Written by Clay Smith
The modified Philadelphia criteria (mPC) had high enough sensitivity to rule out invasive bacterial infection (IBI) in infants 29-60 days old without obtaining CSF. No infants 29-60 days old who were classified as low risk had bacterial meningitis.
Why does this matter?
Only 1/500 (0.2%) of febrile infants 29-60 days old will have bacterial meningitis. Given this low prevalence, the risk is already low. The risk is higher under 28 days, and most advocate for a full workup that includes LP. The step by step was one approach to this issue, but it requires procalcitonin, which may not be available in some centers and “neonate” was defined as ≤21 days in that study. Would the tried and true Rochester and modified Philadelphia criteria help us risk stratify 29-60-day old infants?
Brotherly love for the babies
This was a multicenter case-control study. The 135 cases were well-appearing, previously healthy, febrile (≥38°C) infants ≤60 days with IBI: 118 bacteremia; 17 bacterial meningitis (7 of these >28 days). Cases were matched with 2 controls (n = 249) from the same time and location and were similar except without IBI. Sensitivity of the mPC surpassed the Rochester criteria overall: 91.9% vs 81.5%; P = .01, but specificity was lower at 34.5% vs 59.8%; P < .001, respectively. Rochester criteria classified 25 children with IBI as low risk, 2 with meningitis; mPC classified 11 as low risk, none with meningitis. In the subgroup 29-60 days old, sensitivity was the same for both criteria at 83.6%. For children > 28 days, the mPC didn’t miss any cases of meningitis.
They gave an example. If we had 300 febrile infants 29-60 days old and assume the rate of IBI is ~2% and bacterial meningitis is 0.2%, 6 of them would have IBI; 0.6 would have meningitis. The mPC misses 1 in 6 (sensitivity 84%); so, we would miss 1 infant with bacteremia and none with meningitis, if this study is correct. See this Facebook video.
The AAP did a short video with the lead author and an excellent graphic representation of the results.
ACEP has a Clinical Policy on well-appearing infant < 2 years with fever.
Journal Watch covered this (subscription required) and commented, “Sensitivities of 92% and 82% are not great, but the rarity of invasive bacterial infection in well-appearing infants aged <61 days means that these criteria — or, for that matter, clinical gestalt — will almost always be right when used to send febrile babies home (i.e., poor sensitivity but good negative predictive value). Consensus is emerging that while neonates require lumbar puncture and inpatient care, most older babies can be managed at home and without lumbar puncture.”
Risk Stratification of Febrile Infants ≤60 Days Old Without Routine Lumbar Puncture. Pediatrics. 2018 Nov 13. pii: e20181879. doi: 10.1542/peds.2018-1879. [Epub ahead of print]
Open in Read by QxMD