This is one of those questions that always crops up when evaluating an infant for sepsis and meningitis – should we test and/or empirically cover for herpes simplex virus infection? Just how frequently is this diagnosis made?
The answers, as described in this retrospective, multi-center study, are complex. First, the basics: 26,533 total encounters analyzed, with 112 children ultimately diagnosed with HSV meningitis. Then, it’s basically chaos. The percent of patients whose CSF was tested for HSV ranged from 12.5% to 70.9% across hospitals included, along with empiric coverage with acyclovir ranging from 4.2% to 53.0%. Rates of positive HSV results were unrelated to overall institutional testing or empiric acyclovir coverage rates, excepting in the sense that HSV infection was more frequent in younger infants – and younger infants were more likely to be tested and empirically treated, in general. A handful of patients with ultimate diagnoses of HSV meningitis were not treated or tested initially, and were found on a subsequent visit.
The authors go into some detail regarding the questionable value of empiric treatment, citing a number needed to treat of 152 for infants 0-28 days and an NNT of 583 for infants from 29-60 days. Generally speaking, these authors agree with a prior cost-effectiveness analysis recommending waiting for the initial CSF cell count, and empirically treating those with a CSF pleocytosis. Consequently, these authors would therefore recommend testing only those ultimately treated empirically – but this is naturally a pragmatic consideration, rather than a statistically modeled balance between sensitivity and specificity.
There are a few more nuances within the paper with regard to their gold standard for diagnosis of HSV meningitis, limitations with regard to selection of patients undergoing testing, and generalizability from these tertiary referral settings, but it is still generally an interesting snapshot of data. Unfortunately, their ultimate conclusion is still back at square one – reiterating a call for specific clinical and laboratory data to help guide clinicians in selecting patients for HSV testing and empiric treatment. In the meantime, we’ll just keep doing our best to differentiate the ill child at the bedside based on gestalt and the culture of our practice setting.
“Herpes Simplex Virus Infection in Infants Undergoing Meningitis Evaluation”
http://pediatrics.aappublications.org/content/early/2017/12/29/peds.2017-1688
The problem also exists with adults.
Whenever I admit an adult with a clinical and CSF picture of simple uncomplicated viral meningitis, I don’t prescribe aciclovir.
If I have the slightest doubt it could be an encephalitis, I do start ACV STAT.
Nothing very peculiar I guess.
But I always hear the admitting internist whine that I should have started and you never know.
From what I hear in other places, when the patient is admitted to internal medicine, they give ACV, and when admitted to Infectious Disease they don’t.
I found internal medicine also tend to give meningitis doses of antibiotics for viral like meingitis cases (clinical and CSF) , even covering for listeriosis in patients 20-50 with no risk factor.
Same for you ?
Virtually no one ever gets reprimanded for being too conservative in these medical grey areas … but goodness knows you’ll hear about it if you were to miss something unexpected!