Category: Epidemiology

A study recently published in the journal The Lancet Child and Adolescent Health provides strong support for vaccinating pregnant women against RSV in order to prevent their infants from being hospitalized with RSV infections in the first 5 months of life. The U.S. Department of Health and Human Services (which oversees the CDC) has sent mixed messages about maternal RSV vaccination this year and as a result, many pregnant American women (and their doctors) are confused about whether or not to get vaccinated.

Respiratory syncytial virus (RSV) is a common respiratory virus that generally occurs in the winter. Nearly all children in the U.S. are infected by age 2-years-old and re-infections are very common. In children older than 4-years-old and adults younger than 65-years-old, RSV infections are relatively mild and frequently similar to a common cold. But in newborn infants and the elderly, RSV infections can be deadly. In young children, RSV causes bronchiolitis, a narrowing of the airways that can make it difficult to breathe. The treatment for most infants with bronchiolitis caused by RSV is primarily supportive, so testing for RSV as the specific viral cause of infant bronchiolitis is frequently not done since it will not affect treatment. Therefore, it is difficult to know exactly how many infants are hospitalized because of RSV. Instead, many epidemiologists use the hospital diagnosis of bronchiolitis as a surrogate for a diagnosis of RSV infection when studying infant hospital admissions (and deaths).

The CDC reported that between October 1, 2024 and May 3, 2025, an estimated 10,00 – 23,000 Americans died from RSV and an estimated 190,000 – 350,000 Americans were hospitalized with RSV (all ages combined). Compared to all other age groups, infants have the highest hospitalization rate with one out of every 100 infants in the U.S. hospitalized with RSV during the winter of 2024 – 2025 (light blue line in the graph below from the CDC’s RSV Net Interactive Dashboard). And last winter was a mild RSV season compared to previous years. For example, during the winter of 2022 – 2023, one out of every 49 infants in the U.S. was hospitalized with RSV (dark blue line in the graph below).

These are astoundingly high rates of infant hospitalizations. In contrast, only one out of every 5,300 American children between ages 5 – 18 years-old were hospitalized with RSV last winter. And only one out of every 10,000 adults between ages 18 – 50 years-old were hospitalized with RSV last winter.

The travesty is that most infant hospitalizations are avoidable by vaccinating the mothers during pregnancy. Two years ago, an RSV vaccine was approved by the FDA to be used in pregnant women. Its use is the likely reason that the 2024 – 2025 infant RSV season was so mild compared to previous years (as shown in the graph above). Currently, only the RSV vaccine made by Pfizer under the brand name Abrysvo is approved for use during pregnancy. Two other RSV vaccines are approved only for use in the elderly (Moderna’s mRESVIA and Pfizer’s Arexvy).

The recent study in The Lancet Child and Adolescent Health looked at 391 infants hospitalized with RSV in England and Scotland. All infants were between 0 – 5 months old. These infants were compared to control infants who were not hospitalized with RSV. The authors found that effectiveness of RSV vaccination at any time during pregnancy was 58% but if the RSV vaccination was given more than 2 weeks before delivery, the vaccination was 72% effective in preventing subsequent hospitalization of the infants due to an RSV infection. These results are similar to the study published in the New England Journal of Medicine in 2023 that led to the FDA approval of the Pfizer RSV vaccine for pregnant women.

What if a pregnant woman has already received an RSV vaccine in the past?

Currently, the CDC recommends that women only get vaccinated against RSV once. If a vaccinated woman then has a second pregnancy, then she is not supposed to be re-vaccinated. Instead, it is recommended that her infant receives a monoclonal antibody against RSV (nirsevimab) shortly after birth. This is different than recommendations for vaccinations against other respiratory viruses like COVID and influenza that are recommended to be repeated on a regular basis. However, it has been shown that immunity after an RSV vaccination wanes with time, so there is currently a phase III clinical trial looking at the safety and efficacy of re-vaccinating pregnant women against RSV during subsequent pregnancies. The results of this study will likely not be available for several years.

The CDC has sent confusing messages about RSV vaccination

Robert F. Kennedy, Jr. was appointed as Secretary of the Department of Health and Human Services on February 13, 2025. The immediate result was a retraction of many of the CDC websites containing information about vaccines. Notably, the CDC website “RSV Vaccine Guidance for Pregnant Women” was taken down only to have a court order the website to be restored on February 14, 2025. The top of the website now carries this disclaimer:

It is unclear why this disclaimer exists as there is no reference to gender ideology other than the mention of “pregnant women” on this webpage. But anyone reading the last sentence: “This page does not reflect biological reality and therefore the Administration and this Department rejects it.” would justifiably believe that the Trump administration rejects vaccinating pregnant women against RSV. Since his appointment, RFK, Jr. has fired all of the members of the CDC’s Advisory Committee on Immunization Practices (ACIP) and replaced them with notable vaccine skeptics. Thus far, the ACIP has not changed the recommendations for vaccinating pregnant women against RSV but has dropped the recommendation for vaccinating pregnant women against another respiratory virus, COVID.

The CDC’s current vaccination recommendations for pregnant women are complicated and confusing to many pregnant women. Currently, the CDC recommends vaccination for women who are between 32 weeks pregnant and 36 weeks 6 days pregnant. However, vaccinations should only be given September through January in most of the continental U.S. But if the pregnant women lives in Alaska, southern Florida, Guam, Hawaii, Puerto Rico, U.S.-affiliated Pacific Islands, or the U.S. Virgin Islands, then pregnant women should be vaccinated against RSV year-round. If the mother did not get an RSV vaccination during that particular pregnancy, then the infant should receive an RSV antibody (nirsevimab) injection but only during the months of October – March. For the average pregnant woman reading these recommendations, they are mind-bogglingly bewildering.

All of these factors together have led to public confusion about vaccinating pregnant women against RSV and have led to mistrust in vaccines in general.

So, what is a doctor to do?

More than ever before, it is incumbent on OB/GYN physicians, nurse midwives, family physicians, pediatricians, NPs, and PAs to provide guidance to their pregnant patients and and to mothers of newborn infants. Fortunately, death from RSV in infants is relatively uncommon because of advances in hospital treatment of infants with bronchiolitis. However, hospitalization because of RSV is all too common in American infants. Hospitalization is frightening for both the child and the parents. It necessitates emergently arranging childcare for other children and arranging someone to take care of the family dog. It means unexpected time off work for the parents. Every year, infant hospitalizations due to RSV cost the U.S. $710 million – that works out to $187 per every pregnancy. Even though most infants do not die from their RSV infection, it still extract a great emotional and financial cost.

The disclaimer on the CDC’s website about RSV vaccination and pregnancy is very disturbing because it implies that the Trump administration does not support vaccinating pregnant women against RSV. Moreover, the rise of anti-vaxxers into positions of government authority has resulted in greater skepticism about vaccines by the general public and has resulted in declining vaccination rates across the country. Educating our patients about the importance of RSV vaccination during pregnancy will take valuable time during office visits. But this will be time well spent by physicians, advance practice providers, and pharmacists.

July 23, 2025

First, let me say that COVID vaccines work… they really, really do. But we are vaccinating our patients in the wrong months of the year. Our goal with current COVID vaccinations is not to prevent everyone from getting infected, it is to prevent as many people as possible from dying from a COVID infection. COVID vaccination results in partial immunity against the virus – enough that when you inhale COVID virus into your lungs, your body can get to work making antibodies against that virus faster than if you had never been vaccinated.

I am a COVID unicorn. Unlike the vast majority of Americans, I have never had a COVID infection (that I know of). This is despite spending the first year of the pandemic taking care of COVID patients in our ICU, intubating them, and performing bronchoscopies on them. I was super attentive to wearing an N-95 mask that I had been fit-tested for and to washing my hands constantly – both in the hospital and out of the hospital. I was first in line when the COVID vaccines were released and I have been revaccinated twice a year ever since. Every time I have had a cough or runny nose, I have done a COVID test (or 2, or 3) and all were negative. I was part of a CDC study looking at the incidence of COVID in healthcare workers and all of my serial blood tests looking for antibodies that would indicate past infection were negative. If I have had an infection since that CDC study finished, then it was mild enough that I never had any symptoms. I attribute the reason for not getting a COVID infection in the last 1,273 days since it first arrived in the U.S. to regular vaccinations (and maybe a dose of good genes).

The U.S. Centers for Disease Control and Prevention (CDC) recommends that adults between the ages of 18 and 64 get a new COVID vaccination each year and those over age 65 get a new COVID vaccination every 6 months. But there is no recommendation about what time of the year to get those vaccinations. Most physicians vaccinate against COVID at the same time they give influenza vaccinations – in September or October. Seniors then get a second vaccination 6 months later in March and April. I would argue that we are doing this all wrong and we should be vaccinating everyone in June and seniors again in November. Here is why.

When are you most likely to get a COVID infection?

In my last post, I discussed how COVID is weird compared to other respiratory viruses. Rather than peaking once a year in the winter, it peaks twice a year – once in about August and once in about December. This can be seen in the following graph showing COVID deaths since the beginning of the pandemic with an annual summer surge (red arrows) and a second winter surge in December or January.

Last year, there were more COVID hospitalizations, deaths, and ER visits during the summer surge than during the winter surge. Whether this trend in the summer surge being worse than the winter surge continues in future years is uncertain but it does seem clear that in the foreseeable future, there will continue to be two surges each year. Ideally, you would like to time your vaccinations to coincide with the regular surges of infections.

When does immunity peak after vaccination?

To oversimplify the immunology of vaccinations, a vaccine works three ways. First, vaccines cause your B lymphocytes to make antibodies that then circulate in the bloodstream and help destroy viruses that get into your body. Second, vaccines cause your body to create memory lymphocytes that can make new antibodies when a virus gets in your body and stimulates those memory cells. The ability of antibodies to destroy viruses is called the humoral immunity. Third, vaccines cause your T lymphocytes to be primed to produce gamma interferon and other cytokines that can help fight invading viruses; this is called cellular immunity.

The result is that when a person who has been vaccinated gets a viral infection, that person has some circulating antibodies left over from vaccination to kill the virus but that person can also ramp up new antibody production much faster than an unvaccinated person because of their memory cells. That vaccinated person also has T lymphocytes that already know to produce gamma interferon and other cytokines the moment a virus gets into the body.

Antibodies immediately produced after a vaccination do not last very long in the blood stream with levels peaking about 3 weeks after the vaccine is administered. Cellular immunity peaks a little later than humoral immunity after a vaccine. The result is that COVID vaccines are highly effective in preventing infection 1 month after vaccination but that effectiveness drops somewhat by 5 months after vaccination. Based on this data, a person should ideally get a COVID vaccination 1-2 months before being exposed to COVID.

And the best month of the year to get a COVID vaccination is…

For people over age 65 and those otherwise at higher risk of COVID, the best times to get biannual COVID vaccinations are late June and mid November. This will put them in the sweet spot of optimal humoral and cellular immunity when the summer and winter COVID surges occur. Ideally, this would prevent any COVID infection but realistically, it will primarily prevent an COVID infection severe enough to require hospitalization or cause death. At the least, physicians should get their patients over 65 vaccinated at the earliest signs of COVID surges – when graphs of wastewater surveillance, percent COVID test positivity, and percentage of ER visits attributed to COVID curves first start to rise in the summer and winter (this year, these curves started to turn upward in late June).

For people under age 65, it is a bit more complicated. Should they get their vaccination in late June in time for the summer COVID surge or should they get their vaccination in mid November in time for the winter COVID surge? Based on last year’s data, there were more COVID hospitalizations during the summer surge than during the winter surge so getting a vaccine in June would seem prudent. But since we don’t know if this year will mirror last year, my advice is to either get the vaccination in late June or mid November.

But some caveats about COVID vaccinations:

1. If you have a COVID infection in May, you don’t need a COVID vaccination in June. Just like vaccinations, infections create both humoral and cellular immunity that peaks several weeks after the infection. So, if a person has had a recent COVID infection, they do not need a COVID booster vaccination – their immune system has already gone to work to provide protection for the next few months.
2. More is better. Booster vaccinations are like exercise for your body’s immune system. The more you exercise, the stronger you get. Your immune system will be far better at defending against a COVID infection if you get regular booster vaccinations either once or twice a year.
3. COVID vaccines are not 100% effective. A tetanus shot is really, really good at preventing a tetanus infection for 10 years. COVID vaccines, not so much. Vaccinations impart different degrees of immunity for different viruses. Patients need to be educated that the vaccination against COVID won’t prevent all infections but will reduce the severity of any infection that they get. To the person who says: “I got a COVID vaccination but I still got COVID” I believe that what they should be saying is: “I got a COVID infection but because I got a COVID vaccination, I lived to tell about it.”
4. You can still get a COVID infection any time of the year. August and December are when the peaks of COVID surges occur but there is a low-grade incidence of COVID infection all year round. Getting you COVID booster in June may not help you as much if your kid brings home a COVID infection from kindergarten in May. But timing vaccination boosters is all about playing the odds and the odds are better that you’ll get COVID in August than in May.
5. A COVID infection will cost you a lot, even if it doesn’t kill you. When your employees get COVID infections, they’re going to be off work for several days and if you are going to lose money on them. If child gets a COVID infection, that child is going to be out of school and their education is going to suffer. If that child is in school sports, he or she is going to lose days or weeks of training and end up being less competitive. It is no fun to have a symptomatic infection – it is like you just wasted 5 days of your life. And if you get long COVID, you are going to pay the price of poor health for months.
6. Anytime of the year is better than no time of the year. It is not practical to vaccinate every American on June 30th and November 15th each year. Some people only bother to get vaccinated when they come in for their annual check-up with their primary care provider. Other people get their regular vaccinations when they think about it or when they hear that their family and friends are all getting vaccinated. Immunity after a COVID booster may wane after 5 months but there will still be some immunity up to 12 months, albeit less. So, it is better to get a COVID booster at the wrong time of the year than to not get one at all.
7. Masks still work. Since January, I’ve been on 23 separate airline flights (many were connecting flights). In January and February, I was still wearing an N-95 mask in airports and on planes. Since March, I’ve been flying maskless. But if I fly somewhere in August, I’ll likely wear a mask again. Vaccinations are good at fighting viruses that get into your body but masks prevent the viruses from getting inside of you in the first place.
8. RFK, Jr. is stupid. I rarely resort to using derogatory terms but when I do, it is reserved for only the most extreme people. As Secretary of the Department of Health and Human Services, Robert F. Kennedy, Jr. has unilaterally changed vaccination schedule guidelines to fit his own unfounded beliefs that ignore science. Specifically, he instructed the CDC to no longer recommend COVID vaccinations for pregnant women and for children. This is a really bad idea for several reasons. First, there is ample evidence that pregnant women who get a COVID infection are more likely to have bad pregnancy outcomes and to die whereas there is zero evidence that a fetus exposed to a COVID vaccination has any kind of bad outcome. Second, children are just as likely to get COVID infections as adults, it is just that they are less likely to die from a COVID infection than adults (but some children can and do still die). Third, when you vaccinate your child against COVID, you are vaccinating the entire family against COVID. The infection that your kid picks up at school is likely to then be transmitted to Grandpa who is then more likely to end up in the hospital – if you can prevent the kid from getting infected, you can prevent that kid from infecting the rest of the family. Household-acquired infections can cause serious illness and even death.

With more than 5 years of experience with COVID, we have learned a lot about it. As our understanding of it has improved, our recommendations to prevent infection have evolved. We now have enough experience with it to recommend to our patients the best time to get COVID boosters.

July 15, 2025

Compared to other common respiratory viruses, COVID is weird. Most respiratory viruses peak once a year in the winter. COVID peaks twice a year, once in the summer and once in the winter. The early indicators are signaling that this summer’s COVID season is just starting to ramp up. But how will it compare to previous years?

The strange epidemiology of COVID

Since the beginning of the pandemic, there has been an annual surge in COVID cases and deaths every summer in August or September, The graph above shows deaths due to COVID reported by the CDC. We do not see the same phenomenon with other respiratory viruses like influenza. The reasons for this unusual biannual surge in COVID is not entirely clear but may be due to a confluence of several factors including:

1. Waning immunity. Both immunization against COVID and infection by COVID will provide partial immunity against future infection. But neither is 100% protective and the immunity tends to wane over time. Similar to influenza vaccinations, the realistic goal of COVID vaccinations is not to prevent everyone from getting infected, it is to prevent as many people as possible from dying of an infection. COVID vaccination is extremely effective in reducing (but not completely eliminating) a person’s chance of dying from a COVID infection but not so great in preventing mild infections, especially after about 6 months following the vaccination.
2. COVID mutates easily. Every time the COVID virus mutates, it becomes more resistant to the antibodies in our bloodstream created by either vaccination against previous mutations or infection by previous mutations. One of the reasons that the COVID virus mutates so quickly is that there is just so many cases of COVID infection out there. This equates to more total number of viruses on Earth at any given moment. Statistically, the larger the number of viruses, the greater the opportunity for mutation. In addition, coronaviruses as a group mutate especially easily compared to many other types of viruses.
3. Air conditioning. Closing up the house and turning on the AC doesn’t cause COVID, but when it gets hot outside, people tend to go indoors to cool off. It is the close contact that people have with each other in air-conditioned buildings that allows the COVID virus to be easily transmitted from one person to another.
4. Summer travel. Travel also promotes close contact with other people in confined spaces with limited ventilation. Cars, airports, and airplanes bring lots of people in proximity, breathing each other’s air. Many people who would stay home from work or school if they are feeling a bit poorly are not willing to cancel their summer car trip or airplane trip to the beach thus increasing the chance that they will infect others.

What happened with COVID cases last year?

In the beginning of the pandemic, it was easier to track the number of cases of COVID. The only places you could get COVID tests were in hospitals or approved outpatient testing sites. All of these reported their results to local health departments that in turn reported to the CDC. Now, many people with mild respiratory infection symptoms do not seek testing. Others do home COVID tests themselves that are not reported to health departments. Consequently, we do not know how many people get new infections. Instead, we have to infer epidemiological trends using other metrics such as wastewater surveillance, the percentage of COVID tests reported to health departments that are actually positive for COVID, and the percentage of emergency department visits that are because of COVID infections.

Wastewater. The graph below shows the amount of COVID virus detected in wastewater monitoring locations over the past year. This shows a large peak in August and a smaller peak in January. The implications is that there were a lot of Americans getting COVID infections in the summer of 2024 – more than in the following winter.

Percentage of ER visits due to COVID. The CDC monitors the primary diagnoses for ER visits. When COVID numbers surge, the percentage of visits to emergency rooms on any given day because of COVID will increase. The graph below shows this data since the beginning of the pandemic. The surges in the percentage of ER visits diagnosed with COVID correlates with the number of hospitalizations as shown in the graph at the top of this post.

The graph below shows this data for the past year and demonstrates a larger number of COVID ER visits in the summer than in the winter. The solid red line depicts the data for all ages combined. Last summer, the highest peaks were seen in infants less than 1 year old (yellow fine dotted line) and the elderly older than 75 (purple line).

Percentage of COVID tests that are positive. Percent test positivity also correlates with surges in the number of COVID cases. Like ER diagnosis percentages and wastewater surveillance, test positivity provides an early warning sign of impending case surges – days or weeks before surges in hospitalizations and deaths. Once again, we see that the percent test positivity was greater last summer than it was last winter.

What about COVID hospitalizations last year?

The winter of 2021 – 2022 (green line below) saw the greatest rate of COVID hospitalization of the pandemic. Since that time, the number of hospitalizations have been falling, due primarily to the availability of vaccinations and population immunity from previous infection. The peak in hospitalizations in the summer of 2024 (light red line below) was slightly greater than the peak in hospitalizations the following winter (light blue line below).

There has also been a slow change in the demographic of COVID patients who get hospitalized. The graph below shows the ages of patients hospitalized since the start of the pandemic. Early on, a high percentage of hospitalized COVID patients were in the 18-49 age range (blue) and 50-64 age range (light green). Now, the vast majority of those hospitalized are over 65 (red). This change in demographic is also likely due to population immunity with most Americans having partial immunity from either vaccination or previous infection with the result that the most medically vulnerable (the elderly) now account for most hospitalizations.

How does COVID now compare to other respiratory infections?

The graph below shows hospitalizations caused by the common respiratory viruses last winter. The number of hospitalizations for COVID (red line) was similar to the number of hospitalizations for RSV (green line). There were many more hospitalizations for influenza (blue line) than for either COVID or RSV. Once again, population immunity from COVID vaccinations and previous COVID infections has resulted in COVID causing less severe infections than earlier in the pandemic. The implication is that influenza has regained its crown as the most epidemiologically significant respiratory infection. So, if your patient will only allow you to give them one vaccination this year, give them a flu shot.

The early signs indicate COVID cases are again increasing

Drilling down on the three early predictors of COVID case surges, there is evidence that the summer surge for 2025 is now beginning:

Wastewater. The graph below shows a slight upward trend in the amount of COVID virus found in wastewater over the past 45 days. The national trend is the black line in the middle, western U.S. is green, southern U.S. is purple, northeastern U.S. is blue, and midwestern U.S. is gold.

Percentage of ER visits due to COVID. The graph below shows COVID deaths (blue bars) and the percentage of ER visits diagnosed with COVID (gold line). The peak in ER visits occurs about 2-3 weeks before the peak in deaths. There has been an uptick in COVID ER visits over the past month.

COVID test percent positivity. The graph below again shows COVID deaths (blue bars). The gold line shows the percentage of COVID tests that were positive for COVID. The peak in test positivity precedes the peak in deaths by 2-3 weeks. Once again, we see a slight uptick in the curve (gold line) over the past month.

COVID has not gone away…

We all so want to be done with COVID. For most Americans, the pandemic with its resultant deaths, lockdowns, masks, and social distancing are a distant memory. Shortages of goods with resultant inflation has given way to economic prosperity and record highs in the stock markets. But COVID is still out there, just waiting for us to close the windows, turn on the air conditioning, and pack the kids into the car for summer vacation.

If last year is a predictor for this year, then in August we should see another surge in COVID hospitalizations, especially infants and the elderly. We will probably see at least as many COVID cases in August as we see later in the year in December. ER physicians, hospitalists, and primary care physicians should resume testing patients who present with symptoms of respiratory infections. We should all make the public aware that COVID cases are increasing so that infected persons seek medical care early, before those infections get severe enough to require hospitalization or cause death.

July 14, 2025

As an urban trauma center, our hospital’s ER sees its share of murder victims every year. Murder happens for all sorts of reasons but one thing all murders have in common is eventually, someone will politicize them. If a murder happens in a Republican stronghold, then Democrats will blame lax gun laws and far-right hate groups. If a murder happens in a Democrat stronghold, then Republicans will blame lax crime laws and undocumented immigrants or the homeless. The finger pointing is inevitable. But what is the truth? Does how a city or a state votes in political elections affect its homicide rate?

To answer this question, I looked at whether a state or major U.S. city voted Republican or Democrat in the 2024 presidential election. For city data, I used the result of how the county that the city lies in voted since county voting information is more readily available than city voting information. There are two Federal agencies that track U.S. homicides: the FBI and the CDC. The FBI’s database is derived from reports from local law enforcement agencies but this reporting is not mandatory so many homicides do not get reported to the FBI. The CDC’s database is derived from death certificates that are filled out by coroners. Because coroner reporting of homicides is mandatory, the CDC data is more complete and will show a higher homicide rate than the FBI for any given state or city.

States that voted Republican have a higher homicide rate than states that voted Democrat

Homicides are best reported by their rate per 100,000 population instead of their total number. New York City with its population of 8.8 million has more people than the entire states of Rhode Island, Montana, Delaware, South Dakota, North Dakota, Alaska, Vermont, and Wyoming combined. Therefore, it is to be expected that there would be far more homicides in New York City than in, for example, Montana, simply based on the sizes of the populations.

There is tremendous variation in homicide rates between states ranging from a low of 1.8 per 100,000 population in New Hampshire to a high of 20.7 per 100,000 in Mississippi. In other words, an average person is about 12 times more likely to be murdered in Mississippi than in New Hampshire. The map below shows 2022 homicide rates by state with the darker the shade, the higher the homicide rate.

This map shows that homicide rates are the highest in southern Mississippi Valley states plus New Mexico. Homicide rates are lowest in New England states, upper Midwest states, and Mountain West states. We. But are homicide rates higher in states that voted Republican in the 2024 presidential election or in states that voted Democrat? The graph below shows the homicide rate per 100,000 population in all 50 states. States that voted Republican are in red and states that voted Democrat are in blue.

The average homicide rate for states that voted Republican was 8.5 per 100,000 and the average homicide rate for states that voted Democrat was 5.6 per 100,000. This is a statistically significant difference with a P-value of 0.018 by unpaired T-test. In other words, a person is 50% more likely to to be murdered in a state that voted Republican than a state that voted Democrat.

Most U.S. cities voted Democrat

There is a striking difference between rural America and urban America when it comes to political preference. This was evident to anyone driving across the United States last October – inside city limits, yard signs were mostly for Democrat candidates and outside city limits, yard signs were overwhelmingly for Republican candidates. The Pew Research Center found that in 2023, 60% of urban counties vote Democrat and 60% of rural counties vote Republican; suburban counties are evenly split between Republican and Democrat. Researchers at the Carsey School of Public Policy found that this trend also held in the 2024 presidential election with 61% of large core metropolitan areas voting Democrat and only 26% of non-metropolitan rural areas voting Democrat. The graph below shows the 50 most populous cities in the U.S. in 2023 based on data from the U.S. Census Bureau.

Of these 50 cities, 38 voted Democrat (blue bars) and 12 voted Republican (red bars) in the 2024 presidential election. However, the total population of all 12 Republican cities combined (7,715,274) is lower than the population of the nation’s most populous city, New York City (8,804,199). Note that for all of these cities, their metropolitan area population is much larger than the city population.

There is not a relation between the size of cities and the homicide rate of cities. The graph below shows the homicide rate per 100,000 population for the same 50 largest U.S. cities by population in 2022 based on CDC data. New York City with the largest population has approximately the same homicide rate as Arlington, Texas which has the smallest population.

Notably, many of the nation’s largest cities have homicide rates below the nation average, including New York City, Los Angeles, San Diego, Austin, San Jose, and San Francisco.

Cities that voted Democrat have a higher homicide rate than cities that voted Republican (but it is not statistically significant)

There is no statistically significant association between a city’s political preference and its homicide rate, although there is a trend for cities voting Democrat to have higher homicide rates than cities voting Republican. The graph below shows the same information from the previous graph for the 50 largest cities in the U.S., but instead of being listed left to right by population, the cities are listed left to right by homicide rate. The average homicide rate for cities that voted Democrat in the 2024 presidential election was 13.3 per 100,000. The average homicide rate for cities voting Republican was lower at 9.2 per 100,000; however, the difference between Democrat (blue bars) and Republican (red bars) cities was not statistically significant (P = 0.17).

Another way of examining homicide in the nation’s cities is to look at the cities with the highest homicide rates, regardless of population. This includes many cities that are not in the top 50 cities by population. For example, the U.S. cities with the highest homicide rates in 2022 were New Orleans (58.4 per 100,000) and St. Louis (57.2 per 100,000). However, New Orleans is only the 52rd most populous city in the United States and St. Louis is only the 76th most populous city.

Of these 50 cities with the highest homicide rates, 44 voted Democrat (blue bars) and only 6 voted Republican (red bars) in the 2024 presidential election. Indeed, the top 18 cities with the highest homicide rates all voted Democrat.

Many large cities have lower homicide rates than entire states

New York City is often accused of having a lot of murders. And it is true that because of its population of more than 8 million, a lot of murders do occur in New York City. However, when expressed as the homicide rate per 100,000 population, New York City has a lower homicide rate than 33 of the states in the U.S. In other words, a person is more likely to be murdered in the state of Texas (7.6 homicides per 100,000) than in New York City (5.3 homicides per 100,000). Similarly, in Florida, a person is more likely to be murdered outside the city limits of Tampa (7.2 homicides per 100,000) than inside the city limits (6.8 homicides per 100,000).

Income matters

There is an association between income and homicide rates, both for states and for cities. However, the association is weak and there are plenty of exceptions. The graph below shows homicide rates (blue bars) versus median individual income in 2022 (red line) for the 50 states. The dotted line shows that there is a trend for states with lower incomes to have higher homicide rates but there are exceptions. For example, Maryland has a very high median income but also has a high homicide rate. On the other hand, Maine and Nebraska have relatively low median incomes but also have low homicide rates.

There is a similar trend for cities. The graph below shows city homicide rates in the blue bars and median individual income in the red line. Cities with the lowest median incomes tend to have the highest homicide rates. But once again, there are exceptions – for example, Washington DC has a very high median income but also has a high homicide rate. Given the proximity of Maryland and Washington DC, it is likely that the same factors accounting for the income/homicide rate discrepancy in Maryland also account for that some discrepancy in Washington DC. On the other hand, Grand Rapids, Michigan and Riverside, California both have relatively low median incomes but also have low homicide rates.

What can we as physicians do?

A few years ago, I was interviewing a medical student who was applying to our internal medicine residency program. He was a foreign medical graduate, having trained in the United Kingdom. He said he wanted to come to the U.S. because of our exceptional residency training programs but was afraid to because he was worried that he would be shot and killed. His impression was that everyone in the U.S. had a gun and most Americans carried a gun with them at all times. His fears are not entirely unfounded – the murder rate in the U.S. is 8-times higher than the murder rate in the U.K. And the homicide by firearm rate is 340-times higher in the U.K. than in the U.S. I tried to assure him that most Americans do not own guns and that our streets are actually reasonably safe but he ultimately decided to stay in London for his residency.

As physicians, we should advocate to reduce all preventable deaths. And homicide is a leading cause of preventable death in the U.S. For Americans between the ages of 15 -24, it is the 3rd most common cause of death and for those between the ages of 25-44, it is the 5th most common cause of death. There are striking racial/ethnic differences in homicide. Homicide is the 21st most common cause of death for Asian Americans of all ages, the 20th most common cause of death for White Americans, the 13th most common cause of death for Native Americans, the 12th most common cause of death for Hispanic Americans, and the 7th most common cause of death for Black Americans. For Black Americans between ages 15 – 24, homicide is the number one cause of death. Black Americans account for 14% of the U.S. population but account for 58% of all American homicide deaths.

For these reasons, all physicians should be concerned about our homicide rates. We need to advocate for protection of victims of domestic violence. Because most U.S. homicides are carried out using guns, we need to advocate for common-sense firearm legislation, such as mandatory background checks for all gun sales. We need to advocate for gun safety measures such as firearm training and use of gun safes. And we need to ask about guns in the homes for our patients with mental health disorders or family members with mental health disorders. Lastly, we need to ensure that our hospitals and medical clinics have policies and procedures in place to protect healthcare workers and patients from harm.

There are dozens of reasons why people kill other people. The Brookings Institute identified three correlates with high homicide rates in different communities: poverty, teenage boys not in school, and young men out of work. We need to hold our community leaders responsible for addressing these root causes of high homicide rates including measures to keep teens in school and keep young men employed.

Homicide is not a disease but it results in death, no different than cancer or myocardial infarctions. In this sense, reducing homicide rates should be a priority for all physicians. The nation’s high homicide rate is not a Democrat problem or a Republican problem. It is an American problem.

January 3, 2025

This month, there as been a lot of rhetoric from U.S. politicians about a new era of American colonialism, and specifically about making Canada the 51st state of the United States. Let’s put aside that the chances of that happening are about the same as pigs flying, but if it did happen, what would be the effect on national health?

Canada is slightly larger at 3.855 million square miles than the United States at 3.794 million square miles. However, Canada’s population is much smaller at 41 million than the U.S.’s at 335 million. California’s population is about the same as Canada’s. So, let’s see how Canada’s health metrics line up against both the U.S. as a whole and the 50 states individually.

Life expectancy

Americans don’t live as long as people in most developed countries. Data from the OECD indicate that in 2021, the U.S. life expectancy was 76.4 years. In the graph below, only Mexico and Poland have shorter life expectancies. Canada’s life expectancy was 81.6 years; in other words, Canadians live an average of just over 5 years longer than Americans.

Within the U.S., the state with the longest life expectancy is Hawaii at 80.7 years, still one year less than Canadians. So, if Canada became the 51st state, it would have a longer life expectancy than all the other 50 states, as well as the District of Columbia.

Infant and maternal mortality

The U.S. does not fare very well when it comes to the heath of infants and new mothers. With an infant mortality rate of 5.4 per 1,000 live births in 2021, only Mexico and Turkey have worse infant mortality rates. Canada’s infant mortality rate of 4.3 per 1,000 live births is lower than all but 8 of the 50 states (North Dakota, Massachusetts, New Jersey, Oregon, New Hampshire, Iowa, California, and New York).

Not only are infants more likely to die in the U.S., but so are new mothers. The maternal mortality rate in the United States was 23.2 per 100,000 live births in 2022 with the lowest maternal mortality rate in California at 10.5. The maternal mortality rate in Canada in 2022 was only 8.53. Although some experts have suggested that Canadian maternal mortalities are underreported, even the highest estimate is 11 per 100,000 live births which is still less than half of the U.S. rate and would still likely be the lowest of all the states.

Infants born the the United States are more likely to have a low birth weight (8.6% of births) than infants born in Canada (7.0% of births). Of the 50 states, only three have a lower rate of low birth weight infants than Canada: Washington, Vermont, and Alaska. Low birth weight babies are more likely to develop infection, respiratory difficulties, chronic disease, and developmental disability. So, Canada seems to make better babies.

Cost of healthcare

Okay, so Canadians live longer and have better maternity care than Americans but Canada has nationalized universal healthcare and we all know that government healthcare is a lot more expensive than privatized free market health insurance, right? Well…no.

Americans spend way more money each year on healthcare than any other country. The 2023 OECD estimate is that we spend an average of $13,432 per person. Canadians spend an average of only $7,013 per person. When expressed as a percent of the GDP, the US spends 16.7% of its GDP on healthcare versus 11.2% in Canada.

Healthcare expenses vary considerably from state to state. In 2020 (the most recently available state-specific data), state per capital healthcare expenditures ranged from a low of $7,522 in Utah compared to a high of $14,007 in New York. In 2020, the U.S. average per capital healthcare expenditure was $11,952. So, healthcare in Canada is less expensive than any of the 50 states.

In the U.S., only 91.6% of the population had health insurance compared to 100% of Canadians. Perhaps because of this, Americans were less likely to see doctors (3.6 doctor visits per year) than Canadians (5.5 doctor visits per year).

Obesity

Worldwide, of OECD countries, only Mexico has a higher percentage of obesity than the United States – we are one of the most obese nations in the world. In the U.S. 42.8% of the population measured in the obese range (BMI > 30) whereas in Canada, 24.3% of the population measured in the obese range. When expressed as those who are overweight (measured BMI > 25), 59.8% of Canadians are overweight and 73.1% of Americans are overweight.

There is a considerable variation in obesity in the 50 states, ranging from a low of 24.9% of people in Colorado to a high of 41.2% in West Virginia. With its obesity average of only 24.3%, Canada would be the least obese of 51 states. However, at 23.5% of the population being obese in the District of Columbia, if DC became the 52nd state, it would then become the least obese state in the nation.

One of the reasons Americans are more obese than Canadians is that we consume more calories, fat, and sugar than Canadians. In 2021, Americans consumed an average of 3,911 calories per day compared to the Canadian average of 3,589 calories per day. Americans consumed an average of 178 grams of fat per day compared to an average of 162 grams per day for Canadians. And the average American consumed 66 kilograms of sugar in 2021 compared to the average Canadian who consumed only 50 kilograms.

Obesity is a major risk factor for diabetes so it is not surprising that Americans have a higher rate of diabetes (10.1% of adults) than Canadians (8.7% of adults). However, a number of individual states do have a lower diabetes rate than Canada including Alaska, Wyoming, New Hampshire, South Dakota, Colorado, Montana, and Vermont.

Smoking, alcohol, and drugs

In 2022, there were 18.0 drug overdose deaths per 100,000 population in Canada and 32.4 drug overdose deaths per 100,000 population in the United States. Only three states had a lower drug overdose rate than Canada: Iowa, Nebraska, and South Dakota.

In 2022, 11.6% of American adults smoked cigarettes daily. In that same year, 8.2% of Canadian adults smoked cigarettes daily. In the U.S., state-specific smoking prevalence rates vary from a low of 6.7% of people in Utah to a high of 21.0% of people in West Virginia. Only Utah has a lower smoking prevalence than Canada.

The average American drank 9.5 liters of alcohol in 2021 whereas the average Canadian drank 8.3 liters. However, there were 13 states where the per capita average alcohol consumption was less than that of Canada: Idaho, Utah, Nebraska, Kansas, Oklahoma, Arkansas, Ohio, Kentucky, West Virginia, Maryland, South Carolina, Georgia, and Alabama.

Communicable diseases

Too many Americans have vaccine phobias. As a result, in 2019 Americans are more likely to get measles (0.4 per 100,000 population) than Canadians (0.3 per 100,000 population. Americans are also more likely to get pertussis (5.7 per 100,000 population) than Canadians (5.6 per 100,000 population).

Americans were more likely to be newly diagnosed with AIDS in 2021 (9.8 per 100,000 population) than Canadians (3.8 per 100,000 population). However, several states did have a lower rate of new AIDS infections than Canada including Idaho, Maine, New Hampshire, Montana, Vermont, and Wyoming.

Since the beginning of the pandemic, more than 1.2 million Americans have died from COVID for a rate of 365 per 100,000 population. The COVID death rate in Canada is 149 per 100,000, less than half of the U.S. death rate. The COVID death rate in all of the individual states is also higher than the Canadian COVID death rate, with Hawaii coming the closest at 157 per 100,000.

Annexing Canada would improve the U.S. health statistics

A slogan of the incoming administration is “Make America Healthy Again”. Perhaps the best way to achieve MAHA would in fact be to make Canada the 51st state of the United States. Canada outperforms America in just about every healthcare metric and assimilation of our northern neighbors would improve our average life expectancy, obesity rate, diabetes prevalence, infant mortality rate, maternal mortality rate, infant low birth weight rate, communicable disease rate, healthcare costs, smoking prevalence, alcohol consumption, and drug overdose rate.

But annexing Canada would come at a considerable cost that may be too great for Americans to tolerate. Hockey would need to replace baseball, football, and basketball as the national pastime. McDonald’s and Wendy’s would need to put poutine on their menus. And Presidential candidates would have to stop calling their opponents “losers” and start calling them “hosers”.

December 28, 2024

It is the start of the 2024 Christmas holidays and COVID is on the upswing for its annual winter surge. With 5 years of COVID data, we know a lot about its epidemiology and that allows us to predict what will happen over the next month. Unlike other seasonal respiratory viruses, COVID has had two peaks in case numbers every year – one peak in the summer and another peak in the winter. These twice yearly peaks are illustrated in the graph below showing COVID deaths by week since the onset of the pandemic in 2020.

In the first year of COVID, the best way to track it was case numbers, identified by testing for the virus. Today, case numbers are no longer very useful because many people with COVID either test themselves at home or do not get tested due to relatively mild symptoms – health departments are unable to track these cases. Instead, epidemiologists now rely on other data to track COVID including COVID deaths, virus concentrations in wastewater, the percentage of ER visits diagnosed as COVID, and the percentage of COVID tests reported to health departments that were positive for COVID. Because COVID deaths occur days or weeks after initial infection, death rates do not give physicians real-time information on the number of people infected with COVID at any given time. Consequently, wastewater viral loads, percent positivity of tests, and percent of ER visits give better information about current infections. The graphs below illustrate the temporal relationship between COVID deaths (blue bars) and the percentage of ER visits diagnosed as COVID (gold line, top graph) and the percentage of COVID tests reported to health departments that were positive (gold line, bottom graph).

From these graphs, we can see that COVID is currently on the rise in the U.S., just as it has been in early winter since 2020. The CDC has been tracking the viral load in sewer wastewater since 2022. Like % test positivity and % ER visits, the amount of COVID virus in wastewater is an early indicator of COVID and is also currently on the rise in the U.S.

From these data, the winter COVID surge is already starting and will likely peak next month, in January. So, what can we expect from this winter’s surge? From previous years’ data, we can predict the demographics of those who will become severely ill and those who will die next month.

Predicting the demographics of this season’s COVID surge

In the first two years of the pandemic, COVID was much less discriminating than today. Although older persons were more likely to become severely ill or die, anyone could get infected. Younger people, particularly those with diabetes or obesity, were also at risk of death. Today, things are different. Most Americans have some degree of immunity, either from vaccination or from surviving previous infections. Consequently, the COVID demographics have changed. Let’s take a look at what we can expect this winter.

Age is still the greatest risk for death. In the first year of the pandemic, it was notable that children rarely died of COVID infection, relative to adults. This trend has continued and the most recent COVID surges have shown that the overwhelming number of people hospitalized or dying from COVID are older adults, particularly those older than age 75. The graph below shows death rates by age since the beginning of the pandemic with the insert showing an enlarged portion of the graph for last summer’s surge. Most of those who will die this winter will be over age 75 (solid brown line) with a smaller death rate for those between 65 – 74 (dashed pink line). The overall death rate for all ages is the dashed blue line). Those younger than 65 will have a very low death rate.

Men will be still more likely to die than women. In the early years of the pandemic, the COVID death rate for men was about 25% greater than the death rate for women. Although this gender effect is still present, it has lessened over the past year. Last summer, the death rate for men (gold line) was about 22% higher than for women (purple line) and last winter, the death rate for men was only about 8% higher than for women.

White Americans will have the highest death rate from COVID. In the first COVID surge in April 2020, Black Americans had the highest death rate, peaking at 39.17 per 100,000. In the December 2020 surge, Native Americans had the highest death rate at 50.41 per 100,000. In September 2021, Hawaiian Americans and Pacific Islanders had the highest death rate at 26.10 per 100,000. However, since January 2022, White Americans have consistently had the highest death rate from COVID. Conversely, throughout the pandemic, Asian Americans and Hispanic Americans have had the lowest death rates from COVID.

One of the reasons for the lower COVID death rates in Asian Americans and Hispanic Americans was the early adoption of COVID vaccination by these racial groups compared to White American and Black Americans. By September 30, 2022, fully 75% of Asian Americans had been vaccinated as opposed to only 45% of White Americans and 43% of Black Americans.

White Americans are increasingly more likely to be hospitalized. Early in the pandemic, Black Americans and Hispanic Americans were disproportionately more likely to be hospitalized with COVID than other U.S. racial/ethnic groups. However, since 2021, White Americans are increasingly more likely to be hospitalized than other racial/ethnic groups. The graph below shows the percentage of people hospitalized with COVID by race: White Americans are in blue, Black Americans are in light brown, Hispanic Americans are in dark brown, and Asian Americans are in green. Currently, 79.25% of hospitalized patients are White Americans. Given that White (non-Hispanic) Americans comprise only 58% of the population, it is clear that White Americans are disproportionately hospitalized compared to other U.S. racial and ethic groups.

Older Americans are increasingly more likely to be hospitalized than younger Americans. Early in the pandemic, more than half of Americans hospitalized with COVID were younger than age 65. Since 2022, this has changed with people older than 65-years-old accounting for an increasingly large majority of those hospitalized with COVID, as shown in the graph below. Currently, approximately 70% of hospitalized patients are older than 65 (red bars) with 15% of hospitalized patients between ages 50 – 64 (green bars), about 10% between ages 18 – 49 (blue bars), about 2% age 5 – 17 (brown bars), and 3% age 0 – 4 (purple bars). This is likely due to an overall increase in U.S. “herd immunity” with the result that now COVID preferentially preys on those who are older.

Influenza is catching up but more people will still die from COVID. Historically, influenza was the main seasonal virus causing death in the United States. Since the beginning of the pandemic, COVID has supplanted influenza. Data from last year’s winter respiratory virus season shows that influenza hospitalizations (blue line) are catching up with COVID hospitalizations (gold line). RSV (purple line) continued to be an important reason for hospitalization but only accounted for about half as many hospitalizations last winter as either COVID or influenza. The summer 2024 COVID surge was not accompanied by an increase in either influenza or RSV hospitalizations as is typical since influenza and RSV mainly occur in winter.

However, when it comes to death from respiratory infections, COVID still reigns king. The graph below shows the percentage of U.S. deaths due to COVID, influenza, and RSV. Looking at last winter’s respiratory virus surge, COVID (gold bars) accounted for far more deaths than influenza (blue bars). RSV (purple bars) was an uncommon cause of death. This is in keeping with a recent study published in JAMA showing that unvaccinated patients admitted with COVID were more likely to either die or require mechanical ventilation than unvaccinated patients admitted with either influenza or RSV. Patients who had been vaccinated for COVID who then were admitted with COVID infection were also more likely to die or require mechanical ventilation than patients vaccinated for influenza who were admitted with influenza infection. The bottom line is that we will likely see about the same (or more) patients admitted with influenza this winter as compared to COVID but the patients with COVID will be considerably more likely to die.

Is COVID ever going to go away? At the beginning of the pandemic, there was hope that COVID would act like the Spanish influenza pandemic of 1918 and fade away. By the winter of 1920 – 21, the influenza serotype responsible for the Spanish flu had essentially disappeared from the United States. However, after 4 years, COVID is still going strong. What has changed are the various serotypes of COVID that are responsible for the different seasonal surges. In this way, COVID is very similar to influenza, with each season’s surge caused by different serotypes due to constantly mutating virus. Similarly, like influenza, immunity to last year’s variant does not provide complete immunity to this year’s variant. And immunity wanes over time. Together, these two features of COVID immunity result in people being susceptible to infection every year and many Americans have had multiple episodes of COVID infection over the past 5 years. For these reasons, it is unlikely that COVID is going to go away in the foreseeable future. Perhaps the best that we can hope is that the summer surge eventually disappears and the winter surge more closely resembles influenza in terms of death rates.

So, what should hospitals do to prepare?

First, and most importantly, get as many patients vaccinated as possible. By now, it is very clear that vaccination does not prevent all infections but vaccination does help reduce the severity of infections so that vaccinated persons are less likely to require hospitalization or to die. Vaccinating the community with the most recent COVID vaccine will reduce hospital bed and ICU utilization. An effective way of getting the word out about vaccination is by messaging patients through the hospital’s electronic medical record, especially when the electronic messages come from patients’ primary care providers.

Second, be sure that hospital supplies of personal protective equipment and COVID tests are sufficient. This equipment includes eye protection, N-95 masks, gowns and gloves. Not only does the central hospital lab need to be fully stocked with COVID test reagents, but outpatient clinics should be stocked with point-of-care COVID test kits. Because of the overlap between influenza and COVID symptoms, point-of-care test kits that cover both influenza and COVID should be used instead of influenza-only and COVID-only test kits.

Third, because the prevalence of COVID will increase in January, hospitals should consider resuming day-of-surgery preoperative testing for COVID. Patients with COVID infection may be asymptomatic in the first couple of days of infection and infected patients are more likely to have postoperative complications. This is not only bad for the patients, but is also bad for the hospital since it can increase the cost of elective surgical admissions, thus eroding the DRG profit margin from surgeries. It is also bad for the individual surgeons who can have their surgical outcome statistics negatively affected.

Fourth, ensure that the hospital pharmacy is adequately stocked with dexamethasone, tocilizumab, baricitinib, and remdesivir for patients admitted with severe COVID infection. Because the number of influenza admissions will likely be similar to the number of COVID admissions, the hospital pharmacy should also be adequately stocked with Tamiflu.

Fifth, update hospital COVID inpatient treatment guidelines. The recommendations for managing hospitalized patients with COVID infection have evolved over the past 5 years. Be sure that the hospitalists and others who admit inpatients with COVID infection are up to date with the current recommendations for testing, isolation precautions, oxygen therapy, pharmacologic treatment, and other aspects of inpatient management.

Sixth, update COVID outpatient treatment guidelines. Because a very high percentage of primary care physicians are now hospital-employed, hospitals have the ability to standardize outpatient treatment within their provider networks. Hospitals should be sure that their evidence-based guidelines incorporate updated recommendations for testing and isolation as well as indications for treatment with nirmаtrеlvir/ritonavir (Paxlovid) and indications for hospitalization.

Don’t let your hospital suffer from COVID fatigue

Let’s face it, we are all sick and tired of COVID and it seems like life is back to normal again. You need a reservation again to get into top restaurants. Movie theaters are packed again. And at Thanksgiving, family dinner tables were full again. In the past two months, I’ve been on 12 different airline flights and on each, I was either the only person or one of only a small percentage of travelers wearing a face mask. But COVID has not gone away. Currently, 5,000 – 6,000 Americans are hospitalized each week and 400 – 500 Americans die each week from COVID. Historical trends indicate that these numbers will increase 4-fold to 5-fold over the next month. By making a few simple preparations today, our hospitals will be ready for the upcoming COVID winter surge.

December 23, 2024