Pertussis vaccination coverage and incidence in King County, Washington

From the Halloran Group, Biostatistics, Bioinformatics, and Epidemiology Program, Vaccine and Infectious Disease Division

What do a runny nose, an ear infection, and a cough have in common? These are all symptoms often present in young children, especially those in aggregate settings such as daycare, and can represent minor illnesses as children explore the world with their hands (and mouths!). However, some illnesses that were previously common in childhood, for example pertussis a highly transmissible infectious disease also known as whooping cough, can be detrimental to the health of a young child. Fortunately, in the case of pertussis, highly efficacious vaccines exist that provide protection and limit spread. “Pertussis vaccination rates have been historically high in Washington (WA) state, in particular in King County,” explained Dr. Madhura Rane, a former PhD candidate in Dr. M Elizabeth (Betz) Halloran’s Group, part of the Fred Hutch Vaccine and Infectious Disease Division. However, problems still exist. “Despite statewide pertussis vaccination coverage, WA experienced a large pertussis outbreak in 2012 and another one 2015 with sporadic activity in between, raising questions about whether we are experiencing a resurgence of pertussis in the era of acellular pertussis vaccines,” she continued.

Dr. Rane (now an Epidemiologist with the California Department of Public Health), under Dr. Halloran’s mentorship, sought to investigate this further, noting that “even with high county-level or state-level vaccination coverage, pockets of low vaccination coverage, or vaccination “cold-spots” could exist locally”. Their study, recently published in Epidemics, utilized two distinct statistical models to determine any association between pertussis vaccine coverage with incidence and sociodemographic factors. Summarizing their findings, Dr. Rane said “we showed that there exist local variations in DTaP [diphtheria, tetanus, and pertussis] vaccination coverage at the school-district level, and found that pertussis epidemics were more likely to occur in school-districts with lower vaccination coverage.” “Additionally, individuals living in areas of high socioeconomic deprivation could have lower access to vaccines and timely vaccination, leading to vaccination cold-spots. This could lead to persistence of pertussis, especially in high density areas. In our study, we found that median income and household size were associated with pertussis endemicity,” she added.

To establish the above findings, the authors gathered DTaP vaccine coverage data for children aged between 0-9 years old in King County WA, through the Washington State Immunization Information System. These data along with pertussis case data derived from the Public Health Seattle and King County Department of Communicable Diseases and Immunizations, allowed them to determine that “within King County, the Vashon Island, Enumclaw, and Tukwila regions had lower vaccination coverage compared to Seattle and Bellevue during the study period [2010-2017].  As the susceptible population accumulates in these areas over time, an introduction of a pertussis case could result in an outbreak”, said Dr. Rane. Importantly, these data also enabled the authors to determine that areas of high pertussis vaccine coverage (≥4 doses of DTaP) were linked to lower disease incidence. 

Low levels of pertussis vaccine coverage among children are associated with a risk of pertussis outbreak.
Pertussis incidence and vaccine coverage among children aged 19-35 months in King County, Washington. Figure from the original article (Open Access, creative commons license Attribution 4.0 International (CC BY 4.0))

To interrogate these data further in depth, the authors relied on two statistical models, an endemic-epidemic and an ecological vaccine model, as the use of both “can model occasional large outbreaks beyond regular endemic disease activity and can easily incorporate vaccination coverage as a covariate to study its effect on outbreaks. This makes them a good choice to model infectious disease data which is dependent data, and in particular pertussis outbreaks which have been sporadic in King County over the past decade,” according to Dr. Rane. The first model, “the endemic-epidemic model splits the disease data into the epidemic component which captures the disease peaks, and the endemic component captures the background cases in an area not explained by the epidemic component. We found that the epidemic risk of pertussis increases over 3-fold for every doubling of DTaP under-vaccination rate, in other words, lower vaccination coverage was associated with pertussis epidemicity. We also found that higher endemic risk of pertussis was associated with living in larger households and having lower median income. However, a major drawback of the endemic-epidemic model is that it is not suitable when individual-level inference is the goal of the analysis,” described Dr. Rane. To combat this, the second model, the ecological vaccine model “more appropriately models aggregate infectious disease data and provides less biased, epidemiologically interesting parameters such as vaccine effectiveness, as a measure of association between vaccination disease incidence. Thus, while both models showed that vaccination coverage was associated with disease incidence, the ecological vaccine model provided a more useful interpretation of this association in the form of vaccine effectiveness estimated to be 83%. Our findings suggest that 4 doses of DTaP vaccination are highly effective in preventing pertussis,” outlined Dr. Rane.

The authors went on to note how their study differs from previous research in this field as “while a lot of the focus has been on the role of the vaccine effectiveness of the acellular vaccine in pertussis resurgence, our study suggests that the vaccine is quite effective in preventing pertussis and local vaccination cold-spots could lead to occasional large pertussis outbreaks despite high state or county-level vaccination coverages”. Dr. Rane further described how their “findings emphasize the need to monitor sub-county level DTaP vaccination coverage to assist local health authorities target interventions to most affected areas, in addition to studying direct DTaP vaccine effectiveness and its waning.”

Lastly, for those undertaking research in this field, the authors have this advice going forward: “Vaccination and disease incidence data is often available aggregated in space and time. In such cases, when individual-level data is available, we suggest using the ecological vaccine model when the goal of the analysis is the estimate vaccine effectiveness to get less biased estimates.”

This work was funded by the National Institute of Allergy and Infectious Diseases of the National Institutes of Health.

Fred Hutch/University of Washington/Seattle Children's Cancer Consortium member M Elizabeth Halloran contributed to this work.

Rane MS, Wakefield J, Rohani P, Halloran ME. Association between pertussis vaccination coverage and other sociodemographic factors and pertussis incidence using surveillance data. Epidemics. 2023 May 18;44:100689. doi: 10.1016/j.epidem.2023.100689. Epub ahead of print. PMID: 37295130.