Interesting and a little challenging case!

High TLC is difficult to ignore but studies have shown that high TLC is not very sensitive/specific marker of neonatal sepsis (not as useful as leukopenia is). It would be useful to know serial inflammatory markers (CRP etc). Also, I wonder why CSF was done (symptomatic baby or part of routine sepsis work-up etc). What is CSF red cell count as that can change the CSF picture. Obviously herpes is ruled out as -ve PCR. It is interesting to see CSF white cells count (0) on day-7 in a preterm and I wonder if first CSF was blood stained.

To conclude, if we have other supportive evidence, then will treat as possible bacterial meningitis with 14 days of appropriate antibiotics - this is obviously a pragmatic approach.

Whoever reads this blog (actually not a blog), kindly shift this discussion to forums. Lumbar puncture was done as part of sepsis work up. CSF had only 20,000 RBC. It cannot account for 600 cells in CSF. Never seen a CSF picture like this with increased cells and normal biochemistry in CSF and then all cells just vanishing (0 cell count) within one week of treatment.