Jump to comment:
- Page navigation anchor for Response to Walkin and LoftusResponse to Walkin and LoftusA case-control study design is susceptible to confounding bias. In our study we aimed to minimise this by including differences in major chronic illnesses (as in Table 1) between the meningitis and control groups; thus our study presents odds ratios adjusted for these potential sources of bias. We also controlled for lipid lowering drugs (which have an anti-inflammatory effect) but we did not correct for immunosuppressants for renal disease – though we did adjust for renal disease itself. We therefore believe our approach has minimised potential confounding bias from these sources, especially given the relatively large differences between study arms in prior antimicrobial prescriptions.Nevertheless, there may still be residual confounding and we did suggest in the paper that the most likely alternative explanation for our findings was the possibility that some patients are just more susceptible to infections, including meningitis, than others. This would have to translate into more GP visits (mainly, we presume, with URTIs) and more antimicrobial prescriptions. If there is such a susceptibility this may indeed reflect impaired immune function though this of itself would be an interesting finding as such widespread susceptibility to infection in some patients has not, to our knowledge, previously been reported.We acknowledge that there are potential residual confounders in our study but this is the cons...Show MoreCompeting Interests: None declared.
- Page navigation anchor for The relationship between prior antimicrobial prescription and meningitis: a case–control studyThe relationship between prior antimicrobial prescription and meningitis: a case–control studyArmstrong et al's case-control study found an association between number of antimicrobial prescriptions and subsequent increased risk of both bacterial and viral meningitis. The data illustrated a dose-response relationship, with increasing numbers of antibiotic prescriptions associated with a greater incidence of bacterial meningitis.These findings could support their hypothesis. There is, however, a trend towards narrower spectrum antibiotics and antibiotics for urinary tract infections having a greater associated risk of bacterial meningitis. This appears inconsistent with their hypothesis. Furthermore we noted that cases in this study appear to be less healthy than controls, with a higher prevalence of renal disease and chronic obstructive pulmonary disease (COPD). In light of this we wondered whether the association is confounded by impaired immune function? For example, given the increased prevalence of COPD amongst cases, there may be increased use of medication with immunological effects by cases. Is corticosteroid use more prevalent amongst cases than controls? Could be increased use of immunosuppressants for renal disease in cases compared to controls explain the increased meningitis risk?The greater use of antibiotics in most cases may represent a true need, reflecting increased risk of many infections including meningitis. Should the data be available, it would be interesting to see if prescr...Show MoreCompeting Interests: None declared.