Edit: I updated this post to focus on the newer paper after discussing it with Dr. Nigrovic
You could go ahead and pick out a good book this summer. But why waste time on formulaic plots and slapdash characterizations when you could be reading great articles relevant to the Pediatric Emergency Department. So without further ado, I introduce the PEMBlog Summer Reading List!
Clinical prediction rule for identifying children with cerebrospinal fluid pleocytosis at very low risk of bacterial meningitis
Nigrovic et al.
JAMA. January 3, 2007, Vol 297, No. 1
What this article is about?
Bacterial meningitis is super rare in the post-vaccine era. Most cases are caused by Streptococcus pneumoniae and Neiserria meningiditis. Children with CSF pleocytosis are frequently admitted with parenteral antibiotics on board despite their illness being almost always viral. This study is an effort to validate a decision rule to identify children at risk for bacterial meningitis.
This study follows one originally published in Pediatrics in 2002, and was a multi center retrospective cohort study of over 3,000 children with CSF pleocytosis over a 3 ½ year period. The children were between 29 days and 19 years. 121/3295 (3.7%; 95% CI, 3.1%-4.4%) had bacterial meningitis and 3174/3295 (96.3%; 95% CI, 95.5%-96.9%) had aseptic meningitis. 80% of the aseptic meningitis patients were hospitalized!
1714 patients were categorized as very low risk for bacterial meningitis by the Bacterial Meningitis Score and only 2 of these had bacterial meningitis – both of which were <2 months old;
- Sensitivity 98.3%; 95% CI, 94.2%-99.8%
- NPV 99.9%; 95% CI, 99.6%-100%
Is is the case with most retrospective studies of this type, the authors originally used multivariable logistic regression to identify predictors of bacterial meningitis. They found five:
- CSF gram stain positive for bacteria
- CSF protein > or =80 mg/dL
- Peripheral ANC ≥10 000 cells/mm3
- Seizure before or at time of presentation
- CSF ANC ≥1000 cells/mm3
The Bacterial Meningitis Score was internally validated, and it was found that a score of zero accurately identified kids with aseptic meningitis and did not miss any bacterial meningitis. Children with NONE of the above factors are therefore very low risk.
Why is it important?
Obviously it helps identify children with a potentially life threatening illness, and can help avoid hospitalization for those at low risk. This original study was performed in the post HiB era (which officially kicked off with the “new-and-improved” 1987 version), and has since been re-validated in the post Prevnar era (post 2000).
How you can use it in your practice
First of all – if you think that a child has viral/aseptic meningitis think about whether or not you need to perform a lumbar puncture in the first place. If you do, by all means use the bacterial meningitis score. There’s even a helpful MDCalc tool – based on the 2002 study! Though, per Dr. Nigrovic herself, the important way to really look at this is ZERO vs NOT ZERO. The score of zero kids are the only truly low risk ones.
You can even use it in the littlest of babies (>29 days). The lowest risk babies can get blood and urine alone – but if you have a feeling about a baby, and get an LP you can use this score to send them home OFF ANTIBIOTICS.
What this article is about?
Bacterial meningitis is super rare in the post-vaccine era. Most cases are caused by Streptococcus pneumoniae and Neiserria meningiditis. Children with CSF pleocytosis are frequently admitted with parenteral antibiotics on board despite their illness being almost always viral. This study is an effort to validate a decision rule to identify children at risk for bacterial meningitis.
This study follows one originally published in Pediatrics in 2002, and was a multi center retrospective cohort study of over 3,000 children with CSF pleocytosis over a 3 ½ year period. The children were between 29 days and 19 years. 121/3295 (3.7%; 95% CI, 3.1%-4.4%) had bacterial meningitis and 3174/3295 (96.3%; 95% CI, 95.5%-96.9%) had aseptic meningitis. 80% of the aseptic meningitis patients were hospitalized!
1714 patients were categorized as very low risk for bacterial meningitis by the Bacterial Meningitis Score and only 2 of these had bacterial meningitis – both of which were <2 months old;
- Sensitivity 98.3%; 95% CI, 94.2%-99.8%
- NPV 99.9%; 95% CI, 99.6%-100%
Is is the case with most retrospective studies of this type, the authors originally used multivariable logistic regression to identify predictors of bacterial meningitis. They found five:
- CSF gram stain positive for bacteria
- CSF protein > or =80 mg/dL
- Peripheral ANC ≥10 000 cells/mm3
- Seizure before or at time of presentation
- CSF ANC ≥1000 cells/mm3
The Bacterial Meningitis Score was internally validated, and it was found that a score of zero accurately identified kids with aseptic meningitis and did not miss any bacterial meningitis. Children with NONE of the above factors are therefore very low risk.
Why is it important?
Obviously it helps identify children with a potentially life threatening illness, and can help avoid hospitalization for those at low risk. This original study was performed in the post HiB era (which officially kicked off with the “new-and-improved” 1987 version), and has since been re-validated in the post Prevnar era (post 2000).
How you can use it in your practice
First of all – if you think that a child has viral/aseptic meningitis think about whether or not you need to perform a lumbar puncture in the first place. If you do, by all means use the bacterial meningitis score. There’s even a helpful MDCalc tool – based on the 2002 study! Though, per Dr. Nigrovic herself, the important way to really look at this is ZERO vs NOT ZERO. The score of zero kids are the only truly low risk ones.
You can even use it in the littlest of babies (>29 days). The lowest risk babies can get blood and urine alone – but if you have a feeling about a baby, and get an LP you can use this score to send them home OFF ANTIBIOTICS.
