Bayesian Correction of Misclassification of Pertussis in Vaccine Effectiveness Studies: How Much Does Underreporting...

Area of Science:

• Epidemiology
• Infectious Disease Research
• Vaccinology

Background:

• Accurate diagnosis of pertussis is challenging, potentially biasing disease risk research due to misclassification.
• Pertussis (whooping cough) diagnosis relies on case definitions that may not capture all true cases, impacting vaccine effectiveness studies.

Purpose of the Study:

• To quantify misclassification in pertussis diagnoses.
• To correct for misclassification bias in a case-control study evaluating pertussis vaccine effectiveness (VE).

Main Methods:

• A case-control study was conducted on children aged 3 months to 6 years diagnosed with pertussis between 2011 and 2013 in Philadelphia, Pennsylvania.
• Bayesian techniques were employed to adjust for nondifferential misclassification using the 2014 Council of State and Territorial Epidemiologists pertussis case definition.
• Vaccine effectiveness was calculated as (1 - odds ratio) × 100, comparing vaccinated and unvaccinated individuals.

Main Results:

• Initial (naïve) vaccine effectiveness (VE) was 50% (95% CI: 16%, 69%).
• After correction for misclassification, VE ranged from 57% (95% CrI: 30%, 73%) to 82% (95% CrI: 43%, 95%), depending on assumed underreporting levels.
• Misclassification, particularly false negatives, was significant, especially after incorporating infant apnea into the case definition. Case definition sensitivity varied widely (20%-90%).

Conclusions:

• Misclassification significantly impacts pertussis research and vaccine effectiveness evaluations.
• Accurate assessment of underreporting is essential for reliable pertussis VE estimates.
• Adjusting for diagnostic misclassification provides more accurate insights into the protective effect of pertussis vaccines.