Sepsis Rates Are Up. Here’s What Might Explain Why.

In his MedPage Today column, \u201cFaust Files,\u201d Faust weighs in on the biggest news in medicine in both written and video commentary.<\/p>“,”affiliation”:””,”credential”:”MD, MS, MA”,”url_identifier”:”jf6550″,”avatar_url”:”https:\/\/assets.medpagetoday.net\/media\/images\/author\/Faust_330px.png”,”avatar_alt_text”:”Jeremy Faust”,”twitter”:”https:\/\/twitter.com\/jeremyfaust”,”links”:null,”has_author_page”:1,”byline”:”Editor-in-Chief, MedPage Today”,”full_name”:”Jeremy Faust”,”title”:”Editor-in-Chief, MedPage Today, “,”url”:”https:\/\/www.medpagetoday.com\/people\/jf6550\/jeremy-faust”,”bluesky”:””}]”/>

Recently, the Boston Globe‘s Jessica Bartlett asked me about sepsis for an excellent story published a few days ago. In her story, Bartlett investigates why sepsis cases have dramatically increased in Massachusetts (and likely elsewhere) in recent years.

As always, I said much more in our conversation than what made the story — in this case, one brief quote — so I’d like to flesh out my thoughts on this fascinating topic.

But before we get to why rates are up, we need to define sepsis correctly and take a brief tangent into how sepsis is treated.

What Sepsis Is and Is Not

First, we need to clear up the definition of sepsis. The definition of sepsis is rarely agreed upon, even by seasoned clinicians, I find. Technically, the definition of sepsis is that it’s the body’s counterproductive response to infection. As I told Bartlett, sepsis is self-imposed scorched earth, rather than the destructive powers of infectious microbes themselves. (She got this right in the lede, which I was pleased to see.)

You’ll hear other sepsis definitions (i.e., incorrect ones) bandied about, even by seasoned health and science professionals. Some of these overlap with true sepsis more than others. You’ll hear that sepsis is when an infection spreads to the blood, or that sepsis is when an infection spreads to distant parts of the body. You might hear that sepsis is when an infection causes organ damage. All of these are incomplete, if not outright wrong. In the purest sense, sepsis is an autoimmune problem in response to an infection, not just the infection itself raging out of control.

How Is Sepsis Treated?

Untreated, sepsis is a real killer. Early detection is key. The idea is to beat the infection as quickly as possible. We call this “source control.” Our main tools are antibiotics or removal of infected tissue thought to be the problem (say, the appendix). The hope is that if we do this in time, the body’s maladaptive reaction (i.e., sepsis) will abate before lasting damage, or worse, is done.

While I’m making a distinction that most don’t — that is, the difference between sepsis and insufficiently controlled infections — at present, our treatments for these entities aren’t actually different. For example, steroids might tamp down an overreacting immune system, but it’s often unclear which infected patients might benefit or be harmed by them. Moreover, attempts to treat sepsis with medications directed at the immune system rather than the infection itself have failed, spectacularly in some cases. Beyond source control, the crucial care we provide is “supportive,” and includes intravenous fluids or, in more serious cases, medications that treat dangerously low blood pressure.

How Can Sepsis Rates Be Up?

If sepsis is an immune system problem, not a microbe problem, how can sepsis rates have risen in recent years? (The answer is not COVID-19, nor is it fallout from COVID. Increased sepsis rates preceded all of that, and for the most part, sepsis is caused by bacteria, not viruses.) Possible answers reside in epidemiology. There are two large, not mutually exclusive, categories to consider.

1. Increases in recognition/diagnoses by clinicians
2. True increases in incidence among the population

If increased recognition of sepsis explains higher numbers, then nothing about the world has changed. If increased incidence explains the rise, then something in the world has.

Within the increased recognition category, there are two further possibilities to consider, both within the concept of “capture.” The first is increased awareness, say after a campaign to educate or refresh clinicians about diagnostic criteria. This might correct previous under-diagnosis (i.e., “under capture”; cases were being missed).

The second possibility, ironically, is over-diagnosis. In those cases, the same educational campaign could lead to the over-application of relatively liberal diagnostic criteria in some cases (i.e., “over capture”). For example, high blood pressure is defined as readings over 130/80. But if a doctor adds a hypertension diagnosis to a patient who had a couple of readings of 132/80, but who is otherwise completely healthy and often has lower numbers when checked, is that over-diagnosis? Possibly. Regardless, these examples illustrate how changes in detection can drive increases in case counts, even if the number of cases in the world has not actually changed.

But what could explain true increases in a condition among the population? Here too, there are two possibilities to consider in the case of sepsis: something about the world of microbes or something about the hosts (i.e., the people getting infected) has changed.

It could be that nastier microbes are becoming more common. There’s no compelling evidence that this explains recent rises in sepsis, but it’s technically possible. In fact, over-treating mild infections with powerful antibiotics today could lead to a world with increasingly perilous bugs tomorrow, including ones more likely to cause sepsis. (That means that over-treating mild infections now could lead to problems down the road, including more resistant bacteria that can cause sepsis.)

The other possibility is that the population carries a higher risk of sepsis than prior generations. The two major personal risks for sepsis are age and medical complexity (with compromised immune systems being the major driver). Our population is indeed getting older and more medically complex. In fact, these go squarely together. One of the reasons that people live longer is advances in medicine, including powerful treatments that lead to immune compromise as an acceptable risk.

Case Study: Do Medical Advances Also Lead to Sepsis?

Let’s imagine two 75-year-old cancer patients, one living in the year 2016, the other in 1916. They are otherwise similar. The patient living in 1916 would be dead within a year. The modern patient achieves remission because of chemotherapy. Now, 10 years later, his cancer comes back. Another round of treatment is given. The treatment works, but it compromises his immune system. He gets an infection, which causes a fatal case of sepsis.

Did the modern patient die of sepsis or cancer? The answer is both. (That said, there can only be one underlying cause on a death certificate. These choices are rarely straightforward.)

Regardless, the 20th century patient didn’t get to live long enough to acquire a higher risk of sepsis, while the 21st century patient did. At the population level, this means that sepsis deaths are now more frequent than 100 years ago, even if we adjust for age. Indeed, it is likely that on average, today’s 85-year-olds have a higher risk of sepsis than their 20th century predecessors. That’s because the relatively few people who lived to see age 85 in 1926 were remarkably healthy specimens. People with complex medical problems did not get to live to age 85 in 1926. Today, however, many people with complex medical problems do live to age 85 and beyond, albeit with the acceptable risks that the responsible medical treatments carry.

Importantly, we should not ignore this just because increased sepsis rates might reflect population epidemiology. If we were better at detecting and treating sepsis, our modern 85-year-old could go on to have another 5, 10, or 15 years of quality life — to say nothing of the subset of far younger people who develop sepsis.

These Lessons Apply Broadly

If you’ve made it this far, you now understand some nuances in epidemiology that many do not. These principles apply to many conditions, not just sepsis. Case definitions are supremely important. When we loosen case definitions, more people seem to have a disease. That’s good if this leads to interventions that are genuinely helpful. But the downside is that some people get treatment (with side effects) that they never needed. Also, epidemiologically, if you loosen the criteria for a condition, you might accidentally make that condition appear less dangerous. If we lowered the definition of high blood pressure to 125/80, droves of people would suddenly have high blood pressure. But the increase would be driven by the mildest cases, making the condition look less dangerous overall. That’s because very few people with blood pressures of, say, 128/80 will have massive heart attacks. This is called diluting the denominator.

Interestingly, when it comes to sepsis, Bartlett’s story in the Boston Globe found that increased sepsis rates were driven by severe cases, not more mild cases diluting the denominator. That matters because in recent years, some of the competing sepsis criteria (yes, there are competing criteria) contain less stringent definitions than others. That’s not what explains the rise in sepsis in Massachusetts, it seems.

So, which of the above explanations we explored actually explains it? I don’t know for sure. What I do know is that it’s important that we find out.

A version of this post originally appeared in Inside Medicine.