Vaccinations Needed to Prevent COVID-19 ED Visits, Hospitalization

COVID-19 

According to the US Centers for Disease Control and Prevention (CDC), SARS-CoV-2 infections have damaged American public health. As of the most recent update of the COVID Data Tracker (August 7, 2023), the CDC estimates that there have been 6,233,900 COVID-associated hospitalizations. Beyond that, trends in hospitalization rates have indicated a significant increase, with a 12.5% in COVID-19-related hospital admissions between June 30, 2023, and July 29, 2023.  

Beyond that, data predicts that severe illnesses have contributed to 1,137,057 COVID deaths.  

The CDC estimates that, in the US, over 100 million doses of the updated Pfizer-BioNTech vaccine (BNT162B2) have been administered. Comparatively, roughly 48.5 million and 1.5 million doses of the Moderna (mRNA-1273) vaccine and Novavax vaccine have been distributed, respectively.  

Although mRNA vaccines to prevent COVID-19 cases and minimize severe disease incidence have been available since the end of 2020, only 81.4% of the US population have received their first dose of a COVID-vaccine. Approximately 69.5% of people in the US have completed their primary vaccine series, including the first and second doses. An even smaller portion has received an updated booster vaccine dose, 17.0%.  

This vaccine data implies that a significant portion of the population remains unvaccinated, perpetuating the spread of disease. Researchers have continued to study illness epidemiology to understand the number of vaccinations required to reach herd immunity. 

“Herd immunity is part of the concept of disease spread,” explained Shaun Grannis, MD, MS, Vice President for Data and Analytics at the Regenstrief Institute. “If enough people are immunized to the infection, then the disease can't spread.” 

He compared the number of vaccinated or immunized people as a barrier or wall that prevents further disease spread. However, there has to be a certain level of immunized individuals before herd immunity is effective. 

For example, if you vaccinate 30% of the population with vaccinated individuals randomly distributed, there is still a 70% chance that the disease will spread. 

“Generally, when the immunized population gets to a particular level, that spread stops,” he added. “Herd immunity is a general concept, and there are different theoretical levels for different conditions, but I think it's an important construct to help us think about why we vaccinate.” 

Study Protocol 

Grannis worked on a retrospective cohort study published in the Lancet Regional Health-Americas that analyzed the number needed to vaccinate with a booster dose to prevent emergency department visits and hospitalizations.  

“Understanding the number needed to vaccinate is important,” explained Grannis. “It gives us a target or a sense of the effort we need to put into vaccinating particular groups.” 

The study led by Grannis looked at multiple different groups of patients. One component focused on age groups, including younger patients and older adults. Older patients are at greater risk when they get COVID-19, as they tend to have more comorbidities. 

“It gives clinicians and researchers one way to measure the impact vaccines can have on individuals,” he added. 

“The number needed to vaccinate comes from the original concept of the number needed to treat,” explained Grannis. 

That number refers to how much of the population needs to be provided with the intervention to achieve a positive or intended outcome. Smaller numbers correlate with a more effective intervention. 

“The number needed to vaccinate helps researchers identify which subgroups we might want to focus on and get the educational messaging out to, making sure that people are well-informed,” he noted. “It helps us — when there are limited resources — to identify which groups or where we should focus resources to get the ‘biggest bang for the buck’ out of the vaccine.” 

Data Collection 

The study data collected information during the SARS-CoV-2 Omicron BA.1 predominance between December 2021 and February 2022. Information was extrapolated from the CDC VISION Vaccine Effectiveness Network, also called the VISION Network. VISION includes data from multiple sites, including Texas, New York, Minnesota, Wisconsin, Utah, California, Oregon, Washington, Indiana, Colorado, and Tennessee 

“We pool data from multiple sites to get the most comprehensive view of the country possible,” noted Grannis.“The data we put together is retrospective, so we are regularly capturing COVID data and aggregating that for analysis.” 

The data is updated about once a month, with information dating back to the pandemic’s beginning. The datasets include information on emergency department visits, hospitalizations, laboratory-confirmed COVID-positive and COVID-negative tests, and vaccination status — or lack thereof. However, they excluded identifying patient information to maintain confidentiality and conform to privacy policies. 

“As a medical informatician, I love to talk about how we can better use our data,” exclaimed Grannis. “Here in Indiana, we have one of the country's largest and longest-running health information exchanges called the Indiana Health Information Exchange. That integrates data from hundreds of different sources, standardizes it, harmonizes it, and makes it available for things like patient care and research. To expand these types of networks, health systems must have the capability to standardize and harmonize their data such that there's less effort required to convert it into a standardized format.” 

While healthcare systems and facilities may have technical hurdles to overcome when standardizing and harmonizing data, there are also governance concerns. 

“By governance, I mean the decisions people and organizations make about participating and collaborating with others,” explained Grannis. “Many organizations prioritize certain other things above research.” 

In particular, healthcare facilities and other organizations that manage electronic health records (EHRs) prioritize other factors, such as patient care, before research. While that’s vital for ensuring outcomes, research is just as important as it informs research, which will eventually impact patient care. 

“It does require human and technical resources to participate in these information exchanges, so individuals need to see research as a priority. As our electronic medical record and health information systems become more standardized, it'll be easier for more organizations to participate in this kind of data-sharing network to analyze information.” 

“The FDA is certainly very interested in using observational EHR data to support drug discovery and drug approval process,” he shared. One example is their Real-World Evidence Initiative. 

“Real-world evidence is needed not just for the drug approval process but also for learning about and answering many questions throughout the healthcare ecosystem,” said Grannis. “At the Regenstrief Institute, we've been using real-world evidence. That's largely how we’ve conducted our research for the last 50 years, using EHR data to support research and answer important questions.” 

Protocols 

Grannis explained, “The researchers and I did a retrospective cohort study looking backward at who was vaccinated and who wasn't. Essentially, we broke it down into two outcomes. Emergency department visits and hospitalizations were the endpoints or the outcome of interest.” 

For example, they looked at ED visits and generated risk or odds ratios to compute the number needed to vaccinate. 

“We computed the number needed to vaccinate for each scenario and could stratify the groups by age or other social demographic factors to understand if there were any significant differences between these folks.” 

Conclusions 

The study data included 1,285,032 patients comprising 938 hospitalizations and 2,076 ED visits. The study excluded pediatric and adolescent patients. Approximately 43.2% of patients were 18-49, 28.3% 50-64, and 28.5% 65 and older were included.  

Across the general population, the median number of people needed to vaccinate was roughly 205 individuals. 

Researchers also divided the data into three periods: 

• Period One: December 16, 2021, to January 9, 2022 
• Period Two: January 10, 2022, to February 3, 2022 
• Period Three: February 4, 2022, to February 28, 2022 

“In individuals 65 and over, the number needed to vaccinate is quite low,” noted Grannis. 

For example, in the first period, the number needed to vaccinate was 110 for patients 65 and older. Meanwhile, the numbers for ages 50-64 and 18-49 were significantly higher, at 351 and 1,249, respectively.  

In period two, the numbers needed to vaccinate patients 65 and older, 50–64, and 18–49 were 46, 144, and 506, respectively. Finally, the period three numbers needed to vaccinate were 88, 275, and 974.  

Because younger patients are less likely to be hospitalized by SARS-CoV-2 infection and have stronger immunity, the number needed to vaccinate was much higher. A significant number of people is required to prevent a rarer occurrence. 

“The trends or the risk factors that we saw associated with the number needed to vaccinate numbers were associated with the risk factors that we've been aware of,” said Grannis. 

For example, advancing age and multiple medical conditions or diagnoses, such as hypertension, diabetes, and more, are associated with a lower number needed to vaccinate, as people in these categories were more likely to have COVID-related ED visits and hospitalizations. 

One or more underlying conditions were associated with lower numbers needed to vaccinate, at 77, 33, and 62 individuals for each period. Comparatively, the number needed to vaccinate patients with a medical condition was much higher at 5,195; 2,089; and 4,048.  

Beyond the impacts of age and comorbidities, Grannis explained that the number needed to vaccinate to prevent an ED visit was more significant than that required to avoid a COVID-19 hospitalization. While it may seem counterintuitive that the number of vaccinated people necessary to prevent emergency department visits was more significant, as it is a less severe — and theoretically rarer — outcome of COVID-19 infection. 

“That is, for a variety of complicated reasons. Fundamentally, younger people often use the emergency department more frequently as primary care providers,” he revealed. “We’re seeing a number of them use the emergency department for various reasons, which skewed our numbers.” 

“A take-home message that I like to emphasize from this work is that vaccinating small populations at higher risk can prevent severe COVID-19 outcomes,” concluded Grannis. “This study puts a number on that and provides empirical evidence that even vaccinating a few thousand people will prevent several adverse outcomes. The cost and risks of vaccination are much lower than the potential adverse events from hospitalizations.”