Throat swabs from students and staff were spread onto 5% Columbia blood agar containing polymyxin 25,000 units/L and vancomycin 3 mg/L. Primary incubation was conducted at 37°C for 48 hours in 10% CO₂. Plates showing preliminary growth were sent to the Meningococcal Reference Unit at Manchester Public Health Laboratory for further examination and serotyping. Rifampicin was given to students in year 11, which included all those subsequently found to be carriers of the epidemic strain.

In February 1995, we investigated risk factors for the carriage of any meningococci among the contacts of three ill students at another Welsh secondary school (8). These index patients were students in different school years, and 2 (1.7%) of 119 contacts carried the epidemic strain (type B2b P1.10). In the current outbreak, we identified 11 (1.5%) of 744 students with the epidemic strain (type B15 P1.16). In contrast, outbreaks of disease associated with serogroup C disease are typically accompanied by lower rates of carriage in populations at risk. However, outbreaks of group C disease involving higher carriage rates are occasionally described. In an outbreak at an agricultural college in England, carriage of the epidemic strain of serogroup C organisms among students and staff was 6.2% (9). In the current outbreak, the fact that all carriers of the epidemic strain were in the same school year as the two index patients enabled us to examine risk factors for carriage of the epidemic strain. Risk factors for the carriage of any meningococci may differ. For example, in an outbreak of six cases among first-year students at Southampton University (United Kingdom) in 1997, 0.9% of students surveyed carried serogroup C strains; however, no first year students were carriers (10).

Better knowledge of risk factors for carriage of epidemic meningococci may help identify close contacts who are candidates for antimicrobial therapy to eliminate nasal carriage and prevent spread of disease (11). Carriers of an epidemic strain with the potential to infect others may be missed, and a number of people may receive unnecessary antibiotics, which has implications for the spread of antibiotic resistance.

Carriage of any meningococci was associated in the univariate analysis with the well-described risk factors of increasing age and smoking (active and passive) (12-14). Smoking may predispose to colonization by inhibiting bronchial ciliary action (12). We also found increasing age and active smoking to be associated with carriage in our previous study (8). In the current study, receipt of bad news was associated. No obvious biological mechanism exists to explain what may be a chance finding, although recent stress has been described as a risk factor for meningococcal disease (6). Being in the same year as the index patients and attending the end-of-year party were risk factors for carriage of any strain of meningococci in this outbreak, mainly because these were the only two risk factors associated with carriage of the epidemic strain. These two factors reflect the kinds of social contacts among teenagers and young adults that may permit spread of meningococci. In a review of 22 school-based clusters between 1989 and 1993, the patients in nine of the clusters had contact through extracurricular activities; these activities in four clusters were parties or dances (15). Patronage of a particular bar was implicated in the university outbreak in Illinois (16), and a particular discotheque in an outbreak among eight adults (five of whom were teenagers) in Corrientes, Argentina (17). Such social settings differ from the residential settings, of outbreaks among military recruits and prisoners, where overcrowding and proximity of beds may permit transmission (18).

Attendance at the party may have been the critical factor in carriage of the epidemic strain among year 11 students. However, another hypothesis may account for our observations. Young adults who socialize frequently at discotheques and parties may, particularly if they smoke, be at higher risk for carriage of meningococci of all types. This increased long-term risk for disease may have a protective effect against a virulent outbreak strain (19). In contrast, those who participate infrequently in social events such as the end-of-year party may be at higher risk. This hypothesis may explain some of the risk for first-year university students. In addition to longitudinal studies, combining the results of surveys during outbreaks may help provide a more scientific basis for the management of future outbreaks.