pdnews@elsevier.com

Ask about fever duration during your history taking. The main question is how to define “persistent.” If the fever lasts fewer than 5 days and everything else looks fine, most of the time the child will have whatever illness is going around.

If the fever lasts closer to 10 days, and especially if new symptoms appear, it could be a complication of what is going around. Obviously, once you go beyond 10 days, it is a more significant clinical situation. Then we are in the “fever of unknown origin” (FUO) range, which may require a subspecialist evaluation. The current official definition of FUO is fever lasting greater than 3 weeks with no diagnosis after reasonable outpatient or inpatient evaluation. Shorter episodes of unexplained fever are referred to as “fever without a source.” Most of these will resolve spontaneously or evolve into more easily recognized conditions, usually infections.

The best way to proceed really depends on the presentation. The initial evaluation always includes a careful history, physical examination, and screening labs as needed. Start a more thorough evaluation or refer when things are not adding up. How sick does the child look? Are there atypical findings? Remember your training about the typical course of strep throat, mono, or bronchitis—if you see a significant deviation, evaluate the child more thoroughly. Always ask: Do I have a reasonable working diagnosis that I am comfortable with?

Fever is a common symptom and most often is a feature of a routine viral infection. The bigger concern is a serious condition, such as a bloodstream infection; this child needs to get to the emergency room right away. A child with an unusual rash, especially with petechiae and purpura, would need an immediate referral.

Kawasaki disease is another major concern. If you suspect Kawasaki disease, the child needs to be evaluated by a subspecialist. The diagnosis is based on a fever lasting at least 4 or 5 days and associated findings, such as a rash, red strawberry tongue, or very red lips. The patient will need immediate treatment, ideally within 10 days, to minimize the risk for cardiac and coronary inflammation.

Helpful tests include a complete blood count, sedimentation rate, and C-reactive protein assay. Serum chemistries including liver function tests also may be useful. I do a urinalysis and a urine culture because urinary tract infections are common and a common cause of fever. Blood cultures should be considered.

These screening tests can be reassuring. When the clinical evaluation is benign, the white blood cell count is 5,000 cells/mcL with a normal differential; then you can tell the parent to come back in a few days for follow-up. In contrast, if a child with a persistent fever has a white count of 30,000, you really have to be more careful—it could be a sinus infection, pneumonia, or bacteremia. The other extreme, a child with a white blood count of 1,000, also requires more careful evaluation.

The pace of evaluation depends on how ill the child is, any associated findings, and whether one has a diagnosis or not. Time is your ally when the child is not very ill; watchful waiting will often reveal the nature of the problem.

It is important to take a good travel history. We saw a child with malaria last year, and the key to recognition was the history of a recent return from Ghana. Also ask about animal exposures and if anyone else at home is sick. Those can be important clinical clues for diagnosis of a child with prolonged fever.

Consider the time of year. A test that is not useful, but is often obtained, is a Lyme disease test. Lyme is rarely a cause of prolonged fever, especially in winter. Also, sometimes an extensive evaluation for autoimmune disease is performed too early in the diagnostic process. If you do an initial evaluation and do not have a diagnosis, and the fever persists, then you can move on to secondary and tertiary evaluations, such as you would with an FUO. Also, tests for mono are ordered way too often. The diagnosis of mono depends on more than a prolonged fever, and there are some very specific associated features, such as tonsils that look awful, large cervical nodes, and a palpable spleen.

Make sure the fever is real. Sometimes we see a child who reports persistent fever but is afebrile during each clinical visit. In this case, ask the parents how they take the child's temperature. Do they use a thermometer strip, or do they just touch the child and say the child feels warm? There should be some documentation of a persistent fever before you embark on additional, expensive evaluation.

Occasionally I have a child come to see me who is referred with say, 25 days of fever. Does the child really have one prolonged fever? It is more likely two different episodes—a child has illness No. 1, then a break, followed by illness No. 2. Use the history and clinical findings to distinguish between these two scenarios.

The Child With Persistent Fever