When it's not a fever

When you see a patient with a temperature of 101, what do you note in the chart? Probably, “fever,” right? Maybe not.

When you see a patient with a temperature of 101, what do you note in the chart? Probably “fever,” right?

Not if you've ever discussed the subject with George H. Karam, FACP, infectious disease specialist and professor of medicine at Louisiana State University in Baton Rouge. Internal Medicine 2011 attendees had the chance to learn from Dr. Karam about fever and common misunderstandings of it—including one that indirectly led to the restructuring of graduate medical education—during a precourse on critical care.

“When we write fever as a diagnosis, we often can't be sure that it's truly fever,” said Dr. Karam. “I would like to see you move away from using the word ‘fever’ and begin to use the words ‘elevated temperature’.”

Elevated temperature allows for the possibility that the problem is not fever, but hyperthermia, resulting from the body either overproducing or failing to dissipate heat. “Fever, on the other hand, is going to be resetting of the hypothalamic thermoregulatory center,” Dr. Karam noted. “The dilemma for the clinician is that today we do not have a way that we can distinguish with a laboratory assay one of these diagnoses from the other. It is all going to be a judgment issue.”


To help physicians make these judgments, he reviewed the most likely causes of either diagnosis. “The list of causes of hyperthermia is very well known to you,” said Dr. Karam, listing off exertional hyperthermia, heat exhaustion and heat stroke. Eclampsia/preeclampsia is another possibility, one made trickier to diagnose by the fact that placental ischemia can cause fever as well.

Certain drugs can also lead to hyperthermia. Dr. Karam cited an article from the June 2010 Critical Care Medicine that described seven categories of drug-induced hyperthermia: adrenergic, anticholinergic, antidopaminergic, uncoupling of oxidative phosphorylation, baclofen withdrawal, malignant hyperthermia and serotonin syndrome.

The last of those lay behind the misdiagnosis of elevated temperature that changed graduate medical education forever. In 1984, Libby Zion was diagnosed with a viral syndrome by an intern and resident at New York Hospital. After her death, the 80-hour work rule was instituted in New York, eventually to be followed by national regulations.

“If you really look at what happened in this case, it was an example of the serotonin syndrome in the time before the serotonin syndrome was described,” said Dr. Karam. And although training schedules changed, lack of knowledge about serotonin syndrome has persisted, with 85% of physicians unaware of the diagnosis, according to a study published in 1999. “Very often what we find is that our actions are on the basis of lack of knowledge, not so much fatigue or lack of supervision,” he said.

So he offered some knowledge. Serotonin syndrome is identified by a clinical triad: neuromuscular abnormalities (including tremor, rigidity, clonus and hyperreflexia), mental status changes and autonomic hyperactivity. “Remember that serotonin is going to have effects on the GI tract, so very often the patient will present initially with some diarrhea,” Dr. Karam said.

It can be caused by drugs that reduce serotonin uptake (SSRIs), increase serotonin release (amphetamines, cocaine), cause an excess of serotonin precursors (buspirone, L-dopa) or slow serotonin metabolism (MAO inhibitors). Therapy includes benzodiazepines for symptoms of agitation and drugs that antagonize the effects of serotonin, such as cyproheptadine.

Serotonin syndrome can look similar to neuroleptic malignant syndrome (NMS), but Dr. Karam offered a tip for distinguishing them based on the effects on muscle. “In contrast to serotonin syndrome, where you have hyperflexia, in NMS, it's going to be hypoflexia.” The cause of neuroleptic malignant syndrome may be a genetic defect in calcium regulation, but it is triggered by blocking dopamine receptors (as haloperidol does) or withdrawing dopamine agonists (such as antiparkinsonian drugs).

Malignant hyperthermia is also caused by a genetic calcium problem, specifically excessive release of calcium from skeletal muscle in response to anesthetic agents. “The classic thing is that it happens in the operating room, but there are reports of malignant hyperthermia as late as 12 hours after an operation is completed. So excessive elevations in temperature ought to bring this to mind,” said Dr. Karam.

Both neuroleptic malignant syndrome and malignant hyperthermia are treated by cessation of the triggering drug, cooling, supportive measures and dantrolene sodium. Although its use in NMS has not been proven in trials, dantrolene sodium has shown dramatic success with malignant hyperthermia, Dr. Karam noted. “Mortality rates go from over 70% to less than 5%.” Treatment for NMS can also include bromocriptine mesylate.

Temperature elevations after surgery may also indicate hyperthermia in neurosurgery patients, especially those with autonomic impairment, cerebrovascular accidents and spinal cord injuries. “The dilemma is there are multiple risks for infection in these people, so we often give them antibiotics because we think they have fever. When they don't respond to antibiotics, ask the question, ‘Might this really be some type of hyperthermic condition?’”

What about the patient who responds to antibiotics too fast, for example, a nursing home resident with an elevated temperature and bacteriuria whose temperature normalizes two hours after antibiotics are given? “As an infectious disease doctor, I would love to tell you that antibiotics cure people that fast,” Dr. Karam said.

The truth is, though, that it's probably the fluids the patient received with his antibiotics that made the difference. “The most common reason for hyperthermia in hospitalized patients is going to be volume contraction,” he said. Awareness of this phenomenon could help reduce overuse of antibiotics, Dr. Karam suggested.


Even in cases when hyperthermia can be ruled out as the cause of a patient's elevated temperature, it may still not be time to break out the antibiotics. “Whenever you see people in the ICU who truly meet the criteria for fever, keep in mind infectious versus noninfectious,” said Dr. Karam.

One noninfectious cause of fever is infarcted tissue, such as a myocardial infarction, pulmonary embolism or ischemic bowel. “If you infarct something, and cells die, those cells are going to release inflammatory mediators,” said Dr. Karam. The same process could also occur in a patient who is hemorrhaging.

Fever has also been found to be associated with thromboembolic disease. One study found that 14% of people with pulmonary embolism had an elevated temperature, Dr. Karam reported.

Elevated temperature could also result from inflammatory response in a number of conditions: chemical pneumonitis, pancreatitis, hepatitis and abdominal aortic aneurysm. Syndromes associated with inflammation, such as acute respiratory distress syndrome (ARDS), systemic inflammatory response syndrome (SIRS) and postcardiac injury syndrome (PCIS), can present with fever.

Elevated temperatures may resolve on their own in some of these cases, Dr. Karam said. “Understanding the entity may be the impetus to not give overly prolonged durations of antibiotics.”

Fever can also be caused by immune complex diseases, which at least are consistent in their features. “It's a fairly classic presentation. Immune complexes love to go to skin, they love to go to joint, they love to go to kidney,” said Dr. Karam. An example is acute interstitial nephritis, which could be caused by beta-lactam antibiotics, NSAIDs or cimetidine.

There is one complication to identifying these diseases, though. “More recently, people have said there are some diseases that don't truly have immune complexes but they look exactly like immune complexes,” Dr. Karam said. These autoinflammatory diseases are best exemplified by familial Mediterranean fever, he added.

Two final noninfectious causes for fever: blood transfusions and drugs. “I really wish we had a way to distinguish drug-induced fever,” said Dr. Karam, noting that as few as 20% of patients present with a rash.

It's often a diagnosis of exclusion, and one that should not be excluded itself until the patient has been off the drug for at least seven days. “Realize that normal clearance is supposed to be three or four days but it may take up to a week to clear, so you've not excluded drug fever simply because it didn't get better quickly,” he said.

Quick improvement is possible, however, in physicians' approach to diagnosing these patients, Dr. Karam concluded. “If this talk resonated with you, the way that you will know is that the next time you write a problem list, you will write elevated temperature, not fever.”