There is a fascinating piece on fevers in Science-Based Medicine by Mark Crislip:
As I have mentioned in the past, almost all of my practice is inpatient medicine, doing infectious disease consults in acute care hospitals. I only spend three hours a week in the outpatient clinic, so I have a skewed perception of medicine and disease. The patients I see are sick, really sick, often trying to die and are a complicated collection of abnormal labs and deranged physiology.
I remember finishing residency thinking that a potassium of 2.8, a hemoglobin of 9.8 or a bilirubin of 4.5 wasn’t all that bad, losing track of normal physiology amongst all the medical pathology. I never did lose track of normal vital signs (VS): pulse, respiration, blood pressure and temperature. Like trying to be the fifth Beatle, over the years other values have vied to become the fifth vital sign: pain level or O2 saturation, but none have the importance of the fab four. I can live without pain*, but I can’t live long if the other vital signs are abnormal for extended periods of time. Watching the vital signs return to normal is often an important variable that signifies the patient is improving.
“Fever is a mighty engine which Nature brings into the world for conquest of her enemies.” –Thomas Sydenham 1666.
Of course I have an inordinate interest in fevers: their cause, their patterns and their treatment. Fevers lead to consults and while I say my job is ‘me find bug, me kill bug, me go home’, more often it is ‘me find cause of fever, me treat cause fever, fever goes away, me go home’.
One aspect of fever I harp on year after year and where I am continually ignored is the importance of not treating a fever. It is estimated that the fever response is 400 million years old. How do they know that? Got me. Most molecular techniques are “sufficiently advanced technology indistinguishable” from magic; all I know is that they were not measuring core body in T. rex. Every creature that can make a fever will make a fever when infected. All branches of the immune system function better at 102 than 98.2 (yes, 98.2), but in the calorie poor environment most creatures live in, if we maintained our core temp at 102 we would all starve to death. It is also quite remarkable how many potential pathogens cannot grow at 98.2, much less 102. Being above ambient temperature protects against thousands of molds and bacteria.
Almost every animal and human study demonstrates that outcomes are worse if you treat a fever: increase in mortality and/or complications, although it is not always clear if it is the anti-inflammatory or anti-pyretic effects of medications being used that lead to the poorer outcomes. You cannot find studies to demonstrate benefit in treating infections from suppressing fevers.
There are times when you may want to treat a fever: the patient does not have the physiologic reserve from cardiac or pulmonary disease to tolerate the metabolic stress, or they have had a stroke or heart attack or the fevers are high enough to cause damage. In the hospital there are multiple factors that should be considered before whipping out the acetaminophen for an increased fever.
Fevers are an important, evolved response to infection and you inhibit fevers at your patients peril. If a patient had a pulse of 120 or a respiratory rate of 25, you wouldn’t slow them to normal would you? No. You would treat the underlying cause of the tachycardia or tachypnea and watch the vital signs normalize as evidence that your clinical intervention is effective. The same should be true of fevers, although I know all too well that most people expect their fevers to be treated and that no one will believe you if you suggest their fevers should be allowed to run free.