Good to see worldwide interest in the febrile infant. As authors of the PECARN febrile infant rule being analyzed here, we have some observations:
1) The Velasco study was a retrospective analysis of a registry at one hospital in Spain. In contrast, the PECARN study was a prospective study in 20 centers analyzed by an independent data center.
2) The rate of SBI in Velasco study was 20% which is much higher than the rate published by PECARN and other investigators on this topic. This suggests a different population, different enrollment criteria and/or enrollment bias.
3) Although the PECARN rule was derived on febrile infants age 0 to 60 days, we believe it should be only implemented in those 29 to 60 days-old. We suggest this in our original article jamanetwork.com/journals/jamap… as well as in @mdcalc
4) We suggest in PECARN article (and supplement) and have validated on 1363 more infants that PECARN variable cutoffs should be rounded to ANC 4K and PCT 0.5. Does not significantly change rule accuracy. Rule is accurate, objective and simple: UA, ANC, PCT
5) With rounded cutoffs we have now enrolled ~3200 and have not missed any bacterial meningitis. Sensitivity for SBI ~98%, specificity ~61%. . We are now implementing.
6) if implemented as we suggest, 4 of 5 “missed cases“ of SBI reported in the Velasco article would not be missed by the PECARN rule. Our rule is not meant for implementation in first month of life as described above.
7) Several other concerns w/ Velasco study, including accuracy of parent-reported fever duration, their rule variable cutoffs, and potential investigator bias as authors of another rule @PECARNteam @riseup_sperg @UCDavisEM @PEMUpdates @ALiEMteam @mahajanpemdoc
