Meningitis studies reveal limitations of observational data
Suggestions of harms and benefits resulting from pre-hospital administration of antibiotics to children with suspected meningitis may be statistical artefacts, two papers published in the BMJ suggest.
Observational studies seeking to determine the impact of giving antibiotics
to children before admission to hospital have had conflicting results.
A confounding factor may be a doctor’s perception of the severity
of meningococcal disease, which is likely to determine the choice of
treatment.
One study (2006;332:1295) found that administration of parenteral penicillin
by GPs was associated with increased odds ratios for death and for complications
in survivors. “One likely explanation for the high odds ratio we
obtained for mortality is that there is a strong residual selection bias
towards giving penicillin to the most severely ill children,” the
researchers conclude. They argue that current discrepancies in the data
can only be addressed by a randomised controlled trial.
The authors of the second paper (2006;332:1299) were unable to determine
whether or not antibiotics given before admission to hospital affected
mortality, since their data were consistent with a benefit when a sufficient
population of cases was treated. “Confounding by severity is the
most likely explanation both for the beneficial effect of oral antibiotics
and the harmful effect observed in some studies of parenteral antibiotics,” the
authors conclude.
They argue, however, that a trial to answer the question definitely would
not be feasible. “It is unlikely that studies of sufficient quality
to give evidence of the effectiveness of antibiotic treatment before
admission on survival in meningococcal disease will become available
in view of the anticipated logistical and ethical difficulties,” they
say. |
The
Pharmaceutical Journal
Acrobat
Reader