Letters

Carriage of meningococci in contacts of patients with meningococcal disease

"Kissing contacts" need to be defined

Chemoprophylaxic strategy needs to be determined

Age and other risk factors need to be taken into account

Authors' reply

"Kissing contacts" need to be defined

EDITORIn their study Kristiansen et al show high carriage rates of pathogenic strains of Neisseria menigitidis in household and kissing contacts of patients with invasive meningococcal disease.¹ While it is easy to define a household contact it may be more difficult to define a kissing contact. There are many types of kiss, ranging from a "peck on the cheek" to much more! In some cultures kissing is as common as shaking hands. In such situations widespread chemoprophylaxis to "kissing contacts" may not be appropriate. It would be valuable to know whether Kristiansen et al placed any restrictions on who was defined as a kissing contact.

Andrew Hayward, Lecturer in public health medicine. 
Medical School, University of Nottingham, Nottingham NG7 2UH Andrew.Hayward{at}nottingham.ac.uk

1. Kristiansen BE, Tveten Y, Jenkins A. Which contacts of patients with meningococcal disease carry the pathogenic strain of Neisseria meningitidis? A population based study. BMJ 1998; 317: 621-625[Abstract/Free Full Text]. (5 September.)

Chemoprophylaxic strategy needs to be determined

EDITORThe study by Kristiansen et al addresses important questions about the use of chemoprophylaxis in contacts of patients with meningococcal disease.¹ We believe, however, that the data presented do not fully support the conclusions. Kristiansen et al found high rates of meningococcal carriers among class 1 contacts (12.4%) and advocated the use of chemoprophylaxis in this group, on the basis of their assumption that carrying the pathogenic strain increases the likelihood of contracting the disease. One concern is that this group accounts for only 18 of 42 contacts who were found to be carriers. More than half of the carriers would therefore not receive prophylactic treatment. Should classes 2 and 3 be excluded?

We accept that giving prophylactic treatment to all those in groups 2 and 3 seems excessive. It would have been interesting, however, if Kristiansen et al had assessed the characteristics of these carriers to identify those at highest risk and target chemoprophylaxis more effectively.

In addition, the carrier rates were compared with those in other Norwegian populations in which the definition of pathogenic strain was different. ^{2 3} It would have been more appropriate to compare their carrier rates with the prevalence among people in the Telemark area who had not been in contact with meningococcal disease.

Kristiansen et al's paper does not determine chemoprophylactic strategy conclusively. It also highlights the fact that there is still much to learn about the relation between carriage of meningococci and meningococcal disease.

Peter Dutton, Fourth year medical student. 
Robert Winterton, Fourth year medical student. 
R.I.S.Winterton{at}ncl.ac.uk

Ewan Wright, Fourth year medical student. 
Han San Aw Yeang, Fourth year medical student. 
Department of Epidemiology and Public Health, Medical School, University of Newcastle, Newcastle upon Tyne NE1 7RU

1. Kristiansen BE, Tveten Y, Jenkins A. Which contacts of patients with meningococcal disease carry the pathogenic strain of Neisseria meningitidis? A population based study. BMJ 1998; 317: 621-625. (5 September.)
2. Kristiansen BE, Lind KW, Mevold K, Sorensen B, Froholm LO, Bryn K, et al. Meningococcal disease: studies of bacterium phenotypic and genomic characteristics and of human antibody levels. J Clin Microbiol 1988; 26: 1988-1992[Abstract/Free Full Text].
3. Caugant DA, Hoiby EA, Magnus P, Scheel O, Hoel T, Bjune G, et al. Asymptomatic carriage of Neisseria meningitidis in a randomly sampled population. J Clin Microbiol 1994; 32: 323-330[Abstract/Free Full Text].

Age and other risk factors need to be taken into account

EDITORKristiansen et al used an interesting study design looking for carriers of Neisseria meningitidis among contacts of patients with meningococcal disease.¹ The results show a strong relation between degree of contact and prevalence of carriership, with the highest prevalence rates among household and kissing contacts. The results are clear, and the prevalence of the pathogenic strain of 12.4% among household and kissing contacts is high.

The prevalence of meningococcal carriage is strongly associated with age, following a typical age distribution, and with other risk factors. This has been shown in studies from various countries, including one from Norway.² Peak rates are observed among people aged 15-30.

A recent study that was conducted after a local outbreak of meningococcal disease in the Netherlands made it clear that several risk factors are involved. The prevalence of meningococcal carriers in a systematic sample of the general population varied from 3% among children aged 2-5 to 39% among those aged 16-20 and was lower among those who had recently been taking antibiotics (odds ratio 0.3; 95% confidence interval 0.1 to 0.9). The prevalence of carriers of the pathogenic strain causing the outbreak in the population was low (0.5%).

This raises several questions. Have Kristiansen et al have studied age as a determinant of carriage? To what extent did different age distributions in the three contact classes vary from that in the general population, and did correction for age influence the rates of prevalence in the three groups? Does the prevalence of carriage among class 2 and 3 contacts (non-household and non-kissing contacts) then still exceed the prevalence found in the general population by two to three times? Has the effect of other known risk factorsfor example, crowding or use of antibiotics in the weeks before the studybeen evaluated?

Ralf Reintjes, Medical epidemiologist in infectious diseases. 
Ralf.Reintjes{at}RIVM.NL

Marina A E Conyn-van Spaendonck, Medical epidemiologist in infectious diseases. 
Department of Infectious Diseases Epidemiology, National Institute of Public Health and the Environment, NL-3720 BA Bilthoven, Netherlands

1. Kristiansen BE, Tveten Y, Jenkins A. Which contacts of patients with meningococcal disease carry the pathogenic strain of Neisseria meningitidis? A population based study. BMJ 1998; 317: 621-625. (5 September.)
2. Caugant DA, Hoiby EA, Magnus P, Scheel O, Hoel T, Bjune G, et al. Asymptomatic carriage of Neisseria meningitidis in a randomly sampled population. J Clin Microbiol 1994; 32: 323-330.

Authors' reply

EDITORAs Hayward points out, the meaning of "kissing contact" depends on cultural context. In the Norwegian context, this signifies mouth to mouth contact, and that is how we intended the term to be understood. As Dutton et al observe, exclusion of non-kissing, non-household contacts would mean that more than half of the carriers whom we found would not have received prophylaxis. Indiscriminate inclusion of these contacts would, however, have resulted in 1366 unnecessary courses of antibiotics. This is why we isolate and genetically characterise the carrier strains.

Are there simpler ways of selecting candidates for chemoprophylaxis? Reintjes specifically suggests age as a selective variable, and this prompted us to reanalyse our database. We found that the overall rate of meningococcal carriage is lower in children aged under 5 (17/203, 7.3%) and 5-12 (24/373, 6.4%) than in those aged 13-18 (70/345, 20.3%) and over 18 (121/614, 19.7%). This does not, however, seem to be reflected in the rates of carriage of the patient strains (6/203, 3%; 7/373, 1.9%; 6/345, 1.7%; and 23/614, 3.7%). Crowding is probably a relevant variable but is not easy to assess objectively. Previous antibiotic treatment would be expected to reduce overall carriage rates but may at the same time render the person treated more prone to colonisation by the patient strain. As Dutton et al have noticed, we cannot, on the basis of our data, determine whether or not the rate of carriage of the strains causing disease in non-kissing, non-household contacts is higher than that in the general population; our numbers are small, our confidence intervals are wide, and the best available comparable study of the general population was conducted in a distant part of the country. The correct population for comparison would be non-contacts from Telemark (as Dutton et al suggest), age matched with the contacts, and ideally collected over the same time (not achievable in this case). We acknowledge the value of such a study and consider it a priority in our future research programme.

Andrew Jenkins, Researcher. 
Yngvar Tveten, Consultant. 
Telelab, Telemark Biomedical Centre, Gulset, N-3705 Skien, Norway

Bjørn-Erik Kristiansen, Professor. 
Department of Medical Microbiology, University of Tromsø, N-9037 Tromsø, Norway