BACKGROUND: Pertussis or “whooping cough” is a highly contagious respiratory disease frequently characterized by a paroxysmal cough, post-tussive vomiting and/or an inspiratory “whoop”. Though previous analyses have described individual-level characteristics associated with pertussis infection, there is limited knowledge of whether neighborhood socioeconomic factors are associated with risk of infection. We describe case-based demographic and neighborhood socioeconomic characteristics of pertussis cases in New Mexico.
METHODS: Probable and confirmed pertussis cases (n=2128) between 2012-2015 were linked by census tract (CT) with American Community Survey data and aggregated by socioeconomic factor (poverty, household crowding, and female head of household (HH)). Poverty level was categorized by percent of CT residents below poverty threshold (≤9%, 10-≤19.9%, 20-≤29.9%, and ≥30%). Crowding and female HH were divided into quartiles. Incidence rates were calculated per 100,000 person years with 95% CIs by demographics (age, gender, race/ethnicity) and age-adjusted for socioeconomic factors, overall and stratified by ethnicity.
RESULTS: Incidence of pertussis decreased from 42.8 in 2012 to 11.7 in 2015 with infants and children (0-17) disproportionately affected compared to adults (p<0.05). Hispanics had higher overall incidence rates compared to non-Hispanic whites (NHW) and American Indian/Alaskan Natives (AIAN), 32 (29.3-35.4) vs. 15.5 (14.2-16.9) and 18.6 (13.4-25.3), despite being more up-to-date on vaccination than non-Hispanics (p<0.05). Residing in the lowest poverty CTs was associated with significantly higher rates of pertussis compared to the medium-low, medium-high and high poverty CTs, 29.6 (27.3-32.0) vs. 19.1 (17.7-20.6), 18.6 (16.9-20.4), and 17.2 (15.1-19.5), overall and stratified by ethnicity. Hispanics had rates 3-9 times greater than NHW regardless of poverty level. Residing in lowest crowding CTs was significantly associated with higher incidence compared to medium-low, medium-high and high crowding CTs, 29.4 (26.9-32.1) vs. 21.5 (19.7-23.3), 17.1 (15.6-18.8), and 18.1 (16.6-19.7); similar trends were observed for female HH. After stratifying by ethnicity, however, higher rates of pertussis among lowest crowding and female HH CTs were only significant among NHW. Upon further examination, cases residing in lowest poverty CTs vs. highest poverty CTs were more likely to be laboratory-confirmed (58% vs. 53%) and less likely to be hospitalized (4.4% vs 3.1%) and UTD on vaccinations (52% vs. 58%).
CONCLUSIONS: In contrast to previous respiratory disease neighborhood studies, pertussis incidence was highest among CTs with low poverty, low crowding, and low female HH; though this varied by ethnicity and CT variable. Our findings suggest that further study considering access to care, healthcare utilization, and provider testing practices is needed.