Predicting COVID In The Future

Last week, a little-noticed milestone in the U.S. COVID pandemic occurred – it is the first week that the number of new cases of COVID exceeded the number of vaccine doses administered since the vaccines became available to the general public. It seems that America is ready to be done with COVID and move on. The good news is that both the number of cases and the number of deaths is falling. The bad news is that they are still very high – it’s just that we have developed a national desensitization to the COVID numbers. Last week, 226,618 Americans were diagnosed with COVID, that is enough people to fill Yankee Stadium five times over. In addition, 22,422 people were hospitalized with COVID and 2,290 people died of COVID. To put that in perspective, last week, 4 times more people were admitted with COVID and 10 times more people died of COVID than were admitted and died of H1N1 influenza during the peak week of the 2009 H1N1 influenza outbreak in the U.S. In other words, the best days of COVID are still worse than the worst days of H1N1 influenza. So, what will the COVID picture look like in the next year? We can learn a lot from the epidemiology of other viruses.

Summary Points:

  • COVID is likely to assume seasonal variation in the future, similar to other coronaviruses
  • The current mortality rate of 0.5 – 1.5% can be reduced by vaccination and new drug development
  • Hospitals should prepare for a modest increase in COVID hospital admissions and ICU utilization next winter
  • COVID is not going to go away anytime soon


What we can learn from non-COVID seasonal coronaviruses

Coronaviruses have been infecting humans for as long as there have been humans. There are dozens of different coronavirus species that each have a preferred animal it infects. For humans, there are four seasonal coronavirus strains cause cold symptoms: 229E, NL63, OC43, and HKU1. These coronaviruses are common and cause 15-30% of all common colds. These viruses recur predictably every year as reported by the Public Health Agency of Canada:

There is a striking seasonality to coronavirus infections with most infections occurring in the winter. The graph below shows the average number of coronavirus infections reported by the Public Health Agency of Canada over a 10-year period:

Future COVID seasonality

For the purpose of simplicity, I will use “COVID” (the name of the disease) as synonymous with “SARS-CoV2” (the name of the virus) in this post. At some point in the future, COVID will most likely go the way of other coronaviruses and assume the same seasonal variation with a baseline year-round rate and a rate surge in the winter. We have already seen signs that COVID has a predilection for the winter months in the case rates during the first 3 years of the pandemic. The graph below shows the number of COVID deaths per 100,000 per week in orange and the number of cases per week in blue.

This graph demonstrates that the death rate for COVID peaks about three weeks after the case rate peaks, consistent with the finding that most people who die from the infection do so about 3 weeks after initial diagnosis. There has been a peak in deaths every January (red dashed lines)  followed by a second, smaller peak in deaths every summer (black dashed lines).

Assuming that COVID becomes a seasonal infection, how long will it take to become primarily seasonal? Any answer to this question is speculative but it will likely be several more years before the seasonal epidemiology of COVID resembles that of other coronavirus infections. Until then, it is likely that there will be a moderate or low baseline level of COVID year-round with case numbers increasing in the winter. The main determinant to becoming seasonal is population immunity.

By now, most Americans have either been vaccinated against COVID or have had a COVID infection or both. The result is that most Americans have some degree of immunity. But what we know from the usual seasonal coronaviruses is that immunity fades and it is common for people to get reinfected with the same coronavirus strain. One study found that up to 21% of people get reinfected with the same strain of non-COVID coronavirus within 6 months of the initial infection. Going forward, having immunity from a previous COVID infection or a previous COVID vaccine will not entirely protect a person from getting a future COVID infection. But immunity can reduce the severity of infection, reduce death rates, and reduce transmission. For the population as a whole, compared to people vaccinated with a bivalent booster, unvaccinated people are 3 times more likely to be diagnosed with COVID, 16 times more likely to be hospitalized with COVID, and 10 times more likely to die from COVID. These numbers likely underestimate the protectiveness of vaccines since people most vulnerable to COVID have been the most likely to get the bivalent booster. As an example, people age 65-79 who are are unvaccinated are 14 times more likely to die of COVID than people age 65-79 who are vaccinated with a bivalent booster.

Unfortunately, we seem to have developed a national aversion (or at last indifference) to vaccination. Currently, 92% of adults have received at least one dose of a COVID vaccine but only 79% received a full primary series. Worse, only 20% of adults have received a bivalent booster. The sooner we can overcome our culture of vaccine hesitancy, the sooner we can overcome the consistent high number of non-seasonal COVID cases.

Future COVID mortality rates

Over the past 3 years, one out of every 300 Americans have died of COVID. It is difficult to know the exact mortality rate of COVID infections because we do not know exactly how many Americans have been infected with COVID. Using case numbers reported to the CDC, the average mortality rate since the beginning of the pandemic is 1.63%. This likely overestimates the true mortality rate because many people who get infected either do not get tested at all or do home tests that are not reported to the CDC.  The mortality rate has varied considerably over the past 3 years. In the graph below, the case numbers reported to the CDC are in blue and the mortality rate of infection is in red.

It is quite striking that when the case numbers are high, the mortality rate is low and vice versa. On possible explanation for this curious finding is that when new, more infectious COVID variants emerge, many non-vulnerable people (children and young adults) get infected with these variants at school and in workplaces causing a surge in case numbers. Because these people are often younger or have some degree of immunity from previous infection, they are less likely to die. Over the following several weeks, they then infect vulnerable people who are more likely to die: the elderly, the nursing home residents, and those with chronic diseases. This possible explanation is purely conjecture, however.

Another way of estimating the mortality rate of COVID in the U.S. is to use data from the Commercial Laboratory Seroprevalence Survey. This survey estimated the number of Americans who have had COVID based on COVID antibody tests performed on left-over blood from commercial lab tests. Notably, the antibodies tested for would be produced by COVID infection but not by COVID vaccination. An advantage of using data from this survey is that it picks up those people who either did not get tested for COVID because they had mild or asymptomatic infections and those who did home COVID tests that were not reported to the CDC. As of February 2022, 57.7% of samples contained antibodies against COVID. If we assume that this is reflective of the U.S. population as a whole, then as of February 2022, 57.7% of Americans had been infected by COVID – that translates to 192,306,920 people. At that time, the total number of people reported to have died of COVID was 939,875. Using these numbers, the mortality rate of COVID infection calculates to be 0.5%

From these analyses, it appears that the COVID mortality rate has most likely been somewhere between 0.5% and 1.5% over the course of the pandemic. In the future, the COVID mortality rate will hopefully be lower as more Americans have immunity from repeated infections and from booster vaccinations. However, it is a near certainty that some number of people will continue to die of COVID infections in future years. Decades of experience with influenza has shown us that neither natural immunity (from past infections) nor vaccination immunity prevents all deaths from influenza. There will always be deaths in vulnerable populations such as the elderly, the immunocompromised, the obese, and the diabetic. In these individuals, immunity can reduce but not eliminate the chance of dying from infection.

COVID is a new infection for the human race. A useful lesson from history about new infections is from the European settlement of North and South America in the 15th and 16th centuries. The indigenous peoples of the Americas had never been exposed to infections such as smallpox, a disease that is highly contagious but preferentially kills adults. When the first Europeans arrived, they brought with them these diseases that then rapidly spread throughout the continents. It is estimated that within a few decades of Columbus’s first landing, about 90% of indigenous people had died of infections such as smallpox. After burning through native populations, the smallpox mortality among these populations settled into a lower baseline number. If COVID behaves like smallpox, then it is likely that once COVID burns through the world’s human population that it will settle into a lower steady state mortality rate.

Future COVID hospitalizations

Early in the pandemic, U.S. hospitals were overrun by COVID patients. With no effective treatments, many patients died rapidly and survivors often required prolonged ICU care, lingering in the hospital for weeks. With better treatments and better population immunity, more patients are surviving their COVID hospitalization and they are improving faster, resulting in shorter hospital stays. However, COVID is still resulting in a relatively large number of hospital admissions. The graph below shows new COVID hospitalizations in orange and COVID deaths in blue as reported by the CDC. During the January 2021 surge, 1 person died for every 3 COVID hospital admissions. That ratio has improved so currently, 1 person dies for every 10 COVID hospital admissions.

During the January 2021 COVID surge, the CDC reported that 19% of all U.S adult hospital beds were occupied by COVID patients and 31% of all U.S. adult ICU beds were occupied by COVID patients. A year later, during the January 2022 surge, COVID patients accounted for 23% of adult hospital beds and 31% of adult ICU beds. These two surges put an enormous strain on our country’s hospital resources, particularly our intensive care units. Last week, 3.4% of both adult inpatient and ICU beds as well as 1.5% of both pediatric inpatient and ICU beds are occupied by COVID patients.

Although it is unlikely that we will see the overwhelming spikes in COVID hospitalizations such as we saw in January 2021 and January 2022, it is likely that we will continue to see seasonal fluctuations in hospital utilization as COVID assumes a more seasonal pattern. Because of this, hospitals should start planning now to ensure sufficient hospital beds and staffing for an anticipated spike in COVID admissions next winter. During the most recent COVID surge in January 2023, COVID patients occupied 6.5% of both adult hospital beds and adult ICU beds. Children with COVID occupied 2.4% of pediatric hospital beds and 2.3% of pediatric ICU beds. To be conservative, hospitals should plan on a similar increase in hospital bed and ICU demand next winter.

Future COVID treatments

Experience with other human infections has shown us that science makes incremental advances in treatment resulting in incremental improvement in mortality rates. Examples include tuberculosis, hepatitis C, and HIV. In each of these infections, the earliest treatments were marginally effective but as pharmacologic research advanced, subsequent treatments were better and better. The result is that now, most people infected with these pathogens can either be cured or kept in indefinite remission with current medications.

Over the past three years, we have also seen steady improvement in COVID treatments ranging from the ineffective (azithromycin) to the ludicrous (ivermectin) to the somewhat effective (Molnupiravir) to the highly effective (Paxlovid). If advances in COVID treatment is anything like hepatitis C, HIV, and TB, then we will likely have even better COVID treatments in the future.

Wild cards

Evolution shows that all living things mutate as they reproduce. COVID has been no exception with new variants emerging that are more infectious than the previous variants. The greater the total number of viruses present on earth at any given time, the greater the likelihood of a new variant developing. As worldwide natural and vaccine immunity increases, it is likely that the rate of new variant emergence will slow. This should give vaccine producers more time to create vaccines effective against those new variants, thus improving our ability to stay one step ahead of COVID. Nevertheless, that other variants will arise in the future is a certainty.

One of the reasons that we cannot eliminate COVID from the planet is that it can infect other animals. So far, COVID has been demonstrated to infect more than 30 different kinds of animals. The disease is not as severe or life-threatening as it is in humans but now other animals can serve as viral reservoirs. Even if we could eliminate all human infections today, humans would just get reinfected from deer, pigs, and dogs tomorrow.

Not only will the human race face new COVID variants but we will also likely see new coronaviruses make the jump from other species to ours. This has already happened recently with the coronaviruses that cause MERS (camels) and SARS (bats). There are many, many different coronavirus species with each species affecting different animals. Thus there are different coronavirus species that have been found in cats, dogs, pigs, camels, bats, cows, and chickens. Mutations in any of these coronavirus species can allow them to become infectious to other animals, including humans. The new mRNA vaccine technology now gives us the ability to rapidly develop and distribute vaccines against new coronavirus species that do cross from animals to humans. One challenge is that new vaccines against new viruses require clinical trials to determine vaccine effectiveness and safety. These trials take time and require a large number of subjects. Ideally, we need ways to rapidly predict efficacy and safety without the months required to perform clinical trials.

A future with COVID

It seems clear by now that COVID is not going to go away in the future. It is unrealistic to think that COVID case numbers will steadily go down until COVID drops off the face of the Earth. It will more likely just become one of the many respiratory viruses that humans regularly get infected with. It is likely that there will be a year-round baseline rate and seasonal rate surges. However, we have the ability to control COVID case numbers and case severity by optimizing immunity and by continued research into new medications.

March 6, 2023

By James Allen, MD

I am a Professor Emeritus of Internal Medicine at the Ohio State University and former Medical Director of Ohio State University East Hospital