a SLACK Incorporated newspaper

What's Your Diagnosis?

A monthly case study, with treatment information and discussion to follow.

[Your turn]

September 1999

Instead of the usual diagnostic challenge, the case this month is a baby with complicated otitis media (OM) and a True-False quiz.

A 23-month-old black boy presented to the pediatric clinic for evaluation of fever of one day's duration. He had no cough, nausea, vomiting, diarrhea or skin rash. His appetite and usual level of activity were decreased. There were no known sick exposures at home or elsewhere.

His past medical history is positive for having been a twin born at 28 weeks gestation. He had respiratory distress syndrome with a short period on a ventilator, but had no pulmonary sequelae. Additionally, 16 days prior to this visit, he had undergone a tonsillectomy, adenoidectomy and insertion of tympanostomy tubes because of recurrent OM. His immunizations were up-to-date and documented.

Examination during the visit was positive only for fever of 103° F, bilateral OM, with tubes in place and a somewhat lethargic appearance, for which he received a sepsis work-up; complete blood count; urinalysis; cerebrospinal fluid analysis with cultures of CSF, blood and urine; and a chest radiograph. The only abnormality was the peripheral white blood cell count (WBC), which was 23,000 with a predominance of polys and band forms.

He was admitted for observation and treated with an oral second-generation cephalosporin for the OM with dramatic improvement in his fever and overall well being. He was discharged the next day for out-patient follow-up. The following day he came back to the clinic with the return of lethargy, decreased appetite and fever.

His blood culture, taken two days earlier, was found to be positive but no organism had yet been identified. He was immediately re-admitted, and a repeat sepsis work-up was done. This time he had cloudy CSF with 138 WBCs per cubic mm, 54% of which were polys, 20% were lymphocytes, and a protein of 187 mg/dL. The glucose was 16 mg/dL, with a serum glucose of 149 mg/dL. The Gram's stain is shown in the figure. He was empirically treated with meningitic doses of ceftriaxone (Rocephin, Roche Laboratories), a third-generation cephalosporin, and vancomycin pending culture and sensitivity results.

Which of the following are true and which are false?

1. The child most likely has meningococcal meningitis.
2. His meningitis was induced by the first lumbar puncture.
3. His meningitis most likely resulted from direct extension from the middle ear.
4. There's no scientific evidence that tonsillectomy adds any benefit to adenoidectomy and or tympanostomy tubes for prevention of recurrent OM.
5. The fact that he was partially treated accounts for the relatively low CSF pleocytosis.

Discussion

1. False. The photo shows a Gram's stain of the CSF revealing gram-positive cocci, mostly in pairs. The CSF antigen test was also positive for Streptococcus pneumoniae as well as the culture. When this organism is isolated it is very important to confirm its sensitivities, as a growing percentage of isolates are resistant or relatively resistant to penicillin. These isolates account for up to 30% to 40% in some areas. It is prudent to empirically treat bacterial meningitis with a combination of vancomycin (60 mg/kg/day divided q 6 hours) and a third generation cephalosporin pending confirmation of the organism and its sensitivities.
2. Probably false, but impossible to prove. he theoretical risk of inoculating the CSF with an organism during an LP while bacteremic should not deter performing the procedure if clinically indicated. To reduce or eliminate this possibility, some experts recommend empirically treating any febrile patient with intravenous antibiotics who has received an LP while waiting for the culture results. If this really occurs, it probably is going to happen after a traumatic tap, in which case those patients tend to get treated pending culture results anyway. It's my opinion that a clear CSF with few to no red blood cells in it has almost no chance of being inoculated with blood-borne bacteria. The decision to treat in that case can be made on other CSF or clinical findings.
3. False. Autopsy data suggests that "direct extension" rarely occurs. Concurrent OM and meningitis usually results from bacteremia seeding these areas at about the same time. While an oral antibiotic may help clear the bacteremia and improve the OM, the usual oral doses of most antibiotics will not achieve high enough levels to eradicate the organism from a deeper focus. This probably accounts for the initial improvement, which was quickly followed by a second stage of deterioration.
4. True. Because of the close proximity of the adenoids to the openings of the eustachian tubes, it makes sense that unusually large adenoids may interfere with tube function, thus encouraging the development of OM. There are data that support this theory, but there is no evidence supporting tonsillectomy as therapy for recurrent OM.
5. False. Most data on partially treated meningitis indicates that CSF protein is the only parameter that is statistically changed with brief, oral prior treatment. It makes very little difference in the other parameters, including the WBC count. However, this is a very complex and difficult area to study with many variables. The low WBC count in this case probably represents early detection. In this case the culture was positive, but in many cases of partially treated meningitis the culture will be negative. If that is the case, you are stuck with treating with broad-spectrum coverage for the entire course. Even if you knew the organism from antigen testing, you would not know the sensitivities.

This patient had an excellent outcome. I actually enrolled him in an occult bacteremia study by Jim Bass, MD, that was ultimately published in the Pediatric Infectious Disease Journal in June 1993. He had randomized to the oral therapy arm of the study. However, he had to be removed from the study due to a procedural error and poor communication, resulting in a second-generation cephalosporin treatment instead of amoxicillin/clavulanate (Augmentin, SmithKline Beecham), which was the oral agent being compared to injectable ceftriaxone. I doubt that the choice of oral antibiotic really made any difference, but we still could not use him in the results.

Would he have developed meningitis if he had received ceftriaxone? Probably not, but the hard data to prove it statistically does not exist. He certainly fulfilled the criteria for being at high risk (about 10%) for bacteremia (WBC count >15,000, fever of 103° F or more, and age 3 months to 3 years). However, most of the time the cause is due to Streptococcus pneumoniae and it is much less likely to cause meningitis than Haemophilus influenzae type b (Hib) and Neisseria meningitidis. Now, with the widespread use of conjugated Hib vaccine, the incidence of bacterial meningitis has dramatically decreased. The physician still needs to use clinical judgement to decide who to empirically treat and who can be observed without treatment. There's no cookbook solution.

Vancomycin was not really added in this case because at that point in time (late 1980s, early 1990s), resistant pneumococcus was not that big of a problem. However, today it is routinely added pending sensitivities. While still debatable by some, I would also pretreat with steroids prior to the first dose of antibiotics.

Don't forget

Influenza season is just around the corner. Identify your high-risk patients to call in for vaccine next month.

Have you been sufficiently confused by the issue of thimerosal in vaccines? This is your hint for next month's case.

Acknowledgement: Thanks to James W. Bass, MD, COL, USA (Retired) of Honolulu, Hawaii for contributing the picture of the Gram's stain used in this case.

Your turn

You can express your views on this article, or other relevant themes, in the Infectious Diseases in Children Specialty Forums.

[Infectious Diseases in Children Homepage]
[Current Issue] [Back Issues] [Breaking News]
[Online Seminar] [Specialty Forums] [Industry Link]
[Search]

Copyright 2000, SLACK Incorporated. Revised 15 September 2000.