Epidemiology of Meningococcal Disease, New York City, 1989-2000

Alexandre Sampaio Moura, Ariel Pablos-Méndez, Marcelle Layton, Don Weiss


Emerging Infectious Diseases. 2003;9(3) 

In This Article


Among New York City residents, 615 cases of meningococcal disease were reported to the NYC Department of Health from January 1989 to December 2000, with an average annual incidence of 0.67/100,000; of cases reported, 582 cases (95%) were confirmed and 33 cases (5%) were included as probable. Meningococcemia occurred in 54% of the cases, meningitis in 44%, and pneumonia, septic arthritis or other sterile site infections in 2%. All cases were considered to be sporadic except for two case-patients in 1997 who were contacts of a primary case-patient in a juvenile detention center resident and one culture-negative case-patient in 2000 who was linked to a subsequent confirmed case-patient by household contact.

For the period 1989-2000, the meningococcal disease rate decreased by 33%, compared to the period 1953-1988, and declined by 90%, compared to the period 1905-1952 ( Table 1 ). During the period under study, a 69% reduction occurred at the beginning of the 1990s, with the rates dropping from 1.19 per 100,000 population in 1989 to 0.37 per 100,000 population in 1992 (chi square for trend = 9.1; p<0.01). Since then, rates have increased slightly and remained relatively constant ( Table 1 ). When children <1 year of age are excluded, the declining trend in incidence is no longer statistically significant (chi square for trend = 2.4; p<0.12).

When stratified by borough of patient residence, the average incidence rates were highest in the Bronx (0.88/100,000) and Manhattan (0.81/100,000) and lowest in Brooklyn (0.65/100,000), Staten Island (0.65/100,000), and Queens (0.55/100,000). However, the differences between boroughs were not statistically significant (chi square = 1.4; df=4; p=0.23). Rates by United Hospital Fund neighborhoods ranged from 0.23 to 1.08 per 100,000. The highest rates occurred in two northern Manhattan and one central Bronx neighborhoods; the lowest rates were all in Queens.

The highest average annual incidence rate was observed among patients <1 year of age (8.49/100,000), with substantially lower rates observed for older age groups ( Table 2 ). A statistically significant declining trend for the age groups of <1 years of age (chi square for trend = 21.5; p<0.01) and 1-4 years of age (chi square for trend =14.3; p<0.01) was seen over the four 3-year groups. No other decrease or increase in age-specific incidence trends was statistically significant. The proportion of cases occurring in young children (<5 years of age) decreased from 39% in 1989-1991 to 17% in 1998-2000.

The overall median age of the patients with meningococcal disease was 22 years; stratification by year group showed that median age has increased from 15 years of age in 1989-1991 to 30 years of age in 1998-2000 (Kruskal-Wallis test; chi square = 20.0; df =3; p<0.01). To assess the effect of changes in serogroup on the median age, serogroups B, C, Y, and unknown were sequentially excluded from the computation of median age. Only the removal of serogroup Y resulted in a loss of statistical significance of the trend in median age (Kruskal-Wallis test; chi square = 7.6; df=3; p=0.06).

Overall incidence rates were higher for males (0.73/100,000) than females (0.61/100,000; relative risk [RR] = 1.19; 95% confidence interval [CI] 1.02 to 1.40). However, CFR was higher (20.1% vs. 13.9%; RR=1.45; 95% CI 1.02 to 2.07) for females. No statistically significant differences were found in gender-specific incidence rate by age category.

Serogroup was determined for 423 (72%) of 582 culture-positive cases. From 1989 to 2000, serogroups B, Y, and C were the most commonly identified serogroups (32% [n=137], 28% [n=119], and 27% [n=112], respectively) of the cases for which a serogroup was known. Serogroup W135 constituted 7% (n=28); nongroupable, 3% (n=13); A, 2% (n=9); and other serogroups, 1% (n=5) of the isolates. The median age of the case-patients differed by serogroup, with the highest median age for nongroupable (48 years of age), followed by other (43 years of age), Y (37 years of age), A (34 years of age), W (27 years of age), C (23 years of age), and B (11 years of age).

Incidence rates for serogroup B infections declined in all age groups with the largest decline in the <1-year and 1-4-year age groups in 1989-2000. Serogroup Y incidence rates increased twofold to tenfold in all age groups except 1-4 years during the period ( Table 3 ).

Over the 12-year interval, the proportion of cases caused by strains included in the quadrivalent vaccine available in the United States (A, C, Y, and W135) increased from 28% to 65% of reported cases (Kruskal-Wallis test; chi square = 57.4; df=3; p< 0.01). This increase is due in part to the decline in incidence of serogroup B infections and the decline in the number of cases for which a serogroup could not be determined (Figure 1).

Distribution of meningococcal serogroups by year group, New York City, 1989-2000.

To assess whether changes in serogroup B and Y incidence were independent from the changes observed in age, we performed logistic regression analyses. The likelihood of serogroup Y infection compared with all other serogroups increased by a factor of 2.47 (99% CI 1.84 to 3.33) for each successive year group while controlling for age. The likelihood of serogroup B infections compared with all other serogroups decreased by a factor of 0.77 (99% CI 0.61 to 0.91) for each successive year group.

Information about patient outcome was initially available for 478 (77.7%) of the cases. After the vital records search, one additional death was identified for a patient with missing outcome. The overall CFR during 1989-2000 was 16.9% (104 deaths, 615 cases). The CFR varied during the study period, being lowest in the interval 1992-1994 (14%; 15/109) and highest in 1989-1991 when 20% (39/196) of the case-patients died; however, the difference between year groups was not statistically significant. When we analyzed CFR for each year separately, we found a surprisingly high CFR of 27% (16/59 cases) in 1999.

CFR increased linearly with age after 5 years of age and was lowest for those 5-14 years of age (8%) and highest for ≥65 years of age (33%) ( Table 2 ). Figure 2 shows CFR by age category and year group. The CFR also differed by serogroup and was the highest for serogroup A (44.4%; 4/9), compared to that observed among serogroups C (22.3%; 25/112), Y (18.5%; 22/119), W (17.9%; 5/28), and B (12.4%; 17/137). However, the high CFR for serogroup A should be interpreted cautiously because of the low number of cases in the study period. No statistically significant difference of CFR between serogroups was noted (Bonferoni adjustment for multiple comparisons, p>0.002).

Meningococcal case-fatality rate by age category and year group, New York City, 1989-2000.