A Febrile Six-Week-Old Whose Mother's GBS Screening Was Never Done
A 6-week-old boy is brought at 02:00 with a rectal temperature of 38.4 °C measured at home an hour earlier. He was born at term, birth weight 3.2 kg, and discharged at 48 hours; his mother had no group B streptococcus screening swab because she was admitted as an emergency with bleeding. He has taken about half his usual feed volumes since the evening, vomited once, and is now settled and alert. Examination: heart rate 158/min, respiratory rate 46/min, capillary refill 2 seconds, fontanelle flat, no rash, chest clear, abdomen soft, two wet nappies in 12 hours. The unit has no on-site paediatric intensive care and the registrar covers 30 beds. His 3-year-old brother had a cold sore last week, and the mother has noticed two small vesicles on her perineum that she says appeared yesterday.
1. Justify a full sepsis evaluation including lumbar puncture in a baby who looks well, and state what you would do if he deteriorated before the LP could be done.
Read the worked answer
Why appearance is not enough. In infants under 90 days, history and examination have poor negative predictive value for serious bacterial infection, and 1.5-3% of well-appearing febrile infants under three months have a urinary infection, bacteraemia or meningitis. At six weeks he falls below every validated low-risk threshold, and outside the narrow age band in which observation without cerebrospinal fluid sampling is accepted.
What to send: full blood count, CRP (or procalcitonin), blood culture, paired urine for urinalysis and culture by catheter or suprapubic aspiration - a bag specimen is worthless - and lumbar puncture for cell count, protein, glucose, Gram stain, culture and HSV PCR. If the mother's perineal vesicles are confirmed and active, swab them and the baby's skin and mucous membranes for HSV PCR. Chest radiograph only if there are respiratory signs.
Do not delay antibiotics for the LP. Take cultures and give antibiotics first; the CSF cell count is still interpretable for several hours, but the risk of deferring treatment is not.
If he deteriorates before the LP - shock, hypoxia, coagulopathy, seizures or signs of raised intracranial pressure - stabilise first with antibiotics, fluids and oxygen, and defer the LP, documenting the reasoning. A deferred LP means treating for meningitis or repeating the LP when stable; it never means stopping antibiotics early.
2. Choose your empirical antibiotics, with doses, and justify each component in light of the maternal and family history.
Read the worked answer
Regimen: ampicillin 50 mg/kg IV 6-hourly plus cefotaxime 50 mg/kg IV 8-hourly. Where cefotaxime is unavailable, use ampicillin plus gentamicin 5 mg/kg once daily with level monitoring.
Ampicillin gives reliable cover against Listeria monocytogenes, enterococcus and group B streptococcus, all of which cause neonatal sepsis and meningitis and none of which cefotaxime covers.
Cefotaxime covers Escherichia coli and other Gram-negative organisms and achieves reliable cerebrospinal fluid concentrations; it is chosen over ceftriaxone in this age group because ceftriaxone can displace bilirubin in a jaundiced neonate and precipitate with calcium-containing fluids.
Why the history matters. The mother's GBS screening was never performed and she had an emergency admission, so vertical transmission of GBS cannot be excluded; late-onset GBS disease at six weeks can present as meningitis. The brother's cold sore plus maternal perineal vesicles raise neonatal herpes: add IV aciclovir 20 mg/kg 8-hourly if the vesicles are herpetic or if the lesions are primary, and do not wait for PCR - disseminated HSV is treatable but devastating if missed. Discuss with virology and dermatology, and remember that maternal antibody does not protect against postnatal HSV.
Also measure glucose, since sepsis and herpes both cause hypoglycaemia, and start maintenance fluids containing glucose.
3. At 48 hours he is afebrile, feeding well, and all cultures are sterile. Defend your decision to stop or continue antibiotics, and state what would change it.
Read the worked answer
Stop, but only after observation on the ward. Criteria: afebrile for at least 24-48 hours, feeding normally, no focal signs, no new symptoms, downward CRP trend, and sterile blood, urine and cerebrospinal fluid cultures at 48 hours, with a correctly obtained catheter urine specimen. Then discontinue antibiotics and observe off treatment for a further 12-24 hours before discharge, with explicit safety-net advice and a review within 24 hours.
Why not simply send him home at once. A small proportion of infants with serious bacterial infection have negative cultures - because of prior maternal antibiotics, low colony counts, fastidious organisms or low-volume specimens. The culture-negative pathway exists for precisely this reason, and readmission of a deteriorated six-week-old is far worse than one additional night.
What would change the decision: a positive culture (treat 7-10 days for bacteraemia, 14-21 days for meningitis, tailored to sensitivities), a deferred or traumatic lumbar puncture (treat for meningitis length or repeat the LP), persistent fever, a rising CRP, a new focus such as an abscess or bone infection, or an active maternal herpetic lesion with any suspicious finding in the baby, which mandates a full HSV evaluation and acyclovir. Also check that the urine specimen was not a bag sample; if it was, repeat it.