analyzing phylogenetic similarities across six types of human-infecting coronaviruses and antibody decay rates estimated that unvaccinated persons, under endemic conditions, likely would become reinfected at a median interval of 16 months (level of evidence C-LD).74
Antibody kinetics are not the only means to achieve long-lasting immunity. Cellular immunity that can provide renewed antibody production when challenged by the same antigen can be induced. This has been investigated specifically in COVID-19-naïve and -recovered patients with an mRNA vaccine vaccination. In this small study of 47 patients, COVID-19 vaccination led to strong CD4 T-cell response.75 An interesting finding of this study was that in COVID-19-recovered patients, there was a robust response after the first dose, but less after the second. In contract, the cell-mediated response was strongest after both doses in the COVID-19-naïve group (level of evidence C-LD).
Vaccine Immunity
At this time, it is unclear how long immunity will last. Published data regarding humoral response has been reported at eight-month follow-up for the J&J vaccine and at six-month follow-up for the Moderna vaccine. The Pfizer/BNT vaccine has a preprint article as well, commenting on overall effectiveness but not specific antibody titers. The J&J vaccine demonstrated a relatively stable median neutralizing antibody titer from the peak seen on day 71 and at day 239 post-vaccination.76 It reduced by a factor of 1.8 by day 239. However, this was assessing response to the original WA1/2020 strain. The antibody titer against B.1.617.2 (Delta) was 107 at day 239 compared to 184 against the original strain. Moderna published data 180 days after the second dose of vaccine.77 From the peak titer within two weeks post-vaccination, it dropped by a factor of nearly 10 at six months. The simulated virus neutralization assay showed titers had dropped by a factor of nearly 5 at six months (level of evidence C-LD).
A small study examined the time-dependent features of neutralizing antibody function post-mRNA vaccination.78 Twenty-seven persons had repeat blood tests out to three months. Median IgG concentrations dropped by 50% compared to the peak level (which had increased five-fold), but more concerning was that median inhibition against variants declined as well (range, 12.5% to 31.5%), but the Delta variant was not tested in this study. Furthermore, clinical effect was not measured in this study (level of evidence C-LD). As mentioned earlier, antibody testing may have a role in both vaccine and post-infection mediated immunity to detect patients who would benefit from boosters. Also, it should be noted that neutralizing antibody function is a mechanistic surrogate for clinical effectiveness.
Vaccination for previously infected persons may reduce the chance of getting infected again. In a small study in Kentucky, 246 cases were followed and compared to 492 controls. Not being vaccinated resulted in a 2.34-fold higher rate of infection than in those who were fully vaccinated post-COVID-19 (level of evidence B-NR).79
Booster doses are a moving target and, until very recently, were not recommended for the general population (level of evidence C-LD).80 This has changed quickly. There are certain groups that are immunocompromised, aged, and high-risk for COVID-19 that were approved for booster doses, but that was based on expert opinion.
Now there are several strong lines of research that support booster doses. Once again, the Israeli experience has helped guide recommendations. In a large study of 1,137,804 persons who were older than 60 years of age and fully vaccinated with the Pfizer/BNT vaccine, there was a significant protective effect of a third shot.81
Confirmed infections dropped by a factor of 11.3-fold and serious illness dropped by a larger 19.5-fold factor.81 The time from second to third shot was at least five months (level of evidence B-NR).
Another study (pre-publication) looked at heterologous booster dose combinations (nine in all) and all showed a strong antibody response.82 A total of 458 persons were studied after receiving the Pfizer/BNT, J&J, or Moderna booster at least 12 weeks after full vaccination with any of the three agents. An interesting finding was that homologous boosters increased antibody titers by 4.6- to 56-fold, but heterologous boosters increased titers 6.2- to 76-fold. This difference was not significant but demonstrates that either homologous or heterologous booster dosing can be done and that there probably is not a need to stay with the same vaccine for the whole series. This will have implications going forward if serial boosters will be needed in the future (level of evidence C-LD).
Summary
The COVID-19 pandemic has led to an infodemic of misinformation affecting the ability of the general public to make good decisions about vaccination. Vaccine hesitancy is a byproduct of this infodemic. After reviewing the current available data, the vaccines have an excellent risk/benefit ratio. The vaccines also are very effective, even in the midst of the Delta variant. Finally, sensational claims, such as infertility, and concerns raised by touting the raw VAERS numbers have not been borne out with further analysis.
References
A complete list of references is available at https://bit.ly/3yk7Kmm.
There have been several different attempts to increase COVID-19 vaccination rates, including offering monetary or other benefits for getting vaccinated. Some companies (including healthcare systems) have moved to mandating vaccination as a term of employment. Some do not think mandates are the most effective way to motivate vaccine-hesitant persons to make the choice, but instead think laying out a logical approach to vaccination that addresses the most common concerns and misinformation is the best path forward. This review addresses these concerns and misinformation.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content