Clinical recognition of meningococcal disease

Abstract

Matthew Thompson and colleagues (Feb 4, p 397)1Thompson MJ Ninis N Perera R et al.Clinical recognition of meningococcal disease in children and adolescents.Lancet. 2006; 367: 397-403Summary Full Text Full Text PDF PubMed Scopus (365) Google Scholar studied the presenting symptoms of children admitted to hospital with meningococcal disease. We have some concerns about their conclusions. First, they studied children who died from meningococcal disease, matched with non-fatal cases. Such cases are likely to be biased towards meningococcal septicaemia, and do not represent the whole spectrum of disease. Such bias will affect the symptoms reported and their timing, since septicaemia is a more rapidly progressive disease than meningococcal meningitis. Second, a control group without meningococcal disease was not included, meaning that the reliability of the identifying symptoms of meningococcal disease cannot be ascertained. Third, the data are subject to recall bias. Most of the data on parents' recollection of symptoms were collected by postal questionnaire, with only 30% collected by personal interview; all were collected retrospectively. Accurate recall of symptoms and their timing is unlikely given this delay. Fourth, although Thompson and colleagues describe the method of adjustment for symptom frequency, they do not state if and how this adjustment was validated. Olcén and colleagues2Olcén P Barr J Kjellander J Meningitis and bacteremia due to Neisseria meningitidis: clinical and laboratory findings in 69 cases from Orebro county, 1965 to 1977.Scand J Infect Dis. 1979; 11: 111-119PubMed Google Scholar did a study of pre-admission symptoms on the basis of clinical histories taken on the day of admission; they also included controls without meningococcal disease. Fever was the most common feature noted by patients' relatives at the start of meningococcal disease (74%); however, only vomiting was significantly more common in patients with meningococcal disease than controls.2Olcén P Barr J Kjellander J Meningitis and bacteremia due to Neisseria meningitidis: clinical and laboratory findings in 69 cases from Orebro county, 1965 to 1977.Scand J Infect Dis. 1979; 11: 111-119PubMed Google Scholar We believe that some of Thompson and colleagues' recommendations are unfounded. First, we do not agree that meningococcal disease is a less likely diagnosis if symptoms have lasted more than 24 h. In our unpublished prospective study of 302 children with meningococcal disease, 50 (16·5%) had symptoms that lasted more than 24 h before admission. Of these, ten had meningitis alone, 22 had meningitis and septicaemia, and 18 had septicaemia alone. Second, parents are not good at detecting neck stiffness. In one study,3Riordan FA Thomson AP Sills JA Hart CA Who spots the spots? Diagnosis and treatment of early meningococcal disease in children.BMJ. 1996; 313: 1255-1256Crossref PubMed Scopus (56) Google Scholar only 11% of parents of children with meningococcal disease thought their child had neck stiffness, when this sign was actually present in 33%. Such findings bring into question the accuracy of assessment of symptoms and their timing in Thompson and colleagues' study. Third, we question the feasibility of the recommendation for family doctors to reschedule appointments for 4–6 hours' time “if the clinician has concerns that are not serious enough to warrant urgent admission”. Fourth, we are aware that some parents are now seeking medical advice because their (well) children have “cold hands and feet”. Increasing the number of medical consultations with these features is likely to make it more difficult to identify those with meningococcal disease, not less. Although we appreciate efforts to identify meningococcal disease early in its course, Thompson and colleagues' study does not help. We suggest that a prospective study with controls is devised to assess the features suggested by Thompson and colleagues before they are used to identify children with meningococcal disease. The Liverpool Meningococcal Research Group are: *N Makwana, F A I Riordan, E D Carrol, S J Hackett, A P J Thomson, P B Baines, and C A Hart. We declare that we have no conflict of interest. Clinical recognition of meningococcal disease – Authors' replyWe acknowledge that the original study design we used to identify children with meningococcal disease included a greater proportion of fatal cases (and hence septicaemia) than would be expected in a typical UK cohort of children with meningococcal disease. It was for this reason that we presented the weighted frequency of symptoms in fatal and non-fatal cases using age-specific case-fatality rates. Given that the data were collected retrospectively (but validated where possible with contemporaneous doctors' records), both N Makwana and colleagues and Aidan Kirkpatrick and colleagues echo the limitation noted in the discussion regarding the possibility of recall bias. Full-Text PDF