Imperfect Detections of Influenza and Non-Influenza Respiratory Viruses among Children and Adults Hospitalized with Pneumonia Open Access

Abstract

Community-acquired pneumonia (CAP) is a common cause of hospitalization among all ages in the United States. Respiratory viruses can be detected using real-time polymerase chain reaction (PCR) and serology; limitations of these tests and available results can lead to misclassification of pneumonia attributable to respiratory viruses. The Centers for Disease Control and Prevention Etiology of Pneumonia in the Community (EPIC) study was a prospective, multi-center, population-based, active surveillance study. Using these data, this dissertation addressed potential misclassification in estimating the proportions of CAP due to respiratory viruses and predicting influenza-associated pneumonia. Missing serology results are a potentially important reason for misclassification of pneumonia etiology in the EPIC study. Among children and adults, 98.8-99.5% had naso-/oropharyngeal specimens and 37.5-48.6% had paired serum specimens available. We accounted for missing values using multiple imputation and estimated revised proportions of CAP with adenovirus, human metapneumovirus, parainfluenza virus, and respiratory syncytial virus detected (Aim 1), which were 14.4%, 15.0%, 8.9%, and 29.9%, respectively, among children. Among adults, the respective revised proportions were 4.2%, 4.7%, 4.0%, and 4.0%. These revised proportions were 0.8-3.2% higher than observed EPIC study estimates. In Aim 2, revised proportions of influenza virus detections were estimated to be 11.1% among children and 7.9% among adults, which were 2.1-4.4% higher than observed estimates. For Aim 3, we developed and evaluated prediction models and scores for influenza-associated pneumonia, using factors readily available at clinical presentation. Two definitions of influenza-associated pneumonia were used: one based on imputed results from PCR and serology, and the other was based on observed EPIC study results. Significant predictors were age and influenza season among children and leukocytosis, underlying medical conditions, cough, abdominal pain, and influenza season among adults. Prediction scores among children and adults had consistently high negative predictive values and, overall, low positive predictive values. By accounting for missing serology results, the proportions of CAP with specific respiratory viruses detected were higher than the observed results; thus, observed data may have underestimated the virus-specific burdens. Additionally, our prediction scores may reflect difficulty in predicting influenza-associated pneumonia and differentiating it from other causes of CAP, especially in children.

Table of Contents

CHAPTER 1. Background and research plan...1

Section 1. Background...1

1.1 Background of Community-Acquired Pneumonia (CAP)...1

1.1.1 Description of CAP...1

1.1.2 Case Definitions for Epidemiological Studies...3

1.2 Etiology of CAP in the United States Based on Diagnostic Tests...5

1.2.1 Available Specimens for Detection of Pathogens Associated with Pneumonia...6

1.2.2 Diagnostic Tests for Bacterial Pathogens...7

1.2.3 Diagnostic Tests for Viral Pathogens...11

1.2.4 Diagnostic Tests for Atypical Bacterial Pathogens...15

1.2.5 Etiology of CAP in Children and in Adults...17

1.2.6 Influenza-Associated Pneumonia...21

1.3 Latent Class Analysis...25

Section 2. Dissertation Research Plan...31

2.1 Objective and List of Specific Aims...31

2.2 Date Source: CDC Etiology of Pneumonia in the Community (EPIC) Study...33

2.3 Analytic Plan for Aims...36

Section 3. Novelty, Significance, and Impact of the Dissertation...49

CHAPTER 2. Use of multiple imputation and other methods to calculate revised proportions of non-influenza respiratory viral detections among children and adults hospitalized with community-acquired pneumonia...60

Abstract...60

Introduction...62

Materials and Methods...63

Results...69

Discussion...71

Supplementary Appendix...74

CHAPTER 3. Use of multiple imputation and other methods to calculate revised proportions of influenza virus detections among children and adults hospitalized with community-acquired pneumonia...89

Abstract...89

Introduction...91

Materials and Methods...92

Results...98

Discussion...100

Supplementary Appendix...103

CHAPTER 4. Prediction Models and Scores for Influenza-Associated Pneumonia among Children and Adults Hospitalized with Community-acquired Pneumonia...117

Abstract...117

Introduction...119

Materials and Methods...120

Results...126

Discussion...130

Supplementary Appendix...136

CHAPTER 5. Conclusions and Future Directions...150

Overview of Findings...150

Overall Strengths and Limitations of Data and Analyses...153

Implications and Future Directions...154

REFERENCES...157

About this Dissertation

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.

School	Laney Graduate School
Department	Epidemiology
Degree	PhD
Submission	Dissertation
Language	English
Research Field	Health Sciences, Public Health Health Sciences, Epidemiology
Keyword	Epidemiology Pneumonia Respiratory viruses Multiple imputation
Committee Chair / Thesis Advisor	Jain, Seema, Centers for Disease Control and Prevention Flanders, W Dana, Emory University
Committee Members	Gandhi, Neel R, Emory University Finelli, Lyn, Merck Vidal Graniel, Jorge, Emory University

Last modified

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	Imperfect Detections of Influenza and Non-Influenza Respiratory Viruses among Children and Adults Hospitalized with Pneumonia ()	2018-08-28 12:59:02 -0400	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions