Several decades ago prior to or just around the time of the arrival of the vaccines directed against Haemophilus and Pneumococcus, several studies looked at observational criteria with and without accompanying laboratory data to determine how adequately they might be able to detect a serious bacterial infection in an infant under three months of age. While these scoring systems were not perfect, they did seem effective at helping to sort the sick from the non-sick—but just how good are they today with the current vaccine schedule? One of these scoring systems was the Yale Observation Scale (YOS) initially studied by Dr. Paul McCarthy and colleagues at Yale. Nigrovic et al. (10.1542/peds.2017-0695) re-evaluated the YOS in 26 emergency departments comprising the Pediatric Emergency Care Applied Research Network. The authors compared the YOS with unstructured “clinical suspicion” for the detection of a urinary tract infection, bacteremia, or bacterial meningitis in 4591 febrile infants.
Unfortunately, the sensitivity and specificity of both the YOS and unstructured clinical observation by clinicians appear not to be as accurate as they once were when originally reported. So should we give up hope on our observational skills to detect serious bacterial infection in our youngest patients? We have asked Dr. McCarthy that question, and he responds with guarded optimism about the role of observation when combined with other diagnostic tests in an important commentary (10.1542/peds.2017-1100) that accompanies this study. You may find you will want to read both the study and commentary at a feverish pace, but take your time and fully digest just what role the YOS and even unstructured clinical observations can make even today in helping to identify a seriously ill infant with a bacterial infection.