Symptom-Based Diagnosis in Pediatrics (CHOP Morning Report) 1st Ed.

CASE 11-5

Six-Year-Old Boy




A 6-year-old boy presented with a 2-week history of low-grade fevers. One week prior to admission, he had an episode of nonbilious emesis and subsequently began complaining of neck pain. There was no cough, diarrhea, rash, or abdominal or joint pain. Although he never complained that light bothered his eyes, he did not play outside with his friends during the day. His mother initially attributed this to the summer heat. He received several doses of ibuprofen for complaints of headache. He was brought for evaluation after the mother learned of a meningococcal meningitis outbreak at local school from the evening news broadcast. She became concerned about the possibility of meningitis and, after speaking with her pediatrician, rushed the child to the emergency department. There were no ill contacts. Neither the child nor his two siblings attended the school where several children had developed meningitis.


The child had an unremarkable birth history. At the age of 8 months he required hospitalization for management of rotavirus gastroenteritis-induced dehydration. Two months before admission, he was bitten on the hand while feeding deer during a hiking trip in the Pocono Mountains in Pennsylvania. Three weeks before admission, he developed a severe contact dermatitis on his arms and face that was attributed to poison ivy. He recovered uneventfully after treatment with oral antihista-mines and cool compresses. The family history was unremarkable. The child lived with his parents and two brothers in Southern New Jersey.


T 38.1°C; HR 100 bpm; RR 28/min; BP 101/53; mmHg; SpO2 100% in room air

Weight 50th percentile for age

In general, the child was a lean but healthy-appearing boy. He had mild photophobia. There was no Kernig or Brudzinski sign, but there was difficulty with terminal neck flexion. There was no cervical lymphadenopathy. The heart and lung sounds were normal. There was no hepatomegaly or splenomegaly. The cranial nerve examination was normal. The skin examination revealed a rash that had appeared recently. The rash suggested the diagnosis (Figure 11-5).


FIGURE 11-5. Photograph of patient’s rash.


Complete blood count revealed the following: 8600 WBCs/mm3 (71% segmented neutrophils, 22% lymphocytes, and 7% monocytes); hemoglobin, 11.1 g/dL; and 461 000 platelets/mm3. Serum glucose was 96 mg/dL. Lumbar puncture revealed 21 WBCs (3% segmented neutrophils, 77% lymphocytes, and 20% monocytes) and 1 RBC/mm3. The CSF protein and glucose were 23 mg/dL and 63 mg/dL, respectively. No bacteria were noted on the CSF Gram stain. The CSF and blood cultures were subsequently negative.


The diagnosis suggested by the rash was confirmed by additional testing.



Aseptic meningitis refers to a syndrome of meningeal inflammation without evidence of pathogens by traditional bacterial culture methods. A specific cause is identified in fewer than 60% of cases. Enteroviruses, including echoviruses, cox-sackie viruses, and numbered enteroviruses, are the most common cause of aseptic meningitis. They account for up to 95% of cases when a specific pathogen is implicated. In the summer, other infectious causes of aseptic meningitis include Lyme disease (Borrelia burgdorferi), Rocky Mountain spotted fever (Rickettsia rickettsii), anaplasmosis, and arboviruses such as West Nile virus, St. Louis encephalitis virus, and Eastern and Western equine encephalitis viruses. Some parainfluenza viruses occur throughout the year, including summer. Other viruses include herpes simplex virus, varicella zoster virus, Epstein-Barr virus, and human herpesvirus 6. Less common causes of aseptic meningitis include tuberculosis, syphilis, various fungi (e.g., Cryptococcus neoformans), and parasites. Noninfectious causes include Kawasaki disease, systemic lupus erythematosus, and poly-arteritis nodosum.

This relatively well-appearing child presented with subacute symptoms and a CSF pleocytosis. The CSF WBC count does not point to a particular etiology. Although tuberculous meningitis is important to consider in any patient with aseptic meningitis, it usually manifests with characteristic CSF findings including a dramatically elevated protein and a low glucose concentration—findings not present in this case. Although a rash is associated with several of the above-mentioned conditions, children with Rocky Mountain spotted fever, anaplasmosis, herpes simplex virus, varicella, or syphilis are often quite ill. Additionally, children with Rocky Mountain spotted fever and anaplasmosis typically have leukopenia and thrombo-cytopenia. The rash of herpes simplex virus and varicella, when present, is characteristic and does not resemble the rash on this patient. The CSF WBC differential with a mononuclear cell predominance (i.e., predominance of lymphocytes and monocytes rather than neutrophils) argues against a bacterial process. The rash shown in Figure 11-5 was characteristic of early disseminated Lyme disease.


Examination of the skin revealed multiple annular lesions on the chest, back, and legs, ranging from 5 to 15 cm in diameter (Figure 11-5). These macular erythematous lesions with partial to complete central clearing were characteristic of erythema migrans (EM). This finding of multiple EM lesions combined with headache, photophobia, and mononuclear CSF pleocytosis suggested the diagnosis of Lyme meningitis. Serum antibodies revealed an IgM titer of 27.3 (reference, 0-0.8) and an IgG titer of 0.4 (reference, 0-0.8). This positive IgM test was also confirmed by Western blotting. CSF enterovirus PCR and bacterial culture were negative. An electrocardiogram did not reveal evidence of heart block, a feature associated with early disseminated Lyme disease. The child was treated with intravenous ceftriaxone for 3 weeks and recovered completely.


Lyme disease, caused by the tick-borne spiro-chete, B. burgdorfei, was initially identified during investigation of a cluster of children with arthritis in Lyme, Connecticut. Features of the disease had been previously described in Europe under various names including erythema chronicum migrans, acrodermatitis chronica atrophicans, and Bannwarth syndrome. Lyme disease is now endemic in more than 15 states. The infection is also common in middle Europe, Scandinavia, and parts of Russia, China, and Japan. Several closely related ticks that are part of the Ixodes ricinus complex (Ixodes scapularis, Ixodes pacificus, Ixodes ricinus, and Ixodes persulcatus) comprise the vectors of Lyme disease. They vary in their geographic distribution. In the United States, I. scapularis is the vector in the northeast and midwest, while I. pacificus predominates on the west coast. New England, the mid-Atlantic States, Minnesota, and Wisconsin have the highest prevalence of infected ticks. Lyme disease has a bimodal age distribution, with the greater number of cases occurring between the ages of 5 and 9 years and between the ages of 45 and 59 years.

Most infections occur between May and July. After injection of B. burgdorferi by the Ixodes tick into the skin, the spirochete multiplies locally at the site of the bite and within days to weeks may disseminate to other sites. The risk of Lyme disease after a tick bite will be discussed later.


Lyme disease typically manifests in three stages: localized infection, early (disseminated) infection, and late infection. Localized infection generally occurs 7-14 days (range, 3-32 days) after the tick bite. Early infection occurs 2-6 weeks after the tick bite, and late infection occurs 6-12 weeks (range, up to 12 months) after the tick bite. Localized infection refers to development of EM lesions at the site of the tick bite. It usually begins as an erythematous macule or papule and expands to a median diameter of 15 cm, with intensely erythematous macular borders and central clearing or induration. Occasionally, the central area becomes vesicular or necrotic. Complaints such as malaise, headache, arthralgias, myalgias, fever, and regional lymphadenopathy may accompany the lesion. Most children (60%-70%) present with the localized form of Lyme disease.

Features of early disseminated infection include multiple EM lesions (23% of all Lyme disease cases), carditis (0.5%), cranial nerve palsy (3%), meningitis (2%), and acute radiculopathy (<0.5%). The most common cranial nerve palsy involves the seventh (facial) nerve but palsy of cranial nerves III, IV, or VI may develop. Most children with Lyme meningitis have symptoms for 2-3 weeks before the diagnosis. They initially develop low-grade fevers, followed by mild neck pain or stiffness with extreme flexion. In some patients with Lyme meningitis, EM rash or concomitant cranial nerve palsy, rather than neck pain, brings the child to medical attention, and the findings of EM, cranial nerve palsy, or papilledema are helpful in distinguishing Lyme meningitis from viral meningitis. Cardiac manifestations include fluctuating degrees of atrioventricular block (first degree, Wenckebach, or complete) and, less commonly, myocarditis, pericarditis, or cardiomegaly. Earlier reports suggested that carditis may occur in 5% of untreated patients with Lyme disease, but more recent studies have found a much lower incidence.

Arthritis is the most common late manifestation of Lyme disease, occurring in up to 10% of untreated patients. Children are more likely than adults to develop Lyme arthritis and are also more likely to have arthritis as their presenting complaint. It involves the knee in 90% of cases. The spirochete occasionally affects the hip, ankle, wrist, elbow, or temporomandibular joint. The joint is swollen and warm but, in contrast to bacterial septic arthritis, only mildly tender. If left untreated, Lyme arthritis is often intermittent, with episodes of inflammation lasting weeks to months. Signs and symptoms of systemic illness are rare.


The diagnosis is usually suspected on characteristic clinical findings, history of exposure in an endemic area, and antibody testing. Diagnosis of Lyme meningitis requires a high level of suspicion, careful history and examination, serum antibody studies, and lumbar puncture. Diagnosis of other manifestations of Lyme disease is discussed later.

Serologic diagnosis. For serologic testing, the U.S. Centers for Disease Control and Prevention recommends a two-tiered approach to testing. An enzyme-linked immunoassay (EIA) is performed initially. The EIA test is very sensitive but false-positive results occur due to cross-reactive antibodies in patients with other spirochetal infections or certain viral infections or autoimmune diseases. Those with equivocal or positive results by EIA should undergo confirmatory testing by Western blot of the initial sample. An IgM determination is considered positive if two of three specific bands (23, 39, or 41 kd) are detected by Western blotting. An IgG Western blot is considered positive if at least 5 of 10 specific bands are present.

Only 40% of children with localized EM have detectable antibodies. On repeat testing 4 weeks after detection of the EM lesion, 70% will have detectable antibodies, thereby suggesting that early antibiotic treatment of the EM lesion may blunt antibody response. In children with early disseminated infection such as Lyme meningitis, IgM antibodies are present in more than 95% of cases and IgG antibodies are present in 70%. Children with late infection should have detectable IgG and may have detectable IgM.

Peripheral blood smear. As Ixodes is a vector for multiple infective agents, patients with Lyme disease frequently are coinfected with another tick-borne pathogen. In a study by Krause et al., 22% of patients with Lyme disease were coinfected with basbesiosis, and 4% were coinfected with human granulocytic anaplasmosis. The diagnosis of babesiosis depends on microscopic identification of intraerythrocytic parasites in Giemsa- or Wright-stained thick and thin blood smears. In approximately 50% of patients with anaplasmosis, morulae are detected in peripheral blood neutrophils. Peripheral blood smears should be examined in any patient with Lyme disease who has continued symptoms despite appropriate therapy. These coinfecting pathogens should also be considered in any patient with a more severe presentation than is typically observed with Lyme disease alone or a patient presenting with accompanying leukopenia, thrombocytopenia, anemia, hemolysis, or jaundice—findings that are more typical of other tickborne diseases rather than Lyme disease.

Lumbar puncture. In children with Lyme meningitis, CSF analysis reveals a mild mononuclear (lymphocytes and moncoytes) pleocytosis (mean, 100 WBCs/mm3; range, 10-500/mm3). In a study by Turnquist et al., 23 of 24 children with Lyme meningitis had less than 10% segmented neutrophils in the CSF. The CSF glucose is normal and the protein may be normal or mildly elevated. CSF antibody studies may provide supporting evidence of Lyme meningitis. CSF Lyme PCR is generally not useful in diagnosing Lyme meningitis due to low bacterial counts in the CSF and subsequent poor sensitivity. There have been recent improvements in PCR methods for identification of B. burgdorferi DNA but many laboratories are not capable of accurately performing the test, and its clinical utility remains limited.

Joint aspiration. In cases of Lyme arthritis, joint aspiration typically reveals between 30 000 and 50 000 WBCs/mm3, often with neutrophil predominance. Results of Lyme PCR testing of joint fluid are positive in approximately 85% of patients with suspected Lyme arthritis. False-positive PCR results from joint fluid are uncommon.

Electrocardiogram. An electrocardiogram should be performed in patients with signs and symptoms of cardiac disease, and should be considered in patients with concern for Lyme arthritis. Approximately one-third of children with Lyme meningitis will have ECG abnormalities; heart block and prolongation of the corrected QT interval are most common. ECG abnormalities occur in 50% of children with Lyme meningitis who have fever for 5 or more days and those who are 13 years of age or older at the time of diagnosis; the probability of ECG abnormalities was 83% (95% confidence interval: 50%-96%) in children with both of these findings. As most ECG abnormalities will resolve without consequence, the relevance of detecting these abnormalities is uncertain.

Other studies. There is no role for Lyme PCR testing of urine or blood samples. Because clinically it may be difficult to distinguish Lyme meningitis from other causes of aseptic meningitis, studies to detect other causes (e.g., enteroviral CSF PCR, serum antibodies, and CSF PCR for arboviruses) should be considered. Results of additional studies such as peripheral WBC count and erythrocyte sedimentation rate are relatively nonspecific for Lyme disease, but certain findings (e.g., leukopenia) can suggest an alternate or concomitant infection. If coinfection with Babesia or anaplasmosis is suspected, serologic testing may be indicated, as antibody detection is more sensitive than blood smear in the identification of these conditions.


Treatment of Lyme disease depends on the clinical manifestation (Table 11-5). Localized disease may be treated with amoxicillin or cefuroxime for 14-21 days, or if the patient is 8 years of age or older, doxycycline for 10-21 days. For patients intolerant to these agents, macrolides should be considered, although they have been shown to be less effective. Some controversy exists as to the appropriate management of children with facial nerve palsy. Although some experts recommend CSF evaluation on all patients with a Lyme-associated cranial palsy, others reserve lumbar puncture for those with clinical evidence of CNS infection (e.g., headache, neck stiffness, or photophobia). An oral regimen, as discussed, is reasonable to treat isolated cranial nerve palsy due to Lyme disease.

TABLE 11-5. Manifestations of Lyme disease.


Children with carditis may be treated either orally or parenterally depending on the severity of cardiac involvement. Children with first-degree atrioventricular heart block may be treated with oral antibiotics. Hospitalization and parenteral antibiotic therapy is recommended for children with second- or third-degree atrioventricular heart block, for those with first-degree heart block in whom the PR interval is prolonged to 30 milliseconds or longer, and for those with symptoms (e.g., syncope, dyspnea, chest pain). A 14-day course (range, 10-28 days) of ceftriaxone is the preferred parenteral regimen, but administration of cefotaxime and penicillin G are reasonable alternatives. These children are often transitioned to oral antibiotics for completion of their treatment course. This decision is frequently made in conjunction with an infectious diseases specialist.

Children with evidence of Lyme meningitis or radiculopathy should receive parenteral therapy for 14 days (range, 10-28 days), as described previously. Children with Lyme arthritis often respond well to 28 days of oral therapy. For those with poor initial response to oral therapy, parenteral ceftriaxone should be considered. Children with Lyme arthritis may have recurrence of arthritis. Recurrences of arthritis after successful initial treatment can be management with nonsteroidal antiinflammatory agents.

Although early studies of the efficacy of prophylaxis after tick bites failed to show a protective effect, a more recent study of adults showed that a single 200-mg dose of doxycycline administered within 72 hours after a recognized I. scapularis bite had an efficacy of 87% in preventing EM when compared to placebo. However, adverse events, including nausea and vomiting, occurred in approximately 30% who received doxycycline. Given these findings, routine administration of antimicrobial prophylaxis or serologic testing following a tick bite is not recommended. However, a single dose of doxycycline may be offered to patients 8 years of age or older if each of the following criteria is met: the attached tick can be reliability identified as an I. scapularis, it is estimated to have been attached for 36 hours or longer, based on exposure history and the degree of tick engorgement, prophylactic therapy can be initiated within 72 hours of tick removal, ecologic information suggests that the local B. burgdorferi infection rate is 20% or greater among these ticks, and the patient has no contraindications to doxycycline. In those who cannot receive doxycycline, no prophylactic therapy is recommended, as the benefit of other antibiotics is not known.


1. Esposito S, Bosis S, Sabatini C, Tagliaferri L, Principi N. Borrelia burgdorferi infection and Lyme disease in children. Int J Infect Dis. 2013;17:e153-e158.

2. Feder HM. Lyme disease in children. Infect Dis Clin N Am. 2008;22:315-326.

3. Gerber MA, Zemel LS, Shapiro ED. Lyme arthritis in children: clinical epidemiology and long-term outcomes. Pediatr. 1998;102:905-908.

4. Gerber MA, Shapiro ED, Burke GS, et al. Lyme disease in children in Southeastern Connecticut. N Engl J Med. 1996;335:1270-1274.

5. Hayes EB, Piesman J. How can we prevent Lyme disease? N Engl J Med. 2003;348:2424-2430.

6. Krause PJ, McKay K, Thompson CA, et al. Disease-specific diagnosis of coinfecting tickborne zoonoses: babesiosis, human granulocytic ehrlichiosis, and Lyme disease. Clin Infect Dis. 2002;34:1184-1191.

7. Nadelman RB, Nowakowski J, Fish D, et al. Prophylaxis with single-dose doxycycline for the prevention of Lyme disease after an Ixodes scapularis tick bite. N Engl J Med. 2001;345:79-84.

8. Newland JG, Zaoutis TE, Shah SS. The child with aseptic meningitis. Pediatr Case Rev. 2003;3(4):218-221.

9. Stanek G, Wormser GP, Gray J, Strle F. Lyme borreliosis. Lancet. 2012;379(9814):461-473.

10. Steere AC. Lyme disease. N Engl J Med. 2001;345: 115-125.

11. Turnquist JL, Shah SS, Zaoutis TE, Hodinka RL, Coffin SE. Clinical and laboratory features allowing early differentiation of Lyme and enteroviral meningitis in children. Pediatr Res. 2003;53:106A.

12. Welsh EJ, Cohn KA, Nigrovic LE, et al. Electrocardiograph abnormalities in children with Lyme meningitis. J Pediatr Infect Dis Soc. 2012;1:293-298.

13. Wormser GP, Dattwyler RJ, Shapiro ED, et al. The clinical assessment, treatment, and prevention of Lyme disease, human granulocytic anaplasmosis, and babe-siosis. Clin Infect Dis. 2006;43:1089-1134.