Epidemiologic Profiles: Clinical and Epidemiologic Profiles for Norovirus Outbreaks Open Access

Yu, Joana (2015)

Permanent URL: https://etd.library.emory.edu/concern/etds/7d278t32v?locale=en


Background: Noroviruses are the leading cause of acute gastroenteritis outbreaks and foodborne disease outbreaks in the United States. Laboratory testing for norovirus during outbreak investigations has historically been limited by the availability of molecular-based diagnostics and resource constraints at state and local health departments. In the absence of laboratory confirmation, clinical and epidemiologic profiles, such as the Kaplan criteria (vomiting in ≥50% of cases, mean incubation period of 24-48 h, mean duration of illness 12-60 h, and negative bacterial culture) and the ratios of fever-to-vomiting and diarrhea-to-vomiting have proven useful in distinguishing norovirus from bacterial agents.

Methods: Previously proposed clinical and epidemiologic profiles were reevaluated with outbreaks occurring during 2009-2012 and reported through the National Outbreak Reporting System (NORS), specifically those with the following etiologies: laboratory confirmed norovirus (N=2,939), suspected norovirus (N=1,321), laboratory confirmed non-viral (N=1,544), and unknown etiology (N=3,694). Alternative clinical and epidemiologic profiles were developed and evaluated with classification and regression tree (CART) modeling. The performance of previous profiles and the CART predictors was evaluated by Cohen's kappa statistic, as well as the proportion of outbreaks with all criteria reported, sensitivity, specificity and the likelihood ratio.

Results: The Kaplan criteria remained highly specific (100%, 95% CI: 83.2%-100%) with a Cohen's kappa of 0.34 but only 108 (3.7%) confirmed norovirus and 19 (1.2%) confirmed non-viral outbreaks had all information for the criteria reported. With CART modeling, an alternative clinical and epidemiologic profile was developed with the fever-to-vomit ratio <1, the proportion of cases with vomiting ≤0.34, and the proportion of cases with bloody stool <0.12. The CART predictors had a high likelihood ratio of 12.5, a Cohen's kappa of 0.78 (95% CI: 0.75-0.81), and 706 (24.0%) confirmed norovirus and 605 (39.1%) confirmed non-viral outbreaks had information for all criteria reported.

Conclusion: Relative to the Kaplan criteria, the CART predictors were similarly effective in distinguishing norovirus from non-viral outbreaks, but were reported far more frequently in NORS. These predictors provide a useful alternative profile for identifying likely norovirus etiology during outbreak investigations.

Table of Contents

Chapter I . 1

Background. 1

Global Impact of Norovirus. 1

Norovirus in the United States. 1

Biology and Epidemiology of Norovirus. 3

Surveillance Systems for Norovirus in the United States. 4

Outbreak Investigations in the United States. 6

Challenges with Diagnosing Norovirus. 7

Clinical and Epidemiological Criteria. 8

Classification and Regression Tree Modeling. 10

Conclusions. 11

Chapter II. 13

Manuscript. 13

Introduction. 13

Methods. 17

Results. 23

Discussion. 29

Tables & Figures. 34

Chapter III. .40

Public Health Implications. 40

References. 41

Appendices. 48

A. NORS form. 48

B. Alternative CART models. 54

About this Master's Thesis

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
  • English
Research Field
Committee Chair / Thesis Advisor
Committee Members
Partnering Agencies
Last modified

Primary PDF

Supplemental Files