Background: Foodborne illnesses are an important public health problem in the United States. Investigation of foodborne disease outbreaks can lead to the identification of the causal etiology and food vehicle, which is important for informing food safety efforts. Characteristics of foodborne disease outbreaks that identify a causal etiology are well established; however, characteristics of outbreaks that identify a food vehicle are less understood. This study examined the outbreak characteristics and investigation methods associated with foodborne disease outbreaks where a food vehicle was identified.
Methods: Data were available from the National Outbreak Reporting System (NORS) for single-state outbreaks occurring in Colorado from 2009 – 2019. Outbreak characteristics and investigation methods included: pathogen (norovirus, bacteria, bacterial toxin, other, unknown), case count (≥5 outbreak-associate cases), lead agency (state or a local health department), age group (≥25% pediatric cases or ≥25% senior cases), hospitalized cases (% of cases hospitalized), investigation method (routine investigation, analytic study, other method), exposure setting (retail food establishment, private residences, other, institution, unknown), human specimen collection (≥ 2 specimens collected), number of primary lab confirmed cases per outbreak, and geographical region (multi-county exposure and urban or rural county). Univariate analysis and multivariate analysis (logistic regression) was performed to determine characteristics significantly associated with identification of a food vehicle overall and stratified by detection method (complaint or pathogen-specific surveillance).
Results: A total of 451 foodborne outbreak occurred in Colorado from 2009 – 2019; 121 multi-state outbreaks were excluded. Of 330 single state outbreaks, 153 (46%) identified a food vehicle. When compared to outbreak than did not identify a food vehicle, outbreak that identified a food vehicle were significantly more likely to include: a bacterial pathogen (25% vs. 12%), a larger case count (61% vs. 57% with ≥ 5 cases), investigation led by the state health department (18% vs. 5%), more hospitalized cases per outbreak (average 8% vs. 4%), conducting an analytic study (51% vs. 38%), exposure in a private residence (23% vs. 12%), and a multi-county exposure (9% vs 2%)). For outbreaks detected by complaint, characteristics were: bacterial pathogen, more hospitalized cases, conducting an analytic study, and exposure in a private resident. In contrast, characteristics significantly associated with identifying a food vehicle for outbreaks detected by pathogen-specific surveillance were: a larger case count, and having the investigation led by the state agency. In the multivariable analysis, bacterial outbreaks (Odds Ratio [OR]: 3.44, 95% Confidence Interval [CI]: 0.1.52-7.78), whereas case count (OR: 1.02, 95% CI: 1.00-1.04), state as the lead agency (OR: 2.99; 95% CI: 1.14-7.89) and exposure in private residences (OR: 2.17, 95% CI: 1.09-4.30) were significantly associated with identification of a food vehicle.
Conclusions: The characteristics significantly associated with identifying a food vehicle in Colorado single-state foodborne included pathogen, case count, and lead agency, specifically, larger outbreaks, bacterial etiology, and outbreaks led by state health departments. These findings will be translated into recommendations and best practice guidelines for outbreak training opportunities geared towards public health agencies.
Table of Contents
Burden of Foodborne Illnesses
Foodborne Disease Outbreaks
Detecting Foodborne Disease Outbreaks
Public Health Impact of Foodborne Disease Outbreaks
Solving Foodborne Outbreaks
Colorado Foodborne Disease Outbreaks
Purpose & Aims
TABLES & FIGURES
About this Master's Thesis
|Subfield / Discipline|
|Committee Chair / Thesis Advisor|
|File download under embargo until 13 January 2024||2021-12-02 14:41:38 -0500||File download under embargo until 13 January 2024|