The Climate in Emergency

A weekly blog on science, news, and ideas related to climate change


Climate Change and Medicine

As some of you know, my family has had altogether too much reliance on the healthcare system of late, prompting me to wonder about healthcare and climate change.

I have written before about some of the ways that climate change threatens public health (see here and here and here), but what about ways that healthcare itself threatens the climate? I had never heard the issue raised, and I’d noticed that medicine seems to be one area in which even environmentalists don’t stop to consider carbon footprints. Even I will drive to a medical appointment, and hospitals are obviously intensive users of energy.

Is medicine an area that just has to be bad for the planet? And, if so, does that mean that in the carbon-neutral future that is coming (remember; “unsustainable” means “going to stop eventually, one way or another”) will our standards of healthcare necessarily suffer?

The short answer is no, probably not.

Carbon-footprinting Healthcare

What does healthcare cost the planet in terms of greenhouse gas emissions? Figures seem to vary depending on the source. In the course of researching this article, I have seen healthcare’s share of the US carbon footprint quoted as 7%, 10%, or other numbers, for example. Most likely, the difference is due to variations in how the footprints were calculated–as I’ve explained before, there is more than one right way to calculate any footprint, and carbon footprint analysis is at its most useful when multiple footprints are calculated the same way and compared.

It does appear that the footprint of healthcare varies from one country to another, sometimes dramatically, though figuring out the exact nature of the variation is difficult. Consider that countries vary dramatically in their total carbon footprint, their population sizes, their ability to provide healthcare to their population, and their healthcare outcomes. It’s not immediately obvious how to make a fair comparison.

But some countries are examining their healthcare footprints with an eye towards improvement. An article about the UK’s National Health Service (NHS) explored that country’s healthcare-related emissions in some detail.

The NHS as a whole accounts for 4% of the UK’s total greenhouse gas emissions. Of that 4%:

  • Patient and staff travel accounts for 16%
  • Pharmaceuticals account for 20%
  • Other procured goods and services, including food and medical devices, account for 30%

So, just to be clear, that means 0.8% of the UK’s total greenhouse gas emissions are related to the production of pharmaceuticals, a small figure to be sure, but fairly impressive for a single industry.

Of course, 16+20+30 doesn’t equal a hundred. The article did not explain the remaining 34%, but my guess is that much of it is related to electricity and heating fuel used by hospitals and other facilities.

It’s not clear to me how a similar breakdown might look in a different country, but it seems likely that countries that are broadly similar economically and provide a broadly similar standard of medical care will also have similar emissions sources, even if the total size of the footprints are different.

Let’s see if we can break some of these figures down even further.

Pharmaceutical Production

Since pharmaceuticals account for about a fifth of the UK’s healthcare-related emissions, it’s important to understand where those emissions come from. How does the production and sale of a drug emit greenhouse gasses?

The short answer is I have not been able to find out. Since mass-produced drugs are made in factories, I suspect most of the emissions come from simply running the factories–mostly electricity and refrigerant (for both air conditioning and any stages of production that require chilling), plus fuel, if heating is not electric. Transportation of ingredients and finished products is probably also important, as is the production of plastic for packaging. But it would be useful to know details, since that would enable us to determine what emissions are truly excessive–then we could set reasonable expectations for pharmaceutical factories.

One research team has least approached the question by looking at the carbon footprints of individual drug companies. The group’s focus appeared to be American, but the companies studied are mostly transnational. The team used figures from the few pharmaceutical companies that report their greenhouse gas emissions to calculate annual emissions per million dollars of revenue for each company and for an industry average. They came up with 48.55 tonnes of carbon dioxide equivalent (CO2e) per million dollars of revenue–almost half again the figure for the auto industry.

Now, since the auto industry involves more millions of dollars than the pharmaceutical industry does, its total footprint is still larger (plus cars produce their own emissions post-production and pills don’t), but it’s still a startling figure.

Just as startling is the fact that while the pharmaceutical industry as a whole needs to radically shrink its emissions in order to meet Paris targets, some companies have been working on shrinking and have met their Paris goals already–and these are among the most successful companies, namely Johnson & Johnson, Amgen inc., and Roche Holding AG. It appears the less-green companies have no excuse for not cutting back.

Unfortunately, the numbers don’t tell us as much as they might. Some companies, such as Bayer, produce more than just pharmaceuticals and don’t break down their reported emissions in a way that would let us see what the figures for their pharmaceuticals alone really are. While we can be assured that Bayer really does have a large carbon footprint, we can’t put that number in context. We can’t fairly compare companies if we don’t even know which industries the numbers refer to.

And none of this answers my original question.


Articles on what hospitals can do to reduce their carbon footprints are fairly easy to find, and their content should sound fairly familiar to anyone interested in sustainability (carpool to work, cut back on meat, etc.). Most don’t provide information on what hospital-related emissions actually are.

More interesting is a study comparing greenhouse gas emissions from specific operating rooms in different countries. The big surprise is that anesthesia is a major factor; anesthetics are greenhouse gasses, and they vary in the strength of their greenhouse potential. Desflurane has between 5 and 18 times the global warming potential of its competitors, yet it is a favored choice in some operating rooms. The surgical suits studied in Vancouver and Minnesota both use a lot of desflurane, so anesthesia accounts for over half their carbon footprint, verses less than 5% of the footprint of surgery at Oxford. The total CO2e of the North American sites is actually ten times that of the UK sites, largely because of desflurane.

Another detail that caught my eye is that heating, ventilation, and air-conditioning together account for a much higher proportion of energy use for operating rooms than for other hospital facilities because building standards for operating rooms are different–though the article didn’t explain why. Does energy inefficiency somehow improve patient care? It’s possible it does in some indirect way.

What I want to know–and have been unable to find out so far–is how the footprints of hospitals compare to those of other centers of human activity. No matter what else it is, a hospital is a facility where large numbers of people live and eat together, and where other people come to work. Many of its emissions sources should therefore be similar to what one would find at a university with on-campus housing. Is the footprint of a hospital larger than that of a school of similar size? If it is, how much is waste and how much is just a necessary part of providing excellent healthcare?


While transportation is a factor in the carbon footprint of anything healthcare facilities must move, including food, pharmaceuticals, and waste, the article I linked to earlier counted “transportation” as only involving the movement of people. In many cases, these movements of people are the same as for any other employer–staff coming to work and patients coming in for scheduled treatment, mostly by car. As with any other employer, the quickest way to minimize these emissions may be to minimize the transportation itself–encouraging car-pooling among staff, for example.

But the transportation category also includes the use of ambulances. These are not efficient machines. Gas mileage varies, depending on various factors, but I checked a number of sources, and it looks like the figures are clustering around 10 miles per gallon for both ground-based ambulances and helicopters (that’s diesel fuel for ambulances and either high-grade gasoline or kerosene–jet fuel–for helicopters). And the problem is that while the engines could perhaps be made more efficient, their use can’t be minimized without compromising (or at least radically changing) care.

So what portion of of the healthcare carbon footprint is emergency transport?

I could not find an article that simply answered that question. I could find one that gave a per-capita figure for emissions of ground-based ambulance service in Australia: it’s 0.003 metric tonnes CO2e.

I then looked up the total annual per capita carbon emissions for Australia (20.58 metric tonnes CO2e) and the proportion of Australia’s total carbon footprint that is attributable to healthcare (7%). Some arithmetic reveals that Australia’s per capita healthcare-related emissions are roughly 1.44 metric tonnes CO2e per year, just 0.2% of which is attributable to ground-based ambulance rides. The figure for other first-world countries is likely similar.

Air ambulances–helicopters–are a different story, but one I can’t really tell. The same article that gave me the ground-based figure also said that air ambulances account for almost 200 times the emissions that ground does. Unfortunately, the article did not make clear whether that is a comparison between air and ground services as a whole, or per-trip figures, or per-kilometer figures, or something else. Logic suggests it can’t be all-of-the-above. But since Australia has large areas that are sparsely settled, it likely uses air ambulances much more extensively than, say, the UK does. Its helicopter-based emissions are likely less comparable to that of other countries than their ground-based ambulance figures are.

What we can say is that whether the emissions of ground-based ambulances can be a substantially reduced or not, they are a drop in the bucket. Emergency helicopters may be a more important contributor, however, at least in some countries.

Post-petroleum Healthcare

While many of the articles I found during research were aimed at reducing the carbon footprint of healthcare, my focus, as I mentioned, is a little different. Shrinking footprints is important, of course, but neither I nor most of my readers are in a position to shrink healthcare footprints directly. What are we supposed to do, boycott our own medical care in protest? No, our job here is to support (and demand!) climate-friendly political leadership.

But I want to know what we can look forward to. What will healthcare in a carbon-neutral society look like? Do we have to think about the ethics of a trade-off, restricting healthcare for the good of the rest of the world, or is such a conflict really a non-issue?

What We Know

What do all the facts and figures I’ve collected suggest?

First of all, healthcare as a whole tends to be about 10% of each country’s total carbon footprint or less. That means most countries could make substantial progress towards carbon neutrality without touching healthcare at all. But there are reasons to believe healthcare footprints can shrink without changing the standard of care.

  • A substantial source of emissions must be electricity use by healthcare facilities. Switching the electricity grid to renewables will therefore automatically shrink the healthcare footprint even if the facilities themselves don’t change.
  • Many healthcare-related emissions types are the same as for similarly-sized facilities in other industries, and can therefore be reduced in the same way: buildings can be better-insulated; lights can be switched to LEDs; unused equipment can be turned off; waste can be minimized; food (in hospitals) can be sourced locally and made largely vegetarian; and so on.
  • Many healthcare-specific emissions types can shrink: operating rooms can switch away from desflurane; the footprint of pharmaceutical production can be reduced (as evidenced by Johnson & Johnson); and ground-based ambulances can run on biodiesel.

There is only one area where I suspect major changes might need to be made; air ambulances could use biofuels, too, but since these are likely to be more expensive than petrofuels, fuel-intensive operations, like Australia’s helicopters, might be cost-prohibitive. Various structural changes to the system might be necessary to maintain the standard of care.

That’s OK. Structural changes can be made.

Possible Complications

All of the above suggests that healthcare could stay basically the same and become carbon neutral–but that’s not true because a carbon neutral society will have to change in ways that will in turn influence healthcare. Exactly what these changes might be is impossible to predict, but we can do a little educated speculation.

Improved Health

We know that modern environmental problems cast a healthcare shadow, both directly and indirectly. From pollution, to extreme weather, to increased violence, to mental health concerns, the environmental crisis is bad for people. And some things that are also bad for people, such as a sedentary (automobile-based) lifestyle and a diet rich in animal products, exacerbate the crisis further. This not to say that we’re sicker now than our pre-industrial ancestors were–we’re not–but most of the factors that have improved our health (antibiotics, vaccines, public sanitation) should not be threatened by carbon neutrality.

So a carbon-neutral world should see its healthcare needs drop, thus further shrinking the healthcare footprint.

An Altered Economy

Industrialization makes it possible to concentrate large numbers of people in one place; the cities of the past were smaller, sometimes much smaller, than those of today. Since carbon neutrality is likely to make fuel very expensive, the long-distance transportation of food and other goods will likely become economically nonviable–urban populations will therefore have to shrink. Even if carbon-neutral big cities prove to be possible, we have to face the fact that most of the world’s major cities are going to be lost to climate change, even if we do achieve carbon neutrality soon (remember atmospheric lag); many coastal cities will drown, while many inland cities will run out of water in droughts or simply burn. Some cities will simply become too hot to live in.

So the future will likely have smaller and more spread-out population centers than we have today, a change which will have a huge impact on the economics of hospitals. Consider that big hospitals, the ones that can use economies of scale to offer world-class care, tend to be in or near big population centers–that’s where the patients are. Paying for rural healthcare is hard. It’s going to get harder when patients can’t afford to travel much.

Smaller Population

There are those who disagree with me on this, but I hold that carbon neutrality will require a smaller human population, at least over the long-term. Hopefully we can make the change through attrition alone. But fewer people also means reduced healthcare needs, further reducing the carbon footprint of care–and making it harder to pay for.

The Vision

Earlier this year, I published a novel set in a post-petroleum society. While healthcare as such is not covered in the novel, in the course of world-building I did think about how healthcare in the future might work. Here are some of the ideas I came up with.

  • Because populations are smaller and more spread-out, doctors, dentists, psychotherapists, and other such professionals travel. Rather than making house calls, they set up temporary offices, either in tents at the weekly market or, for those who need specialized equipment, in clinics that are shared with other traveling professionals. For example,everybody in a small town might have their semi-annual dental cleaning and check-up the same month, when the dentist and a team of hygienists visit. The next month, some of those people will return to the same office, because now it’s the office of an orthodontist or an oral surgeon.
  • Each small town will have a tiny clinic that has space for traveling professionals and an emergency center, a birthing center, and perhaps a dozen or so beds for in-patient treatment. On good days these clinics will be mostly empty. The idea is to minimize travel for patients, most whom will now live within two or three miles of basic medical services.
  • Large, full-service hospitals will exist for specialized services. These will function as small cities, with food production, machine shops, and staff and visitor housing all on site. Most patients will have traveled long distances to get to the hospital after having exhausted the capacity of local options.
  • Emergency transportation exists and is powered by biofuels. It is minimized by the use of online consultation; techs working either in the local clinic or, in some cases, in the patient’s home, can collect diagnostic data and send it to teams of relevant experts elsewhere. Some treatments can also be given by techs, nurses, general practitioners, and even robots working at the direction of experts who are far away (usually in the major hospitals).
  • In some cases air transport is accomplished, not by helicopters, but by semi-autonomous drones. The drone carries a paramedic and relevant equipment to the patient’s location, where the paramedic stabilizes the patient and, if necessary, loads the patient into the drone. The drone flies autonomously to the nearest medical facility, adjusting the flow of medication or oxygen on the way if necessary. Its onboard AI can also talk to the patient and record messages. Meanwhile, the paramedic hires a horse-drawn cab to get back to the clinic. Because the drone is only carrying a single human being at a time, it can be much smaller and use much less energy than a helicopter, which must carry a pilot and a medical team in addition to the patient.

All of the above is, of course, speculation on my part. But informed speculation can be useful; it makes the future seem a good deal less scary.


How Heat Hurts

I got heat exhaustion today. Unfortunately, this is not an unusual occurrence for me–I seem to be unusually susceptible. I don’t know why. Heat exhaustion is one of several types of heat-related illnesses. It is not, in itself, normally dangerous, but can progress to heat stroke, which can kill you.

Heat is a matter of weather—but it is also a matter of climate. Obviously, global warming means more hot days, but the increase doesn’t work quite the way intuition says it should. Intuitively, an average warming of, say, one degree Fahrenheit, should add one degree on to typical daily temps. So if your normal summer day was 90 degrees, now it’s 91 degrees. Not a big difference. But that’s not how it works.

As I have addressed before, a small increase in average temperature results in a large increase in the frequency of heat waves. This is because there is a well-established link between rarity and severity across many different types of variation, from body height to intelligence to air temperature. A slight increase above average (a few degrees, a few inches….) translates into a dramatic decrease in frequency of occurrence. How often do you see people who are six feet tall? Now often do you see people who are seven feet tall? The difference seems larger than what a mere twelve inches would imply.

A hotter normal means that severe heat waves that used to be very rare become common-place, while the human vulnerability to heat injury remains roughly the same.

My illness today is not particularly a climate change story–it was not one of those events that make it obvious normal has changed. Hot days in mid-June are not new for Maryland. But the experience did inspire me to do some reading about how heat stroke actually works—more people die from heat than from all other natural disasters combined. I decided I wanted to know more about what happens in the body when it gets hot.

Please note that I’m skipping over issues like how to recognize and treat heat-related problems. For that information, look up a public health website maintained by a reputable medical institution


When I received emergency medical training years ago, I learned that heat exhaustion is essentially a form of aggravated dehydration—the body is not hotter than it should be, but keeping cool is taking too much effort, including loss of so much water through sweat that blood volume drops. The symptoms are mostly the body’s attempt to compensate for lost blood volume in order to keep adequate blood flow to the brain. Heat injury and then heat stroke, in contrast, result when the body’s cooling system fails (sometimes because dehydration has become critical and the sweat response shuts off—when a person who should be sweating isn’t, that’s a very bad sign) and body temperature rises uncontrollably. Doctors then have hours or even minutes to act before the patient literally cooks to death.

Like most simple explanations, this one is not quite right. For example, brain damage in heat stroke is not caused by the brain tissue heating up, as I’d been led to believe–instead, excessive heat causes the blood/brain barrier to become leaky, allowing substances into the brain that should not be there, and that causes damage. Heat stroke, though triggered by heat (either through passive exposure to high temperature or to excessive exercise in hot weather or under too much clothing), actual injury—and often death—is not the direct result of the body cooking. After all, cooking occurs at specific temperatures (that’s why recipes work), but the temperature at which heat injury occurs is variable. There are documented cases of people surviving core temperatures above 107 degrees Fahrenheit, but there are also many cases of people dying at much lower temperatures. The body is a complex system. Heat-related injury and death are the result of complex responses to heat, not the heat itself.

The information in this post, except where noted, is taken from a document produced–or at least presented–by the US Military (service members are at high risk for heat stroke, therefore the military is interested in the issue). The “report date” of the PDF is listed as 2012, although since it is evidently a chapter in a longer book, I don’t know if the report date is earlier or later than the copyright date of the book. I don’t know how old this information is. It’s a dense read, but I’ve attempted to summarize the main points below.

How Heat Stroke Works

Not everyone is equally vulnerable to heat stroke. There are long lists of circumstances that create higher risk, so many that it might seem everybody must belong to at lest one of them—but it’s important to note that some risk factors are a matter of choice (running marathons on hot days) and some are not (being very young, very old, already ill, or poor). There are obvious social justice issues here, as I’ve discussed before.

Interestingly, several risk factors do not involve simple vulnerability to heat (as in our marathon runner, or a home-bound elderly person without an air conditioner) but rather impairments of the body’s ability to respond. A sunburn or a heat rash can impair the body’s ability to cool itself, for example. Illness or inflammation (e.g., pneumonia) makes heat stroke more likely. Heart problems, certain medications, or low potassium or sodium levels also either make heat stroke more likely or more dangerous. These facts alone should suggest the medical complexity of the problem.

Heat stroke is also a much more drawn out process than the idea of cooking would imply. Literal cooking ceases as soon as the object being cooked cools, but heat stroke isn’t over when the victim’s core temperature is brought back to normal. If he or she lives long enough, the bodily changes initiated by the heat will continue to play out. The patient will probably run a fever (which actually helps the body heal), and may also go through periods of abnormally low body temperature.  Kidney failure will probably occur between two and 24 hours after the initial collapse. The liver will likely fail after 24 to 48 hours. Mortality rates often rise about a month after mass heat stroke events (like heat waves), after patients have been discharged. The risk of dying from cardiovascular, kidney, or liver disease can remain elevated for 30 years. There may be long-term cognitive impairment. And since many illnesses or deaths are either never recognized as related to a patient’s heat-stroke history, or never reported as such, the true prevalence of these problems is likely much higher than the data we have indicate. There has been little research done on how these long-term problems happen, and no one really knows what to do about it yet.

The bottom line is that the number of people who die of a heat wave is much higher than the number of people who die in a heat wave.

Heat stroke is actually several processes, although the whole story is not yet clear even to scientists.

The dominant process may actually be an immune response called Systemic Inflammatory Response Syndrome (SIRS). This is the same–or at least very similar–to what happens when an infection enters the bloodstream, a condition called sepsis or, less technically, “blood poisoning.” Its symptoms include fever and a whole series of both helpful and non-so-helpful biochemical changes.

Heat-induced SIRS is actually not caused directly by heat. Instead, when the body redirects more blood flow to the skin (heat stroke victims are typically bright red), the internal organs necessarily get less. Insufficient blood flow can damage the gut lining, causing it to leak endotoxin into the blood. The endotoxin, in turn, triggers SIRS–if severe enough, the endotoxin or SIRS (I’m actually not clear which–it looks as though scientists might not be sure, either), destroys the major vital organs, causing death.

Injection of endotoxin alone (into animals) triggers the clinical symptoms of heat stroke.

Another important process is DIC, which stands for Disseminated Introvascular Coagulation. Essentially, the blood starts clumping up, leaving the blood remaining in circulation way too thin. DIC can be caused either by tissue damage (sepsis is listed as a common cause, suggesting that DIC can be caused by SIRS–the immune response I just described–although that is not clear to me from the article) or by direct heat injury to the vascular system.  Besides the real risk of bleeding to death, DIC also causes, or helps cause other problems associated with heat stroke.

DIC can cause kidney failure, for example. But kidney failure can the proteins released by muscles damaged by SIRS, or by heat toxicity itself.  It can be difficult to tell which problems are causes and which are results.

Heat stress is one of several possible triggers for the release of cytokines, a class of messenger proteins that in some circumstances are a necessary part of healing–but experimental injection of these proteins triggers heat stroke symptoms including excessive body heat. In other words, the body doesn’t just get sick because it gets too hot—it also gets hot because it’s sick. Exactly what role cytokines play in actual heat stroke isn’t known, yet, but cytokines are involved in many of the processes and subprocesses of heat stroke.

There are several possible treatments for heat stroke being developed based on this more detailed understanding of the malady, but so far, heat stroke is much easier to prevent than to treat. Prevention consists not just of staying cool, but also in becoming adequately acclimatized–general good health and fitness, plus a recent history of being uncomfortably but not dangerously hot fairly often dramatically increase the body’s ability to safely withstand heat. In other words, HAVING a working air conditioner can save your life, but using it often (hiding from summer heat) puts you more at risk for those times when you do have to get by without it–if, for example, there is a power outage during a heat wave.


All of this might sound like unrepentant geeking out on my part. I am, in fact, an unrepentant geek, but my primary motivation for this post is, as I said, to take a close look at a malady likely to become ever more familiar, both to us individually and as a matter of public health policy.

One study that looked at the UK has predicted that, as a result of global warming, the incidence of death from heat stroke in that country will double by 2050. That’s only just over thirty years away.