You can tell what a doctor does for a living by the ICD-9 codes they have memorized. There is an ICD-9 code for nearly every medical condition. Weightlessness is 994.9. Must be there for NASA, I have yet to see a weightless patient. Decapitation by guillotine is E978. There, I suppose, in case Marat returns from the dead. There is an ICD-9 code for the initial visit after being sucked into jet engine (V9733XA) and one for subsequent visits (V9733XD). Why do I suspect V9733XD has yet to be used?

780.6 is my personal favorite. Fever. All my patients have fever and 780.6 was certainly the first ICD-9 code I committed to memory. I have an endless interest in fever and after last Fridays post I thought I would toss in my two cents worth. I will remind my readers that I am an adult ID doctor (who I treat, not necessarily how I behave) and unless specifically mentioned, all that follows applies to those who can legally drink, vote and serve in the military.

98.6 F. It is not normal body temperature. Well it is. But it is not. 98.6 F as average body temperature is an enduring medical myth.

In 1868 Carl Reinhold August Wunderlich published Das Verhalten der Eigenwärme in Krankenheiten  (The Course of Temperature in Diseases), then the definitive text on temperature. He was one of the first to recognize fever as a symptom of disease rather than a disease. As only the obsessed of Victorian times seemed to do (see The Professor and the Madman, an account of how the Oxford English Dictionary came into being, as another example) without a spreadsheet or a TI hand calculator he averaged a million observations from 25,000 subjects and came up with 37°C (98.6°F) as the average normal temperature. The book makes me wish I knew German. Well, almost. There is an English translation for the interested.

Because of his book 98.6 became medical dogma, enshrined forever by a red line on a mercury thermometer. My children certainly knew the importance of that red line. If the mercury went past it they did not have to go to school. Ah, mercury thermometers. Endless fun breaking them open as a kid and playing with the mercury. Didn’t do me any harm.  Kids today, with their electronic thermometers, will never know the fun for trying to vacuum up mercury spilled into a shag carpet.

While Wunderlich did suggest that 38°C (100.4°F) is the upper limit of normal, his equipment, while the best at the time, fell a wee bit short by modern standards: they were a foot long, took at least 20 minutes to equilibrate and were mis-calibrated by as much as 1.4 to 2.2°C (2.6 to 4.0°F) higher than today’s instruments.

Unfortunately his 400 page epic, while a verbose and interesting read, is based on slightly inaccurate tools. What does modern thermometry have to add?

A 1992 an analysis of 700 baseline oral temperature observations from 148 healthy men and women found

  • There was a range of 35.6°C (96.0°F) to 38.2°C (100.8°F),
  • an overall mean of 36.8 ± 0.4°C (98.2 ± 0.7°F),
  • a median of 36.8°C (98.2°F),
  • a mode of 36.7°C (98.0°F);

The maximal temperature (as reflected by the 99th percentile) varied from a low of 37.2°C (98.9°F) at 6 AM to high of 37.7°C (99.9°F) at 4 PM.  So people are 98.6 twice a day: once on the way up and once on the way down.
Age did not significantly influence temperature within the age range studied (18 to 40 years) (linear regression, p = 0.99).

Women had a slightly higher average oral temperature than men (36.9°C [98.4°F] versus 36.7°C [98.1°F].
Black subjects exhibited a slightly higher mean temperature and slightly lower average diurnal temperature oscillations than white subjects (36.8°C [98.2°F] versus 36.7°C [98.1°F] and 0.51°C [0.93°F] versus 0.61°C [1.09°F], respectively); these differences approached, but did not quite reach, statistical significance (t test, p = 0.06, df = 698).

Oral temperature recordings of smokers did not differ significantly from those of nonsmokers. The two big exceptions being ovulating females (but not ovulating males) will be relatively warmer in the early morning and those who work the night shift, who will have the normal variation shifted by 12 hours.

As best I can tell when I quiz them, most medical students are taught that 98.6 is normal temperature, an observation supported by the literature:

Seventy-five percent of the 268 physicians and physicians-in-training surveyed offered 37 degrees C (98.6 degrees F) as their definition of “normal body temperature.” Only 10 (4%) specified a particular body site (eg, oral or rectal) of temperature measurements in their definition. Although 98% believed that body temperature normally varies during the day, there was not a consensus as to the magnitude of such variability. There was also considerable disagreement as to the specific temperatures defining the lower and upper limits of the febrile range. Subjects exhibited a clear preference for the Fahrenheit scale in their responses

So much for knowing about a so called vital sign. But it could just as likely be that they were given the correct information and it didn’t overwrite a lifetime of misinformation.

Not knowing what normal temperature is does have occasional consequences. Every couple of years I see a patient who takes their temperature in the late afternoon, and what do you know, they have a fever. Their temperature is 99.9. They check the temperature every day and every afternoon it is around 100, so they see their doctor who also thinks it is a fever and they start a FUO (fever of unknown origin) evaluation. Work-up, of course, is negative, so after several thousand dollars, the patient is sent to me and I explain normal physiology to them.

From a practical point of view, I start to worry about people when they have a temperature above 101, especially hospitalized patients if they have more than one elevated temperature. The height of the fever and pattern of the fever is usually of little interest (there are a few exceptions unless you are a homeopath), and there is, to my mind, no such thing as low grade fever. Like pregnancy, where either you are or you are not, either you have an abnormal temperature or you do not. And the occasional person who insists “My normal temperature is 97, so 98.6 is a fever for me”? Nothing in the literature to support the assertion and I wait for the first well patient to consistently have a normal temperature that is lower than the usual variation.

It is quite remarkable how tightly normal temperature is regulated, and 98.2 is the optimal balance between caloric requirements and keeping most germs at bay. Being above ambient temperature is an important defense as most organisms cannot grow well at 98.2, much less 102.

…we present a minimal, parsimonious model to account for the cost of maintaining a high body temperature in mammalian organisms. A body temperature of 36.7°C maximizes fitness by restricting the growth of most fungal species relative to its metabolic cost.

Being warm blooded is important to keep us from becoming compost:

The paucity of fungal diseases in mammals relative to insects, amphibians, and plants is puzzling. We analyzed the thermal tolerance of 4802 fungal strains from 144 genera and found that most cannot grow at mammalian temperatures. Fungi from insects and mammals had greater thermal tolerances than did isolates from soils and plants. Every 1 degrees C increase in the 30 degrees C-40 degrees C range excluded an additional 6% of fungal isolates, implying that fever could significantly increase the thermal exclusion zone.

And may even be part of why we dominate instead of dinosaurs:

Given that most fungal species grow best at ambient temperatures, the high body temperature of endothermic animals must provide a thermal barrier for protection against infection with a large number of fungi. Fungal disease is relatively common in birds but most are caused by only a few thermotolerant species. The relative resistance of endothermic vertebrates to fungal diseases is likely a result of higher body temperatures combined with immune defenses. Protection against fungal diseases could have been a powerful selective mechanism for endothermy in certain vertebrates. Deforestation and proliferation of fungal spores at cretaceous-tertiary boundary suggests that fungal diseases could have contributed to the demise of dinosaurs and the flourishing of mammalian species.

And may be something we should watch so we do not go the way of the dinosaurs:

The relatively high resistance of mammals has been attributed to a combination of a complex immune system and endothermy. Mammals maintain high body temperatures relative to environmental temperatures, creating a thermally restrictive ambient for the majority of fungi. According to this view, protection given by endothermy requires a temperature gradient between those of mammals and the environment. We hypothesize that global warming will increase the prevalence of fungal diseases in mammals by two mechanisms: (i) increasing the geographic range of currently pathogenic species and (ii) selecting for adaptive thermotolerance for species with significant pathogenic potential but currently not pathogenic by virtue of being restricted by mammalian temperatures.

So we probably evolved endothermy in part to keep the microscopic world at bay. We can make it even harder on the organism trying to kill us by increasing our core temperature. Every animal that can raise its temperature will raise its temperature in response to infection, in the case of lizards by moving to environments with increased temperature. Although that is only clinically applicable to Fox News analysts.*

And it is estimated that the febrile response is ancient:

A febrile response has been documented in the Vertebrata, Arthropoda, and Annelida. These observations suggest that the febrile response evolved more than 400,000,000 years ago at about the time evolutionary lines leading to arthropods and annelids diverged.

Although how they estimated that number I am uncertain. I doubt they went back in time and took a rectal temperature in T. tex, since with those forelegs it could not have taken their own oral temperature. Maybe one of Wunderlich’s thermometers would have worked.

There are numerous beneficial physiologic effects that occur as part of the febrile response. Virtually all aspects of the physiologic response to infection are, dare I say it, boosted, and many wings of the immune system function better at higher temperatures.

I feel the naturalistic fallacy creeping over me: fever is an almost universal response to infection that evolved millions of years ago and helps enhance the response to infection. So suppressing a fever during infection should be bad. Right?

This becomes a little less clear cut. There are issues as to how the fever is suppressed: mechanical, such as ice or alcohol baths. Pharmacologic: acetaminophen, steroids, aspirin or nonsteroidal anti-inflammatories, all of which have effects that go beyond suppressing fevers.

What infection is occurring?  A response that has evolved as a response to the infections found on the plains of East Africa could be counter productive against diseases of modernity like Pseudomonas sepsis in the ICU or MRSA endocarditis in a heroin user.

All that glitters is not gold, all who wander are not lost, and not all who are febrile are infected.  Many pathologic processes will increase temperature, so perhaps it is non-infectious fevers that require antipyretics.

And who is having the fever? An otherwise healthy child? A adult with poor physiologic reserve from congestive heart failure or severe emphysema? Or someone with a new stroke or heart attack?

Which of those factors is involved makes the literature on treating a fever less clear cut, and there is the question as to whether treating fever is of benefit, causes harm, or does nothing. Or are the outcomes even clinically relevant? When you read the literature, depending on the population studied, you can find all manner of interesting consequence of treating fevers.

ICU Fevers

A  review suggests

Observational studies on ICU populations have reported associations between fever magnitude and patient mortality. Especially recent findings indicated that infected patients may significantly benefit from temperature elevation, while high fever may be maladaptive for non-infected ones. Aggressive antipyretic treatment of ICU patients has not been followed by decreased mortality in randomized trials. However, fever suppression and return to normothermia improved outcomes of septic shock patients. ”

Aggressive fever treatment with acetomenaphen lowers temperature by a whopping 0.2 degrees C compared with routine fever management. No harm to the patient as a result but another suggests aggressive acetaminophen led to increased mortality in a surgical ICU

And the increase in mortality is a more consistent finding:

For aggressive versus permissive antipyretic treatments, a reduction in mean daily temperatures favoured the aggressive group (MD, -1.09, 95% CI -1.37, -0.81, P<0.001) with a trend towards higher mortality for aggressive treatment (RR, 6.05, 95% CI 0.78, 46.95, P=0.09).”

But it may depend of whether the fever is associated with sepsis or not:

In non-septic patients, high fever (≥39.5°C) independently associated with mortality, without association of administration of NSAIDs or acetaminophen with mortality. In contrast, in septic patients, administration of NSAIDs or acetaminophen independently associated with 28-day mortality, without association of fever with mortality.

As in another study post op fevers worsened outcome in the icu, but

there was a trend toward a protective effect from an infectious etiology of fever.

And whether the fever was prolonged or not:

The mortality in the group with prolonged fever was 62.5% (10/16) compared to 29.6% (16/54) in patients with fever of less than 5 days’ duration, a highly significant difference (p < 0.0001).

Treat a fever in the ICU?  Depends. Probably if non-infectious but the effect is not clear cut given the heterogeneity of the patients studied.

Influenza and fevers

In animal models, treatment with antipyretics for influenza infection increases the risk of mortality. There are no randomized placebo-controlled trials of antipyretic use in influenza infection in humans that reported data on mortality and a paucity of clinical data by which to assess their efficacy.

Although there some data to suggest antipyretics prolongs influenza A.

It has been suggested that aspirin was partly responsible for the increased deaths in the 1918 pandemic, although not due to the antipyretic effect:

In 1918, the US Surgeon General, the US Navy, and the Journal of the American Medical Association recommended use of aspirin just before the October death spike. If these recommendations were followed, and if pulmonary edema occurred in 3% of persons, a significant proportion of the deaths may be attributable to aspirin.

The multifactorial effects are one of the problems with pharmacologic interventions. NSAIDs, for example, may make pneumonias worse:

Of the 90 patients included, 32 (36%) had taken NSAIDs prior to hospital referral. Compared with nonexposed patients, they were younger and had fewer comorbidities but similar severity of disease at presentation, despite a longer duration of symptoms before referral. However, they more often developed pleuropulmonary complications, such as pleural empyema and lung cavitation (37.5% vs 7%; P = .0009), and had a trend to more-invasive disease, with a higher frequency of pleural empyema (25% vs 5%, P = .014) and bacteremia, especially in those not having received concomitant antibiotics (69% vs 27%, P = .009).

And aspirin increases mortality in S. pneumoniae infections:

A twofold increased risk of mortality was found with aspirin treatment in animal models of S pneumoniae infection. No relevant human studies were identified

But whether it is due to the effects on fever or other parts of the inflammatory response is unknown.

Antipyretics may make the common cold worse

Sixty healthy volunteers were challenged intranasally with rhinovirus type 2 and randomized to one of four treatment arms: aspirin, acetaminophen, ibuprofen, or placebo. Fifty-six volunteers were successfully infected and shed virus on at least 4 days after challenge. Virus shedding, antibody levels, clinical symptoms and signs, and blood leukocyte levels were carefully monitored. Use of aspirin and acetaminophen was associated with suppression of serum neutralizing antibody response (P less than .05 vs. placebo) and increased nasal symptoms and signs (P less than .05 vs. placebo). A concomitant rise in circulating monocytes suggested that the suppression of antibody response may be mediated through drug effects on monocytes and/or mononuclear phagocytes. There were no significant differences in viral shedding among the four groups, but a trend toward longer duration of virus shedding was observed in the aspirin and acetaminophen groups.

And aspirin prolongs shedding of rhinovirus, although by what mechanism is uncertain.

Fevers and children

Certainly the use of antipyretics can help resolve fever and make the child more comfortable, but is it of benefit in making the child better faster? Probably not.  As an example, acetaminophen does not make chickenpox symptoms better and may prolong the disease.

I would advise against this naturopathic treatment of fever unless you want to make sure you sleep alone:

Onions sliced and placed in wool or cotton socks and applied to the feet. It also recommended sliced onions in a bowl by the bed. It is believed that the onions have the capacity to draw out the fever.

This is by no means a comprehensive review of the risk and benefits of treating fever. Outside of comfort to the patient I cannot find a consistant benefit of treating infectious fevers. The preponderance of information suggests treating infectious fevers is almost always detrimental. Whether to treat fever; with chilliness; from putting the hands out of bed or fever; after sexual intercourse, there are no reliable clinical trials for guidance.

While there may be non-infectious reasons to treat a fever, stroke  and CNS injury as examples, all the information suggests that treating a fever is probably counterproductive with the caveat in the ICU that the hemodynamic effects of fever may be worse than the potential complications of antipyretics. As in all acute illness, a careful consideration of the risks and benefits of treating a given patients is required, not the automatic attempt to keep everyone euthermic.

That being said, when my boys had fevers as a child and it was my turn to stay home with them, I did not treat the fevers. I think the bulk of data suggests treating fevers is a bad idea and the fevers has the added benefit that the kids were quiet and I could get some work done. Treat the fever and they become too active. Then my wife would take her shift and the kids would get Tylenol.

Peu d’hommes ont esté admirés par leurs domestiques 

Not that my wife is my domestique, but you get my meaning.


*You can fill in the noun of what ever group you most think behaves like a reptile, says the juvenile emotionally immature hateful left wing extremist, or so I have been referred to on iTunes.



Posted in: Basic Science, Evolution, History, Science and Medicine

Leave a Comment (51) ↓