Principles and Practice of Pediatric Infectious Diseases 6e May 16 2022 - 0323756085 - Elsevier 6th Edition Sarah S. Long Download PDF
Principles and Practice of Pediatric Infectious Diseases 6e May 16 2022 - 0323756085 - Elsevier 6th Edition Sarah S. Long Download PDF
com
                    DOWLOAD EBOOK
                             OR CLICK LINK
       http://ebookstep.com/product/principles-and-
     practice-of-pediatric-infectious-diseases-6e-may-
    16-2022-_-0323756085-_-elsevier-6th-edition-sarah-
                           s-long/
      https://ebookstep.com/product/house-of-flame-and-shadow-sarah-j-
      maas/
      https://ebookstep.com/product/biocontrol-of-plant-diseases-by-
      bacillus-subtilis-basic-and-practical-applications-1st-edition-
      makoto-shoda-author/
      https://ebookstep.com/product/illustrated-handbook-of-rheumatic-
      and-musculo-skeletal-
      diseases-2e-mar-2-2024_3031473787_springer-2nd-edition-pelechas/
      https://ebookstep.com/product/principles-of-corporate-
      finance-13th-edition-brealey/
Neurología (Sexto edicíon; Elsevier) 6th Edition Juan
José Zarranz
https://ebookstep.com/product/neurologia-sexto-edicion-
elsevier-6th-edition-juan-jose-zarranz/
https://ebookstep.com/product/house-of-flame-and-shadow-ciudad-
medialuna-3-sarah-j-maas/
https://ebookstep.com/product/the-perils-of-pleasure-julie-anne-
long/
https://ebookstep.com/product/principles-of-corporate-finance-
mcgraw-hill-irwin-series-in-finance-insurance-and-real-
estate-12th-edition-brealey/
https://ebookstep.com/product/solution-manual-for-elements-of-
electromagnetics-6e-sodiku/
Principles and Practice of
PEDIATRIC
INFECTIOUS DISEASES
                                                                          SIXTH EDITION
                                       EDITOR
                                SARAH S. LONG, md
                                  Professor of Pediatrics
                          Drexel University College of Medicine
                      Chief Emeritus, Section of Infectious Diseases
                          St. Christopher’s Hospital for Children
                                Philadelphia, Pennsylvania
                                ASSOCIATE EDITORS
                              CHARLES G. PROBER, md
                  Professor of Pediatrics, Microbiology, and Immunology
             Founding Executive Director, Stanford Center for Health Education
                    Senior Associate Vice Provost for Health Education
                         Stanford University School of Medicine
                                   Stanford, California
                              DAVID W. KIMBERLIN, md
              Professor and Vice Chair for Clinical and Translational Research
                    Co-Director, Division of Pediatric Infectious Diseases
                                  Department of Pediatrics
             University of Alabama at Birmingham Heersink School of Medicine
                            and the Children’s Hospital of Alabama
                                    Birmingham, Alabama
Elsevier
1600 John F. Kennedy Blvd.
Ste 1600
Philadelphia, PA 19103-2899
Previous editions copyrighted 2018, 2012, 2008, 2003, 1997 by Elsevier, Inc.
No part of this publication may be reproduced or transmitted in any form or by any means, electronic or
mechanical, including photocopying, recording, or any information storage and retrieval system, without
permission in writing from the publisher. Details on how to seek permission, further information about the
Publisher’s permissions policies, and our arrangements with organizations such as the Copyright Clearance
Center and the Copyright Licensing Agency can be found at our website: www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by the Publisher
(other than as may be noted herein).
All chapters published herein that are authored or co-authored by an employee of the US government are in
the public domain. In addition, the following chapters are in the public domain:
Chapter 113: Management of HIV Infection
Chapter 218: Flaviviruses
Chapter 278: Diphyllobothriidae, Dipylidium and Hymenolepis Species
Chapter 285: Blood Trematodes: Schistosomiasis
Notice
 Practitioners and researchers must always rely on their own experience and knowledge in evaluating
 and using any information, methods, compounds, or experiments described herein. Because of rapid
 advances in the medical sciences in particular, independent verification of diagnoses and drug dosages
 should be made. To the fullest extent of the law, no responsibility is assumed by Elsevier, authors, editors,
 or contributors for any injury and/or damage to persons or property as a matter of products liability,
 negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas
 contained in the material herein.
ISBN-13: 978-0-323-75608-2
Printed in India
                                                                                                                                                              v
                         Acknowledgments
                                              With special contributions of clinical images by James H. Brien, DO,
     Adjunct Professor of Pediatrics, Infectious Diseases, Texas A&M University College of Medicine, McLane Children’s Hospital, Baylor Scott & White
                                                                     Health, Temple, Texas
vi
                  With our spouses (Bob, Laura, Lisa, and Kim)
      our children (Stephen, Suzanne, and Caroline; Meghan and Andrew;
               Sydney and Madison; Will, Claire, and Katherine)
                              and other loved ones
                 whose patience and endurance are our bedrock,
                                                                                       vii
                               Contributors
       Mark J. Abzug, MD                                    Upton D. Allen, Jr., MBBS, MSc                    Shai Ashkenazi, MD, MSc
       Professor of Pediatrics–Infectious Diseases,         Professor of Paediatrics, University of Toronto   Professor and Dean, Adelson School of
       University of Colorado School of Medicine;           Faculty of Medicine; Division of Infectious       Medicine, Ariel University, Department of
       Division of Infectious Diseases, Children’s          Diseases, Hospital for Sick Children, Toronto,    Pediatrics A, Schneider Children’s Medical
       Hospital Colorado, Aurora, Colorado                  Ontario, Canada                                   Center, Petach Tikva, Israel
       Introduction to Picornaviridae; Enteroviruses        Adenoviruses                                      Plesiomonas shigelloides; Shigella Species
       and Parechoviruses
                                                            Gerardo Alvarez-Hernández, MD, PhD, MPH           Liat Ashkenazi-Hoffnung, MD
       Elisabeth E. Adderson, MD, MSc                       Professor of Medicine and Health Sciences,        Sackler Faculty of Medicine, Tel-Aviv
       Associate Professor of Pediatrics, University of     Universidad de Sonora, Hermosillo, Mexico         University, Tel-Aviv, Israel; Head, Day
       Tennessee Health Sciences Center; Associate          Rickettsia rickettsii (Rocky Mountain Spotted     Hospitalization Department, Attending
       Member, Department of Infectious Diseases,           Fever)                                            Physician, Pediatric Infectious Diseases Unit,
       St. Jude Children’s Research Hospital and St.                                                          Schneider Children’s Medical Center, Petach-
       Jude Graduate School of Biomedical Sciences,         Krow Ampofo, MBChB                                Tikva, Israel
       Memphis, Tennessee                                   Professor of Pediatrics, University of Utah       Plesiomonas shigelloides
       Infectious Complications of Antibody                 School of Medicine; Attending Physician,
       Deficiency                                           Infectious Diseases, Primary Children’s           Edwin J. Asturias, MD
                                                            Hospital, Salt Lake City, Utah                    Professor of Pediatrics, University of Colorado
       Aastha Agarwal, MD                                   Streptococcus pneumoniae                          School of Medicine; Professor of Epidemiology,
       Senior Resident, Dermatology, Dr RML                                                                   Associate Director, Center for Global Health,
       Hospital, ABVIMS, Delhi, India                       Evan J. Anderson, MD                              Colorado School of Public Health; Jules Amer
       Superficial Fungal Infections                        Professor of Pediatrics and Medicine, Emory       Chair, Department of Community Pediatrics,
                                                            University School of Medicine; Division of        Children’s Hospital Colorado, Aurora,
       Allison L. Agwu, MD, ScM                             Infectious Diseases, Children’s Healthcare of     Colorado
       Professor of Pediatrics, Johns Hopkins School        Atlanta, Atlanta, Georgia                         Escherichia coli
       of Medicine, Baltimore, Maryland                     Campylobacter jejuni and Campylobacter coli;
       Infectious Complications of HIV Infection            Other Campylobacter Species                       Kestutis Aukstuolis, DO
                                                                                                              Division of Allergy and Infectious Diseases,
       Lindsey Albenberg, DO                                Grace D. Appiah, MD, MS                           Department of Medicine, University of
       Assistant Professor of Pediatrics, University        Medical Epidemiologist, Waterborne Disease        Washington, Seattle, Washington
       of Pennsylvania Perelman School of                   Prevention Branch, Centers for Disease Control    Immunologic Development and Susceptibility
       Medicine; Attending Physician, Division of           and Prevention, Atlanta, Georgia                  to Infection
       Gastroenterology, Hepatology, and Nutrition,         Vibrio cholerae (Cholera)
       Children’s Hospital of Philadelphia, Philadelphia,                                                     Vahe Badalyan, MD, MPH, MBA
       Pennsylvania                                         Monica I. Ardura, DO, MSCS                        Attending Physician, Department of
       Anaerobic Bacteria: Clinical Concepts and the        Associate Professor of Pediatrics, The Ohio       Gastroenterology, Hepatology, and Nutrition,
       Microbiome in Health and Disease                     State University College of Medicine; Medical     Children’s National Hospital, Washington,
                                                            Director, Host Defense Program, Division of       District of Columbia
       Jonathan Albert, MD                                  Infectious Diseases, Nationwide Children’s        Acute Hepatitis
       Fellow, Division of Infectious Diseases,             Hospital, Columbus, Ohio
       Children’s National Hospital, Washington,            Risk Factors and Infectious Agents in Children    Carol J. Baker, MD
       District of Columbia                                 With Cancer: Fever and Granulocytopenia;          Professor of Pediatrics–Infectious Diseases,
       Peritonitis                                          Clinical Syndromes of Infection in Children       McGovern Medical School, University of Texas
                                                            With Cancer; Citrobacter Species                  Health Science Center, Houston, Texas
       Kevin Alby, PhD, D(ABMM)                                                                               Bacterial Infections in the Neonate;
       Assistant Professor of Pathology and                 Stephen S. Arnon, MD, MPH                         Streptococcus agalactiae (Group B
       Laboratory Medicine, University of North             Founder and Chief, Infant Botulism Treatment      Streptococcus)
       Carolina School of Medicine; Director,               and Prevention Program, California
       Bacteriology and Susceptibility Testing, UNC         Department of Public Health, Richmond,            Karthik Balakrishnan, MD, MPH
       Medical Center, Chapel Hill, North Carolina          California                                        Associate Professor of Otolaryngology,
       Mechanisms and Detection of Antimicrobial            Clostridium botulinum (Botulism)                  Stanford University School of Medicine,
       Resistance                                                                                             Lucile Packard Children’s Hospital, Stanford,
                                                            Naomi E. Aronson, MD                              California
       Grace M. Aldrovandi, MD, CM                          Professor of Medicine, Director, Infectious       Otitis Externa and Necrotizing Otitis Externa
       Professor of Pediatrics, David Geffen School         Diseases Division, Uniformed Services
       of Medicine at UCLA; Chief, Division of              University of the Health Sciences, Bethesda,      Elizabeth D. Barnett, MD
       Infectious Diseases, UCLA Mattel Children’s          Maryland                                          Professor of Pediatrics, Boston University
       Hospital, Los Angeles, California                    Leishmania Species (Leishmaniasis)                School of Medicine; Attending Physician,
       Immunopathogenesis of HIV-1 Infection                                                                  Department of Pediatrics, Boston Medical
                                                            Ann M. Arvin, MD                                  Center, Boston, Massachusetts
                                                            Lucile Salter Packard Professor of Pediatrics,    Infectious Diseases in Refugee and
                                                            Professor of Microbiology and Immunology,         Internationally Adopted Children; Protection
                                                            Stanford University School of Medicine,           of Travelers
                                                            Stanford, California
                                                            Varicella-Zoster Virus
viii
                                                                                                                                  Contributors
                                                                                                                                                         ix
        Contributors
    Rebecca J. Chancey, MD                             C. Buddy Creech, MD, MPH                          H. Dele Davies, MD, MHCM
    Medical Officer, Parasitic Diseases Branch,        Professor in Pediatrics, Director, Vanderbilt     Professor of Pediatrics and Public Health,
    Center for Global Health, Centers for Disease      Vaccine Research Program, Edie Carell             University of Nebraska Medical Center,
    Control and Prevention, Atlanta, Georgia           Johnson Chair, Division of Pediatric Infectious   Omaha, Nebraska
    Clonorchis, Opisthorchis, Fascioloa, and           Diseases, Vanderbilt University School of         Infections Related to Biologic Response
    Paragonimus Species; Antiparasitic Agents          Medicine and Medical Center, Nashville,           Modifying Drug Therapy; Infectious
                                                       Tennessee                                         Complications of Corticosteroid Therapy
    Cara C. Cherry, DVM, MPH, DACVPM                   Musculoskeletal Symptom Complexes;
    Veterinary Epidemiologist, Rickettsial             Myositis, Pyomyositis, and Necrotizing            Fatimah S. Dawood, MD
    Zoonoses Branch, Centers for Disease Control       Fasciitis; Transient Synovitis; Staphylococcus    Medical Officer, Influenza Division, Centers
    and Prevention, Atlanta, Georgia                   aureus                                            for Disease Control and Prevention, Atlanta,
    Coxiella burnetii (Q fever)                                                                          Georgia
                                                       Jonathan D. Crews, MD, MS                         Influenza Viruses
    Silvia S. Chiang, MD                               Assistant Professor of Pediatric–Infectious
    Assistant Professor of Pediatrics, Warren Alpert   Diseases, Baylor College of Medicine/             J. Christopher Day, MD
    Medical School of Brown University; Attending      Children’s Hospital of San Antonio, San           Assistant Professor of Pediatrics, University of
    Physician, Infectious Diseases, Hasbro             Antonio, Texas                                    Missouri at Kansas City; Attending Physician,
    Children’s Hospital, Providence, Rhode Island      Other Gastric and Enterohepatic Helicobacter      Division of Infectious Diseases, Children’s
    Mycobacterium tuberculosis                         Species                                           Mercy–Kansas City, Kansas City, Missouri
                                                                                                         Lymphatic System and Generalized
    Mary Choi, MD, MPH                                 Donna Curtis, MD, MPH                             Lymphadenopathy; Mediastinal and Hilar
    Medical Officer, Viral Special Pathogens           Associate Professor of Pediatrics, University     Lymphadenopathy
    Branch, Centers for Disease Control and            of Colorado School of Medicine; Section
    Prevention, Atlanta, Georgia                       of Infectious Diseases, Children’s Hospital       M. Teresa de la Morena, MD
    Bunyaviruses; Filoviruses and Arenaviruses         Colorado, Aurora, Colorado                        Professor of Pediatrics, University of
                                                       Pneumonia in the Immunocompromised Host           Washington School of Medicine; Division
    John C. Christenson, MD                                                                              of Immunology, Seattle Children’s Hospital,
    Professor of Clinical Pediatrics, Ryan White       Nigel Curtis, DCH, DTM&H, MRCP,                   Seattle, Washington
    Center for Pediatric Infectious Diseases and       MRCPCH, PhD                                       Immunologic Development and Susceptibility
    Global Health, Indiana University School           Professor of Paediatrics, The University of       to Infection
    of Medicine; Riley Hospital for Children,          Melbourne; Head of Infectious Diseases,
    Indianapolis, Indiana                              The Royal Children’s Hospital Melbourne,          Gregory P. DeMuri, MD
    Histoplasma capsulatum (Histoplasmosis);           Leader of Infectious Diseases Group, Murdoch      Professor of Pediatrics, University of Wisconsin
    Laboratory Diagnosis of Infection Due to           Children’s Research Institute, Parkville,         School of Medicine and Public Health; Division
    Bacteria, Fungi, Parasites, and Rickettsiae        Victoria, Australia                               of Infectious Diseases, American Family
                                                       Infections Related to the Upper and Middle        Children’s Hospital, Madison, Wisconsin
    Susan E. Coffin, MD, MPH                           Airways; Mycobacterium Nontuberculosis            Sinusitis; Preseptal and Orbital Infections
    Professor of Pediatrics, Perelman School of        Species
    Medicine at the University of Pennsylvania;                                                          Dickson D. Despommier, PhD
    Division of Infectious Diseases, Children’s        Lara A. Danziger-Isakov, MD, MPH                  Emeritus Professor of Microbiology and
    Hospital of Philadelphia, Philadelphia,            Professor of Pediatrics, University of            Immunology, Columbia University College of
    Pennsylvania                                       Cincinnati College of Medicine; Division of       Physicians and Surgeons, New York, New York
    Healthcare-Associated Infections                   Infectious Diseases, Cincinnati Children’s        Tissue Nematodes
                                                       Hospital Medical Center, Cincinnati, Ohio
    Amanda Cohn, MD                                    Infectious Complications in Special Hosts         Daniel S. Dodson, MS, MD
    Chief Medical Officer, National Center for                                                           Assistant Professor of Pediatrics, University of
    Immunization and Respiratory Diseases,             Toni Darville, MD                                 California Davis School of Medicine; Division
    Centers for Disease Control and Prevention,        Professor of Pediatrics, University of North      of Pediatric Infectious Diseases, UC Davis
    Atlanta, Georgia                                   Carolina at Chapel Hill; Chief, Infectious        Health, Sacramento, California
    Neisseria meningitidis                             Diseases, Department of Pediatrics, North         Escherichia coli
                                                       Carolina Children’s Hospital, Chapel Hill,
    Despina G. Contopoulos-Ioannidis, MD               North Carolina                                    Stephen J. Dolgner, MD
    Clinical Associate Professor of Pediatrics,        Chlamydia trachomatis                             Assistant Professor of Pediatrics, Baylor College
    Division of Infectious Diseases, Stanford                                                            of Medicine; Division of Cardiology, Texas
    University School of Medicine, Stanford,           Gregory A. Dasch, PhD                             Children’s Hospital, Houston, Texas
    California                                         Biotechnology Applications Laboratory             Endocarditis and Other Intravascular
    Toxoplasma gondii (Toxoplasmosis)                  Director, Rickettsial Zoonoses Branch, Centers    Infections
                                                       for Disease Control and Prevention, Atlanta,
    James H. Conway, MD                                Georgia                                           Clinton Dunn, MD
    Professor of Pediatrics, University of Wisconsin   Other Rickettsia Species                          Department of Medicine, Division of Allergy
    School of Medicine & Public Health; Associate                                                        and Infectious Diseases, University of
    Director, Global Health Institute, University of   Irini Daskalaki, MD                               Washington, Seattle, Washington
    Wisconsin–Madison, Madison, Wisconsin              Global and Community Health Physician             Immunologic Development and Susceptibility
    Mastoiditis                                        Coordinator, Princeton University Health          to Infection
                                                       Services, Princeton, New Jersey
    Margaret M. Cortese, MD                            Corynebacterium diphtheriae; Leptospira           Jonathan Dyal, MD, MPH
    Captain, United States Public Health Service,      Species (Leptospirosis); Other Borrelia Species   Epidemic Intelligence Service Officer, Viral
    Division of Viral Diseases, National Center        and Spirillum minus                               Special Pathogens Branch, Centers for Disease
    for Immunization and Respiratory Diseases,                                                           Control and Prevention, Atlanta, Georgia
    Centers for Disease Control and Prevention,                                                          Filoviruses and Arenaviruses
    Atlanta, Georgia
    Rotaviruses
x
                                                                                                                                  Contributors
†Deceased.
                                                                                                                                                          xi
          Contributors
xii
                                                                                                                                      Contributors
Julia C. Haston, MD, MSc                            Peter J. Hotez, MD, PhD                               Sophonie Jean, PhD, D(ABMM)
EIS Officer, Division of Foodborne,                 Dean, National School of Tropical Medicine,           Assistant Professor of Pathology and
Waterborne, and Environmental Diseases,             Professor of Pediatrics and Molecular Virology        Laboratory Medicine, The Ohio State
National Center for Emerging and Zoonotic           & Microbiology, Head, Section of Pediatric            University College of Medicine; Clinical
Infectious Diseases, Centers for Disease            Tropical Medicine, Baylor College of Medicine;        Microbiology and Immunoserology
Control and Prevention, Atlanta, Georgia            Texas Children’s Hospital Endowed Chair               Laboratories, Nationwide Children’s Hospital,
Trypanosoma Species (Trypanosomiasis)               of Tropical Pediatrics, Director, Center for          Columbus, Ohio
                                                    Vaccine Development, Texas Children’s                 Enterobacter, Cronobacter, and Pantoea
Sarah.J. Hawkes, MBBS, PhD                          Hospital, Baker Institute Fellow in Disease and       Species; Citrobacter Species; Less Commonly
Professor of Global Public Health, Institute        Poverty, Rice University, Houston, Texas              Encountered Enterobacterales
for Global Health, University College London,       Classification of Parasites; Intestinal
London, United Kingdom                              Nematodes                                             Ravi Jhaveri, MD
Treponema pallidum (Syphilis)                                                                             Professor of Pediatrics, Northwestern
                                                    Katherine K. Hsu, MD, MPH                             University Feinberg School of Medicine;
Taylor Heald-Sargent, MD, PhD                       Professor of Pediatrics–Pediatric Infectious          Interim Division Head, Division of Pediatric
Assistant Professor of Pediatrics, Northwestern     Diseases, Boston University Medical Center,           Infectious Diseases, Ann & Robert H. Lurie
University Feinberg School of Medicine;             Boston, Massachusetts; Medical Director,              Children’s Hospital of Chicago, Chicago,
Attending Physician, Division of Infectious         Division of STD Prevention, Bureau of                 Illinois
Diseases, Ann & Robert H. Lurie Children’s          Infectious Diseases Prevention, Response, and         Fever Without Localizing Signs
Hospital of Chicago, Chicago, Illinois              Services, Massachusetts Department of Public
Streptococcus pyogenes (Group A Streptococcus)      Health, Jamaica Plain, Massachusetts                  Kateřina Jirků-Pomajbíková, DVM, PhD
                                                    Neisseria gonorrhoeae; Other Neisseria Species        Researcher, Institute of Parasitology, Biology
†J.   Owen Hendley, MD                                                                                    Centre of the Czech Academy of Sciences,
Professor of Pediatric Infectious Diseases,         Felicia Scaggs Huang, MD, MSc                         České Budějovice, Czech Republic
University of Virginia, Charlottesville, Virginia   Assistant Professor of Pediatrics, University         Balantioides coli (Formerly Balantidium coli)
Rhinoviruses                                        of Cincinnati College of Medicine; Attending
                                                    Physician, Division of Infectious Diseases,           Nadia A. Kadry, PhD
Adam L. Hersh, MD, PhD                              Cincinnati Children’s Hospital Medical Center,        Cell and Molecular Biology, University of
Professor of Pediatrics, University of Utah         Cincinnati, Ohio                                      Pennsylvania, Philadelphia, Pennsylvania
School of Medicine; Pediatric Infectious            Clinical Syndromes of Device-Associated               Haemophilus influenzae
Diseases, Primary Children’s Hospital and the       Infections
University of Utah Health, Salt Lake City, Utah                                                           Mary L. Kamb, MD, MPH
Principles of Anti-Infective Therapy                David A. Hunstad, MD                                  Team Lead, Elimination and Control
                                                    Professor of Pediatrics and Molecular                 Epidemiology, Parasitic Diseases Branch,
Joseph A. Hilinski, MD                              Microbiology, Washington University School            Division of Parasitic Diseases and Malaria,
Clinical Associate Professor of Pediatrics,         of Medicine; Service Chief, Pediatric Infectious      Centers for Disease Control and Prevention,
University of Washington School of Medicine,        Diseases, St. Louis Children’s Hospital, St. Louis,   Atlanta, Georgia
Seattle, Washington; Attending Physician,           Missouri                                              Blood and Tissue Nematodes: Filarial Worms
Pediatric Infectious Diseases/Children’s            Infectious Complications in Special Hosts
Infections and Immune Deficiency Clinic, St.                                                              Ronak K. Kapadia, MD, MSc, BSc
Luke’s Children’s Hospital, Boise, Idaho            W. Garrett Hunt, MD, MPH, DTM&H                       Assistant Professor of Clinical Neurosciences,
Myocarditis; Pericarditis                           Associate Professor of Pediatrics, The Ohio           University of Calgary, Calgary, Alberta, Canada
                                                    State University College of Medicine; Section         Parainfectious and Postinfectious Neurologic
Susan L. Hills, MBBS, MTH                           of Infectious Diseases, Nationwide Children’s         Syndromes
Medical Epidemiologist, Arboviral Diseases          Hospital, Columbus, Ohio
Branch, Centers for Disease Control and             Enterobacter, Cronobacter, and Pantoea Species        Ben Z. Katz, MD
Prevention, Fort Collins, Colorado                                                                        Professor of Pediatrics, Northwestern
Flaviviruses                                        Loris Y. Hwang, MD                                    University Feinberg School of Medicine;
                                                    Associate Professor of Pediatrics, David              Attending Physician, Division of Infectious
David K. Hong, MD                                   Geffen School of Medicine at the University of        Diseases, Ann & Robert H. Lurie Children’s
Associate Professor (Affiliated) of Pediatrics–     California, Los Angeles; Division of Adolescent       Hospital of Chicago, Chicago, Illinois
Infectious Diseases, Stanford University School     and Young Adult Medicine, University of               Epstein-Barr Virus (Mononucleosis and
of Medicine, Stanford, California; Head of          California Los Angeles, Mattel Children’s             Lymphoproliferative Disorders)
Viral Respiratory Infections R&D, Janssen           Hospital UCLA, Los Angeles, California
Pharmaceutical Companies of Johnson &               Human Papillomavirus                                  Sophie E. Katz, MD, MPH
Johnson, South San Francisco, California; Staff                                                           Assistant Professor of Pediatrics, Division of
Physician, Pediatric Infectious Diseases, Santa     Christelle M. Ilboudo, MD                             Infectious Diseases, Vanderbilt University
Clara Valley Medical Center, San Jose, California   Assistant Professor of Child Health, University       Medical Center, Nashville, Tennessee
Osteomyelitis; Infectious and Inflammatory          of Missouri Health Care, Columbia, Missouri           Myositis, Pyomyositis, and Necrotizing
Arthritis; Diskitis                                 Other Gram-Positive Bacilli                           Fasciitis; Transient Synovitis
                                                                                                                                                            xiii
          Contributors
      Gilbert J. Kersh, PhD                               Andrew T. Kroger, MD, MPH                            Eloisa Llata, MD, MPH
      Branch Chief, Rickettsial Zoonoses Branch,          Medical Officer, National Center for                 Medical Epidemiologist, Division of STD
      Centers for Disease Control and Prevention,         Immunization and Respiratory Diseases,               Prevention, Centers for Disease Control and
      Atlanta, Georgia                                    Centers for Disease Control and Prevention,          Prevention, Atlanta, Georgia
      Coxiella burnetii (Q fever)                         Atlanta, Georgia                                     Pelvic Inflammatory Disease
                                                          Active Immunization
      Muhammad Ali Khan, MD, MPH                                                                               Kevin Lloyd, MD
      Assistant Professor of Pediatrics, George           Matthew P. Kronman, MD, MSCE                         Research Instructor, Department of
      Washington University School of                     Associate Professor of Pediatrics, University of     Pediatrics, George Washington University
      Medicine and Health Sciences; Division of           Washington School of Medicine; Division of           School of Medicine and Health Sciences;
      Gastroenterology, Hepatology, and Nutrition,        Infectious Diseases, Seattle Children’s Hospital,    Fellow, Division of Infectious Diseases,
      Children’s National Hospital, Washington,           Seattle, Washington                                  Children’s National Hospital, Washington,
      District of Columbia                                Endocarditis and Other Intravascular                 District of Columbia
      Chronic Hepatitis                                   Infections                                           Appendicitis
xiv
                                                                                                                                        Contributors
                                                                                                                                                              xv
          Contributors
xvi
                                                                                                                                Contributors
Mikael Petrosyan, MD, MBA                        Susan M. Poutanen, MD, MPH                         Adam J. Ratner, MD, MPH
Associate Professor of General and Thoracic      Associate Professor of Laboratory Medicine         Associate Professor of Pediatrics and
Surgery, George Washington University School     and Pathobiology & Medicine, University            Microbiology, NYU Grossman School
of Medicine; Associate Chief, Department         of Toronto Faculty of Medicine; Medical            of Medicine; Chief, Division of Pediatric
of General and Thoracic Surgery, Children’s      Microbiologist & Infectious Diseases Physician,    Infectious Diseases, NYU Langone Health, New
National Medical Center, Washington, District    Mount Sinai Hospital & University Health           York, New York
of Columbia                                      Network, Toronto, Ontario, Canada                  Sexually Transmitted Infection Syndromes;
Appendicitis                                     Human Coronaviruses                                Epididymitis, Orchitis, and Prostatitis
                                                                                                                                                        xvii
            Contributors
        Shannon A. Ross, MD, MSPH                          Kabir Sardana, MD, DNB, MNAMS                     Samir S. Shah, MD, MSCE
        Associate Professor of Pediatrics, University      Professor and Director of Dermatology, Dr         Professor of Pediatrics, University of
        of Alabama at Birmingham Heersink School           RML Hospital, ABVIMS, Delhi, India                Cincinnati College of Medicine; Director,
        of Medicine; Division of Infectious Diseases,      Superficial Fungal Infections                     Division of Hospital Medicine, Attending
        Children’s Hospital of Alabama at the                                                                Physician in Infectious Diseases and Hospital
        University of Alabama, Birmingham, Alabama         Jason B. Sauberan, PharmD                         Medicine, James M. Ewell Endowed Chair,
        Cytomegalovirus                                    Clinical Research Pharmacist, Neonatal            Cincinnati Children’s Hospital Medical Center,
                                                           Research Institute, Sharp Mary Birch Hospital     Cincinnati, Ohio
        G. Ingrid J.G. Rours, MD, PhD, MSc                 for Women and Newborns; Consultant                Acute Pneumonia and Its Complications;
        Pediatrics, Kinderplein, Medical Center            Pharmacist, Helen Bernardy Center for             Chlamydophila (Chlamydia) pneumoniae;
        for Quality of Life; Physician, Medical            Medically Fragile Children, Rady Children’s       Mycoplasma pneumoniae; Other Mycoplasma
        Microbiology and Infectious Diseases, Erasmus      Hospital, San Diego, California                   Species; Ureaplasma urealyticum
        Medical Center, Rotterdam, Zuid Holland, the       Pharmacokinetic and Pharmacodynamic
        Netherlands                                        Basis of Optimal Antimicrobial Therapy;           Nader Shaikh, MD, MPH
        Chlamydia trachomatis                              Antibacterial Agents                              Professor of Pediatrics, University of Pittsburgh
                                                                                                             School of Medicine; Attending Physician,
        Peter C. Rowe, MD                                  Joshua K. Schaffzin, MD, PhD                      UMPC Children’s Hospital of Pittsburgh,
        Professor of Pediatrics, Johns Hopkins             Associate Professor of Pediatrics, University     Pittsburgh, Pennsylvania
        University School of Medicine; Division            of Cincinnati College of Medicine; Division       Otitis Media
        of Adolescent and Young Adult Medicine,            of Infectious Diseases, Director, Infection
        Department of Pediatrics, Johns Hopkins            Prevention & Control Program, Cincinnati          Andi L. Shane, MD, MPH, MSc
        Children’s Center, Baltimore, Maryland             Children’s Hospital Medical Center,               Marcus Professor of Hospital Epidemiology
        Myalgic Encephalomyelitis/Chronic Fatigue          Cincinnati, Ohio                                  and Infection Prevention, Emory University
        Syndrome (ME/CFS)                                  Granulomatous Hepatitis; Acute Pancreatitis;      School of Medicine and Children’s Healthcare
                                                           Cholecystitis and Cholangitis; Clinical           of Atlanta, Atlanta, Georgia
        Anne H. Rowley, MD                                 Syndromes of Device-Associated Infections         Infections Associated With Group Childcare;
        Professor of Pediatrics and Microbiology/                                                            Approach to the Diagnosis and Management of
        Immunology, Northwestern University                Sarah Schillie, MD, MPH, MBA                      Gastrointestinal Tract Infections
        Feinberg School of Medicine; Attending             Medical Epidemiologist, Division of Viral
        Physician, Infectious Diseases, The Ann            Hepatitis, Centers for Disease Control and        Eugene D. Shapiro, MD
        & Robert H. Lurie Children’s Hospital of           Prevention, Atlanta, Georgia                      Professor of Pediatrics and Epidemiology
        Chicago, Chicago, Illinois                         Hepatitis B and Hepatitis D Viruses               (Microbial Diseases and Investigative
        Kawasaki Disease                                                                                     Medicine), Yale University School of Medicine;
                                                           Jennifer E. Schuster, MD, MSCI                    Attending Physician, The Children’s Hospital
        Lorry G. Rubin, MD                                 Associate Professor of Pediatrics, University     at Yale–New Haven, New Haven, Connecticut
        Professor of Pediatrics, Zucker School of          of Missouri-Kansas City School of Medicine;       Fever Without Localizing Signs; Borrelia burg-
        Medicine at Hofstra/Northwell, Hempstead,          Division of Infectious Diseases, Children’s       dorferi (Lyme Disease)
        New York; Director, Infectious Diseases, Cohen     Mercy–Kansas City, Kansas City, Missouri
        Children’s Medical Center of New York, New         Human Metapneumovirus                             Jana Shaw, MD, MPH
        Hyde Park, New York                                                                                  Professor of Pediatrics, SUNY Upstate Medical
        Capnocytophaga Species; Francisella tularensis     Kevin L. Schwartz, MD, MSc, DTM&H                 University, Syracuse, New York
        (Tularemia); Legionella Species; Streptobacillus   Assistant Professor, Dalla Lana School of         Infections of the Oral Cavity
        moniliformis (Rat-Bite Fever); Other Gram-         Public Health, University of Toronto Faculty of
        Negative Coccobacilli                              Medicine; Physician, Infectious Diseases, Unity   Avinash K. Shetty, MD
                                                           Health Toronto; Infectious Diseases Physician,    Associate Dean for Global Health, Professor of
        Edward T. Ryan, MD                                 Infection Prevention and Control, Public          Pediatrics, Director, Global Health Education,
        Professor of Medicine, Harvard Medical             Health Ontario, Toronto, Ontario, Canada          Wake Forest School of Medicine; Chief, Section
        School; Professor of Immunology, Harvard           Protection of Travelers                           of Pediatric Infectious Diseases, Director,
        T.H. Chan School of Public Health; Director,                                                         Pediatric HIV Program, Atrium Health-Wake
        Global Infectious Diseases, Massachusetts          Bethany K. Sederdahl, MD, MPH                     Forest Baptist, Brenner Children’s Hospital,
        General Hospital, Boston, Massachusetts            Attending Physician, Department of Obstetrics     Winston-Salem, North Carolina
        Antiparasitic Agents                               & Gynecology, Christiana Care Healthcare          Rubella Virus; Mumps Virus; Rubeola
                                                           System, Newark, Delaware                          Virus: Measles and Subacute Sclerosing
        Alexandra Sacharok, BS                             Other Campylobacter Species                       Panencephalitis
        Graduate Student Researcher, Department
        of Pediatrics, The Children’s Hospital of          Jose Serpa-Alvarez, MD                            Timothy R. Shope, MD, MPH
        Philadelphia, Philadelphia, Pennsylvania           Associate Professor of Medicine–Infectious        Professor of Pediatrics, University of Pittsburgh
        Classification of Bacteria                         Diseases, Baylor College of Medicine, Houston,    School of Medicine; Attending Physician,
                                                           Texas                                             Children’s Hospital of Pittsburgh of UPMC,
        Thomas J. Sandora, MD, MPH                         Taenia solium, Taenia asiatica, and Taenia        Pittsburgh, Pennsylvania
        Associate Professor of Pediatrics, Harvard         saginata: Taeniasis and Cysticercosis; Taenia     Infections Associated With Group Childcare
        Medical School; Hospital Epidemiologist,           (Multiceps) multiceps and Taenia serialis:
        Department of Pediatrics, Division of              Coenurosis                                        Linda M. Dairiki Shortliffe, MD
        Infectious Diseases, Boston Children’s                                                               Stanley McCormick Memorial Professor
        Hospital, Boston, Massachusetts                    Kara N. Shah, MD, PhD                             Emerita, Department of Urology, Stanford
        Clostridioides difficile                           Optima Dermatology, Liberty Township, Ohio        University School of Medicine, Stanford,
                                                           Urticaria and Erythema Multiforme                 California
        Sarah G.H. Sapp, PhD                                                                                 Urinary Tract Infections, Renal Abscess, and
        Biologist, Division of Parasitic Diseases and                                                        Other Complex Renal Infections
        Malaria, Centers for Disease Control and
        Prevention, Atlanta, Georgia
        Diphyllobothriidae, Dipylidium and
        Hymenolepis Species
xviii
                                                                                                                                  Contributors
                                                                                                                                                         xix
         Contributors
xx
                                                                                                                                    Contributors
                                                                                                                                                           xxi
   PART                 Understanding, Controlling, and Preventing Infectious Diseases
     I
SECTION A: Epidemiology and Control of Infectious Diseases
Epidemiology is the study of the distribution and determinants of disease          Case Definition
or other health-related states or events in specified populations and the
application of this study to the control of health problems.1 A key com-           Establishing a standard case definition is a necessary first step for sur-
ponent of this definition is that epidemiology focuses on populations,             veillance and description of the epidemiology of a disease or health
an emphasis that distinguishes epidemiology from clinical case studies,            event.2 Formulation of a case definition is particularly important when
which focus on individual subjects.                                                laboratory diagnostic testing results are not definitive. More restric-
   Health events can be characterized by their distribution (descriptive           tive case definitions have greater specificity and minimize misclassi-
epidemiology) and by factors that influence their occurrence (analytic             fication of persons without the condition of interest as cases; however,
epidemiology). In both descriptive and analytic epidemiology, health-              they can exclude true cases and may be most useful when investigating
related questions are addressed using quantitative methods to identify             a newly recognized condition, in which the ability to determine etiol-
patterns or associations from which inferences can be drawn and inter-             ogy, pathogenesis, or risk factors is decreased by inclusion of noncases
ventions developed, applied, and assessed.                                         in the study population. A more inclusive definition can be important
                                                                                   in an outbreak setting to increase sensitivity to detect potential cases
                                                                                   for further investigation or to inform application of preventive inter-
DESCRIPTIVE EPIDEMIOLOGY                                                           ventions (e.g., reactive vaccination campaigns). Multiple research or
                                                                                   public health objectives can be addressed by developing a tiered case
Surveillance                                                                       definition that incorporates varying degrees of diagnostic certainty for
The goals of descriptive epidemiology are to define the frequency of               confirmed, probable, and suspected cases. The Council of State and Ter-
health-related events and determine their distribution by person, place,           ritorial Epidemiologists (CSTE) provides uniform surveillance case defi-
and time. The foundation of descriptive epidemiology is surveillance, or           nitions for nationally notifiable infectious and noninfectious conditions
case detection. Retrospective surveillance identifies health events from           (https://wwwn.cdc.gov/nndss/case-definitions.html).
existing data, such as clinical or laboratory records, hospital discharge
data, and death certificates. Prospective surveillance identifies and col-         Sensitivity, Specificity, and Predictive Value
lects information about cases as they occur, for example, through ongoing
laboratory-based reporting.                                                        Sensitivity, specificity, and predictive values can be used to quantify the
    With passive surveillance, case reports are supplied voluntarily by clini-     performance of a case definition or the results of a diagnostic test or algo-
cians, laboratories, health departments, or other sources. The complete-           rithm (Table 1.1). Sensitivity and specificity are intrinsic measures of a
ness and accuracy of passive reporting are affected by whether reporting           case definition or diagnostic test, whereas predictive values vary with the
is legally mandated, the ease of establishing a definitive diagnosis for the       prevalence of a condition within a population. Even with a highly specific
disease under surveillance, illness severity, interest in and awareness of         diagnostic test, if a disease is uncommon among the people tested, a large
the medical condition among the public and the medical community, and              proportion of positive test results will be false positives, and the positive
by whether a report will elicit a public health response. Because more             predictive value will be low (Table 1.2). If the test is applied more selec-
severe illness is more likely to be diagnosed and reported, the severity           tively, such that the proportion of people tested who truly have disease
and clinical spectrum of passively reported cases often differ from those          is greater, the test’s predictive value will be improved. Thus, predictive
of all cases of an illness. Weekly and annual state counts of passively            values depend both on test sensitivity and specificity, and on the disease
collected reports of nationally notifiable diseases are available on the           prevalence in the population in which the test is applied, also called the
National Notifiable Diseases System (NNDSS) Data and Statistics web                pre-test probability.
page (https://wwwn.cdc.gov/nndss/data-and-statistics.html).                           Often, the sensitivity and specificity of a test are inversely related.
    In active surveillance, an effort is made to ascertain all cases of a condi-   Selecting the optimal balance of sensitivity and specificity depends on the
tion occurring in a defined population. Active case finding can be pro-            purpose for which the test is used. Generally, a screening test should be
spective (through routine contacts with reporting sources), retrospective          highly sensitive, whereas a follow-up confirmatory test should be highly
(through record audit), or both. Population-based active surveillance, in          specific.
which all cases in a defined geographic area are identified and reported,
provides the most complete and unbiased ascertainment of disease and is            Incidence and Prevalence
optimal for describing the rate of a disease and its clinical spectrum. By
contrast, active surveillance conducted at only one or several participat-         Characterizing disease frequency is one of the most important aspects of
ing facilities, often referred to as sentinel surveillance, can yield biased       descriptive epidemiology. Frequency measures typically include a count
information on disease frequency or spectrum based on the representa-              of new or existing cases of disease as the numerator and a quantifica-
tiveness of the patient population and the size of the sample obtained. The        tion of the population at risk as the denominator. Cumulative incidence
range and severity of clinical symptoms also influence whether active sur-         is expressed as a proportion and describes the number of new cases of
veillance is able to ascertain all cases of a disease; individuals with mild       an illness occurring in a fixed at-risk population over a specified period
disease may not present to a physician for diagnosis.                              of time. The incidence density or incidence rate is the rate of new cases
                                                                                                                                                                   1
     PART I Understanding, Controlling, and Preventing Infectious Diseases
     SECTION A Epidemiology and Control of Infectious Diseases
     TABLE 1.1   Definitions and Formulas for the Calculation of Important Epidemiologic Parameters
     Measures of test        Sensitivity: Proportion of true positive (diseased) with a        A/(A + C)
       accuracy                positive test result
                             Specificity: Proportion of true negative (nondiseased) with a     D/(B + D)
                               negative test result
                             Positive predictive value (PPV): Proportion of positive test      A/(A + B)
                               results that are true positives
                             Negative predictive value (NPV): Proportion of negative test      D/(C + D)
                               results that are true negatives
     Measures of data        Variance: Statistic describing variability among individual       [1 (n − 1)][ x1 − x )2 +
       dispersion and           members of a population                                            ( x 2 − x )2 + … + ( x n − x )2 ]
       precision
                             Standard deviation (SD): A second, more commonly used                      Variance
                                statistic describing variability among individual members
                                of a population
                             Standard error (SE): Statistic describing the variability           SD
                                of sample-based point estimates (P) around the true
                                                                                                  √n
                                population value being estimated
                             Confidence interval: A range of values that is believed to        —
                               contain the true value within a defined level of certainty
                               (usually 95%)
                                                                                                               10       Secondary case
     Proportion With         Positive Predictive           Negative Predictive
     Condition               Value                         Value
             1%                      8%                          >99%
                                                                                                               5
            10%                    50%                             99%
            20%                    69%                             97%
            50%                    90%                             90%
                                                                                                                    1   7   13   19    25   31      6                  12   18   24   30   6    12   18   24
                                                                                                                            October                   November                                 December
    of disease in a dynamic at-risk population; the denominator typically is                                                                         Date of onset
    expressed as the population-time at-risk (e.g., person-time).                            FIGURE 1.1. Example of an epidemic curve for a common source outbreak
       Because the occurrence of many infections varies with season, extrap-                 with continuous exposure. Cases of hepatitis A by date of onset in Fayetteville,
    olating annual incidence from cases detected during a short observation                  Arkansas, from November to December 1979. (From Centers for Disease Control
    period can be inaccurate. In describing the risk of acquiring illness dur-               and Prevention, unpublished data.)
    ing a disease outbreak, the attack rate, defined as the number of new cases
    of disease occurring in a specified population and time period, is a useful
    measure. Finally, the case-fatality rate, or proportion of cases of a disease            the impact of prevention programs. The timing of illness in outbreaks can
    that result in death, is used to quantify the mortality resulting from a                 be displayed in an epidemic curve (Fig. 1.1) and can be useful in defin-
    disease in a particular population and time period.                                      ing the mode of transmission or incubation period, or for assessing the
       Prevalence refers to the proportion of the population having a con-                   effectiveness of control measures.
    dition at a specific point in time. As such, it is a better measure of
    disease burden for chronic conditions than is incidence or attack rate,                  ANALYTIC EPIDEMIOLOGY
    which identify only new (incident) cases. Prevalent cases of disease
    can be ascertained in a cross-sectional survey, whereas determining                      Study Design
    incidence requires longitudinal surveillance. When disease prevalence
    (P) is low and incidence (I) and duration (D) are stable, prevalence                     The goal of analytic epidemiologic studies is to assess for and quantify
    is a function of disease incidence multiplied by its average duration                    the association between an exposure and a health outcome. This goal can
    (P = I × D).                                                                             be addressed in experimental or observational studies. In experimental
                                                                                             studies, hypotheses are tested by systematically allocating an exposure of
                                                                                             interest to subjects in separate groups to achieve the desired comparison.
    Describing Illness by Person, Place, and Time                                            Such studies include randomized, controlled, double-blind treatment
    Characterizing disease by person, place, and time is often useful. Demo-                 trials as well as laboratory experiments. By carefully controlling study
    graphic variables, including age, sex, socioeconomic status, and race or                 variables, investigators can restrict differences among groups and thereby
    ethnicity, often are associated with the risk of disease. Describing a dis-              increase the likelihood that the observed differences are a consequence
    ease by place can help define risk groups, for example, when an illness is               of the specific factor being studied. Because experiments are prospective,
    caused by an environmental exposure or is vector borne, or during an                     the temporal sequence of exposure and outcome can be established, mak-
    outbreak with a point source exposure. Time also is a useful descriptor of               ing it possible to define cause and effect.
    disease occurrence. Evaluating long-term (secular) trends provides infor-                   By contrast, observational studies test hypotheses using observational
    mation that can be used to identify emerging health problems or to assess                methods to assess exposures and outcomes among individual subjects in
2
                                                                                                                   Principles of Epidemiology and Public Health           1
 TABLE 1.3   Types of Observational Studies and Their Advantages and Disadvantages
 Type of Study               Design and Characteristics                      Advantages                                     Disadvantages
 Cohort                      Prospective or retrospective                    Ideal for outbreak investigations in defined   Unsuited for rare diseases or those with
                                                                                populations                                   long latency
                             Select study group                              Prospective design ensures that exposure       Expensive
                                                                                preceded disease
                             Observe for exposures and disease               Selection of study group is unbiased by        Can require long follow-up periods
                                                                                knowledge of disease status
                             Outcome measures used: Relative risk (RR) or    RR and HR accurately describe risk given       Difficult to investigate multiple exposures
                               hazard ratio (HR) of disease given exposure     an exposure
 Cross-sectional             Nondirectional                                  Rapid, easy to perform, and inexpensive        Timing of exposure and disease can be
                                                                                                                               difficult to determine
                             Select study group                              Ideal to determine knowledge, attitudes,       Biases can affect recall of past exposures
                                                                                and behaviors
                             Determine exposure and disease status
                             Outcome measures used: Prevalence ratio for
                               disease given exposure
 Case-control                Retrospective                                   Rapid, easy to perform, and inexpensive        Timing of exposure and disease can be
                                                                                                                               difficult to determine
                             Identify cases with disease                     Ideal for studying rare diseases, those with   Biases can occur in selecting cases and
                                                                                long latency, new diseases                     controls and determining exposures
                             Identify controls without disease                                                              OR only provides an estimate of the RR if
                                                                                                                              disease is rare
                             Determine exposures in cases and controls
                             Outcome measures used: Odds ratio (OR) for
                               an exposure given disease
populations and to identify statistical associations from which inferences           an exposure. However, in a cohort study it can be difficult to investigate
regarding causation are drawn. Although observational studies can-                   multiple exposures as risk factors for a single outcome. Cohort studies
not be controlled to the same degree as experiments, they are practical              also are impractical for studying rare diseases or conditions with a long
in circumstances in which exposures or behaviors cannot be assigned.                 latent period between exposure and the onset of clinical illness. In gen-
Moreover, the results often are more generalizable to a real population              eral, cohort studies are unsuited for investigating risk factors for new or
having a wide range of attributes. The 3 basic types of observational stud-          rare diseases or for generating new hypotheses about possible exposure-
ies are cohort studies, cross-sectional studies, and case-control studies            disease relationships.
(Table 1.3). Hybrid study designs, incorporating components of these 3                  Cohort studies provide data not only on whether an outcome occurs
types, also have been developed.3 In planning observational studies, care            but also, for those experiencing the outcome, on when it occurs. Analysis
must be taken in the selection of participants to minimize the possibility           of time-to-event data for outcomes such as death or illness is a power-
of bias. Selection bias results when study subjects have differing prob-             ful approach to assess or compare the impacts of preventive or therapeu-
abilities of being selected and the probability of selection is related to the       tic interventions. The probability of remaining event-free over time can
risk factors or outcomes under evaluation.                                           be expressed in a survival curve where the event-free probability is ini-
   In contrast to experimental or observational studies that analyze infor-          tially 1 and declines in a step-function as the outcomes of interest occur
mation about individual subjects, ecologic studies draw inferences from              (Fig. 1.2A). Time-to-event data also can be displayed as the cumulative
data on a population level. Causal inferences from ecologic studies must             hazard of an event occurring among members of a cohort that increases
be made with caution because relationships observed on a population                  from 0 at enrollment (Fig. 1.2B). These 2 approaches are related in that
level do not necessarily apply on the individual level (a problem known              the hazard reflects the incident event rate, whereas survival reflects the
as the ecologic fallacy). Because of these drawbacks, ecologic studies are           cumulative nonoccurrence of that outcome.4,5 With time-to-event analy-
suited best for generating hypotheses that can be tested using other study           sis, the association between exposure and disease often is expressed as a
methods.                                                                             hazard ratio. Like relative risk, the hazard ratio is a comparative measure
                                                                                     of risk between exposed and unexposed groups. The primary difference
Cohort Studies                                                                       is that the hazard ratio compares event experience over the entire time
                                                                                     period, whereas the relative risk compares cumulative event occurrence
In a cohort study, subjects are categorized based on their exposure to a             at the study endpoint.6
suspected risk factor and are observed for the development of disease or
other health-related outcome. Associations between exposure and dis-                 Cross-Sectional Studies
ease are expressed by the relative risk of disease, or risk ratio, in exposed
and unexposed groups (Table 1.4). Cohort studies typically are pro-                  In a cross-sectional study, or survey, a sample is selected, and at a single
spective, with exposure defined before disease occurs. However, cohort               point in time exposures and outcome are determined. Outcomes can
studies also can be retrospective, in which the cohort is selected after             include disease status or behaviors and beliefs, and multiple exposures
the outcome has occurred. In this case, exposures are determined from                can be evaluated as explanations for the outcome. Associations are char-
existing records that preceded the outcome, and thus the directional-                acterized by the prevalence ratio, similar to the risk ratio in cohort stud-
ity of the exposure-disease relationship is still forward. Characterizing            ies. Because neither exposures nor outcomes are used in selection of the
exposures before development of disease is a major benefit of cohort                 study group, prevalence is an estimate of that in the overall population
studies because this approach minimizes selection bias and simplifies                from which the sample was drawn. National survey data characterizing
inference of cause and effect. Another advantage of cohort studies is that           health status, behaviors, and medical care are available from the National
they can be used to assess multiple potential outcomes resulting from                Center for Health Statistics (http://www.cdc.gov/nchs/index.htm).
                                                                                                                                                                          3
     PART I Understanding, Controlling, and Preventing Infectious Diseases
     SECTION A Epidemiology and Control of Infectious Diseases
    Case-Control Studies                                                                     hospital-based studies, differential referral also can bias selection of a
                                                                                             study sample. This bias would occur if, for example, the frequency of an
    In a case-control study, the investigator identifies a group of people with a            exposure varied with socioeconomic status and a hospital predominantly
    disease or outcome of interest (cases) and compares their exposures with                 admitted persons from either a high-income group or a low-income
    those in a selected group of people who do not have disease (controls).                  group. Bias also can occur when eligible subjects refuse to participate in
    Differences between the groups are expressed by an odds ratio, which                     a study.
    compares the odds of an exposure in case and control groups (Table 1.4).                     Determination of exposures can also be affected by several types of
    The odds ratio is not the same as a risk ratio; however, it provides an                  bias. Recall of exposures can be different for persons who have had an
    estimate of the risk ratio if the disease or outcome in question is rare.                illness compared with people who were well. This bias occurs in either
    Case-control studies are retrospective in that disease status is known and               direction: patients may be more likely to remember an exposure that
    serves as the basis for selecting the 2 comparison groups; exposures are                 they associate with their illness (e.g., what was eaten before an episode
    then determined by reviewing available records or by interview.                          of diarrhea) or less likely to recall an exposure if a severe illness affected
       A major advantage of case-control studies is their efficiency in study-               memory. Interviewers can introduce bias by questioning cases and con-
    ing uncommon diseases or diseases with a long latency. Case-control                      trols differently about their exposures. Misclassification of exposures can
    studies also can evaluate multiple exposures that may contribute to a                    also result from errors in measurement such as can occur with the use of
    single outcome; study subjects frequently can be identified from existing                an inaccurate laboratory test. Although systematic misclassification can
    sources (e.g., hospital or laboratory records, disease registries, or surveil-           result in bias, misclassification of exposure often is random rather than
    lance reports), and after identification of suitable control subjects, data              systematic.
    on previous exposures can be collected rapidly. Case-control studies also                    Even a carefully designed study that minimizes potential biases can
    have several drawbacks. Bias can be introduced during selection of cases                 lead to erroneous causal inferences. An exposure can falsely appear to be
    and controls and in determining exposures retrospectively, and inferring                 associated with disease because it is closely linked to the true, but unde-
    causation from statistically significant associations can in some situations             termined, risk factor. For example, race often is found to be associated
    be complicated by difficulty in determining the temporal sequence of                     with the risk of a disease, but in many instances the true risk factor is
    exposure and disease.                                                                    likely an unmeasured variable that is associated with race, such as socio-
                                                                                             economic status. The risk of making incorrect inferences can be mini-
    Causal Inference and the Impact of Bias                                                  mized by considering certain general criteria for establishing causation.
                                                                                             These criteria include the strength of an association, the presence of a
    The impact of potential bias is particularly important in observational                  dose-response effect, a clear temporal sequence of exposure to disease,
    studies. The validity of a study is the degree to which inferences drawn                 the consistency of findings with those of other studies, and the biologic
    from a study are warranted. Internal validity refers to the correctness of               plausibility of the hypothesis.8
    study conclusions for the population from which the study sample was
    drawn, whereas external validity refers to the extent to which the study
    results can be generalized beyond the population sampled. The validity of
                                                                                             Statistical Analysis
    a study can be affected by bias, or systematic error, in selecting the study
    participants (sampling), in ascertaining their exposures, or in analyzing
                                                                                             Characteristics of Populations and Samples
    and interpreting study data. For errors to result in bias, they must be sys-             While epidemiologic analysis seeks to draw valid conclusions about popu-
    tematic, or directional. Nonsystematic error (random misclassification)                  lations, the entire population rarely is included in a study. An assumption
    decreases the ability of a study to identify a true association but does not             underlying statistical analysis is that the sample evaluated was selected
    usually result in detection of a spurious association.                                   randomly from the population. Often, this criterion is not met and calls
       Several sources of bias can occur in selection of study participants                  into question the appropriateness and interpretation of statistical analyses.
    (Box 1.1).7 Diagnosis bias results when persons with a given exposure are                   The mean, median, and mode describe central values for samples and
    more likely to be diagnosed as having disease than are people without                    populations. The arithmetic mean is the average, determined by sum-
    the exposure (or vice versa); this can occur because diagnostic testing                  ming individual values and dividing by the sample size. When data are
    is more or less likely to be done based on exposure or because the inter-                not normally distributed, or skewed, calculation of a geometric mean can
    pretation of a test may be affected by knowledge of exposure status. For                 limit the impact of outlying values. The geometric mean is calculated by
4
                                                                                                                         Principles of Epidemiology and Public Health   1
                          1.0
                                                                                       BOX 1.1 Potential Sources of Bias in Observational Studiesa
                          0.3
Cumulative hazard
                          0.2
                                                                                     (Table 1.1). For a normally distributed population, 68% of values fall
                                                                                     within 1 standard deviation of the mean, and 95% of values fall within
                          0.1                                                        1.96 standard deviations.
                                                                                        When analyzing a sample, the mean or other statistics describing the
                                                                                     sample represent a point estimate of that parameter for the entire popula-
                          0.0                                                        tion. If another random sample were drawn from the same population,
                                                                                     the point estimate for the parameter of interest likely would be differ-
                                0   10   20        30         40        50     60
                                                                                     ent, depending on the variability in the population and the sample size
                                              Age (months)                           selected. The standard error is used to describe the precision of a point
                                                                                     estimate (e.g., mean, odds ratio, relative risk) and depends on sample
                                                   HBV                               standard deviation and the sample size (Table 1.1).
                 %                                                                      A confidence interval defines a range of values that includes the true
                                                   PncCRM
                                                                                     population value within a defined level of certainty. Most often, the 95%
FIGURE 1.2. Example of Kaplan-Meier and cumulative hazard curves. (A)                confidence interval is presented (Table 1.1).
Survival plot for critically ill patients with Streptococcus pneumoniae bacteremia
treated with monotherapy or combination therapy. (B) Cumulative hazard of
tympanostomy tube placement from 2 months until 4–5 years of age in children         Absolute and Relative Measures of Association
who received pneumococcal conjugate vaccine (PncCRM) or a control vaccine            Measures of association are used to assess the strength of an association
(hepatitis B vaccine [HBV]). ([A] Redrawn from Baddour LM, Yu VL, Klugman KP,
                                                                                     between an exposure and an outcome. In a cohort study, the absolute risk
et al. Combination antibiotic therapy lowers mortality among severely ill patients
                                                                                     reduction (also known as excess risk, attributable risk, or risk difference)
with pneumococcal bacteremia. Am J Respir Crit Care Med. 2004;170:440–444.
                                                                                     is the difference in the incidence of the outcome between exposed and
[B] Redrawn from Palmu AAI, Verho J, Jokinen J, et al. The seven-valent
pneumococcal conjugate vaccine reduced tympanostomy tube placement in
                                                                                     unexposed subjects. The number needed to treat (NNT) is a measure of the
children. Pediatr Infect Dis J. 2004;23:732–738.)                                    number of individual subjects who must receive a treatment to prevent a
                                                                                     single negative outcome and is calculated as the reciprocal of the absolute
                                                                                     risk reduction. In addition to absolute measures, relative measures also
                                                                                     are useful for describing the strength of an association. In a cohort study
taking the nth root of the product of all the individual values, where n is          or survey, the relative risk or risk ratio compares the risk of disease for
the total number of individual values. For example, immunogenicity of                subjects with versus subjects without an exposure (Table 1.4). In case-
vaccines is usually expressed by the geometric mean titer. The median, or            control studies, association is assessed by the odds ratio, which compares
middle value, is another way to describe nonnormally distributed data.               the odds of exposure among subjects with and without a disease or health
The mode, or most commonly occurring value in a sample, rarely is used.              outcome; when disease is uncommon (<10%) in both exposed and unex-
   Several measures can be used to describe the variability in a sample.             posed groups, the odds ratio approximates the relative risk. For time-
The range describes the difference between the highest and lowest value,             to-event analyses, the comparative risk is expressed as the hazard ratio.
whereas the interquartile range defines the difference between the 25th              Odds ratios, relative risks, and hazard ratios >1 signify increased risk
and 75th percentiles. Variation among individual elements most often                 given exposure, and values <1 suggest that exposure decreases the risk
is characterized by the variance or standard deviation. The variance is              of an outcome. Because observational studies generally do not include
the mean of the squared deviation of each observation from the sam-                  all members of a population, these measures of association represent an
ple’s mean. The standard deviation is the square root of the variance                estimate of the true value within the entire population. Statistical analyses
                                                                                                                                                                        5
     PART I Understanding, Controlling, and Preventing Infectious Diseases
     SECTION A Epidemiology and Control of Infectious Diseases
    can help guide investigators in making causal inferences based on point            significant difference among treatment groups. However, only 3 of these
    estimates of these measures of association.                                        trials were adequately powered to exclude a 10% difference in mortality,
                                                                                       thus showing that many studies potentially missed a clinically significant
    Statistical Significance                                                           difference.11
                                                                                           In some situations, an investigator would want to detect a significant
    Statistical tests are applied to assess the likelihood that the study results      difference among study groups as soon as possible, for example, when a
    were obtained by chance alone rather than representing a true differ-              therapeutic or preventive intervention could be applied once a risk group
    ence within the population. Most investigators consider a P value <0.05            is identified or when concerns exist about the safety of a drug or vaccine.
    as being statistically significant, indicating a <5% risk that the observed        One approach to this situation is to include in the study design an interim
    association is the result of chance alone (designated a type I error, the          analysis after a specified number of subjects are evaluated. Because the
    probability of which is the alpha level). Although use of this cutoff for          likelihood of identifying chance differences as significant increases with
    significance testing has become conventional, ignoring higher P values             the number of analyses, it is recommended that the threshold for defining
    can lead to missing a real and important association, whereas blind faith          statistical significance should become more stringent as the number of
    in the significance of lower P values can lead to erroneous conclusions.           planned analyses increases.12 If each interim analysis can lead the inves-
    Statistical testing should contribute to, but not replace, criteria for evalu-     tigators to stop the trial, this study design is considered a group sequen-
    ating possible causation.                                                          tial method.13 Another example of a group sequential design is when
       Statistical significance also can be defined based on 95% confidence            concordance or discordance in outcome is tabulated for each matched
    intervals, which approximately correspond to a P value of 0.05. An odds            set exposed to alternate treatments. Results for each set are plotted on a
    ratio, relative risk, or hazard ratio is considered statistically significant if   graph, and data collection continues until a preset threshold for a signifi-
    the 95% confidence interval does not include 1. An advantage of using              cant difference among study groups is crossed or no significant difference
    confidence intervals to define statistical significance is that they provide       is detected at a given power.12
    information on whether a finding is statistically significant and on the
    possible range of values for the point estimate in the population, with            Statistical Inference
    95% certainty.
       One pitfall in interpreting statistical significance is ignoring the mag-       Statistical testing is used to determine the significance of differences
    nitude of an effect in favor of its “significance.” A very large, overpowered      among study groups, and thus it provides guidance on whether to accept
    study can identify as significant a small, perhaps trivial, difference among       or reject the null hypothesis. Although providing details of specific statis-
    study groups. Some epidemiologists have proposed that, despite statisti-           tical tests is outside the scope of this chapter, Table 1.5 gives examples of
    cal significance, odds ratios <2 or 3 in an observational study should not         statistical tests that can be applied in analyzing different types of exposure
    be interpreted because unidentified bias or confounding could account              and outcome variables.
    for a difference of this magnitude.9 Conversely, the relative risk or odds            Using appropriate analytic and statistical methods is important in
    ratio associating an exposure and outcome can be large, but if the expo-           identifying significant predictors of an outcome (i.e., risk factors) cor-
    sure is uncommon in both groups, it cannot explain most cases of ill-              rectly. Confounding variables are associated with the disease of inter-
    ness. The public health importance of an exposure can be described by              est and with other exposure variables and are not part of the causal
    the population-attributable fraction, or the proportion of the disease in a        pathway between the other exposure(s) and the outcome. For example,
    population that is related to the exposure of interest.                            consider a study attempting to determine whether meningococcal vac-
                                                                                       cination decreases nasopharyngeal carriage of Neisseria meningitidis. If
    Sample Size                                                                        frequent attendance at social events is associated with increased car-
                                                                                       riage and also with a decreased chance of receiving vaccine, then failure
    Another type of error in epidemiologic studies is when a study fails to            to adjust for social event attendance as a confounding variable could
    identify a true risk factor as statistically significant (designated a type II     lead to overestimation of the relationship between vaccination and car-
    error, the probability of which is the beta level). The probability of a type      riage reduction. Meanwhile, effect modifiers interact with other risk fac-
    II error is higher when the sample size is small. Often, the type II error         tors to affect their impact on outcome but may or may not be associated
    rate is set at 0.2, indicating acceptance of a 20% likelihood that a true dif-     with the outcome on their own. Frequently, age is an effect modifier,
    ference exists but would not be identified by the study. Statistical power is      with an exposure associated significantly with an outcome in one age
    defined as 1 − β and is the complement of the probability of committing a          group but not in another.
    type II error (β); that is, power is the probability of correctly identifying a       Several approaches are used to control for confounding variables and
    difference of specified size among groups, if such a difference truly exists.      effect modifiers. In study design, an extraneous variable can be controlled
    The problem of inadequate sample size in clinical studies was highlighted
    in an analysis of “negative” randomized controlled trials reported in 3
    leading medical journals between 1975 and 1990. Of 70 reports, only 16%
    and 36% had sufficient statistical power (80%) to detect a true 25% or              TABLE 1.5   Types of Statistical Tests Used to Evaluate the Significance
    50% relative difference, respectively, among treatment groups.10                    of Associations Among Categorical and Continuous Variables
       In calculating sample sizes for testing hypotheses, investigators must
    select acceptable rates of type I and type II errors and define the magni-                                      Dependent Variable (Disease, Outcome)
                                                                                        Independent
    tude of the difference in outcomes that is deemed clinically important.             Variable (Exposure,     Categorical and
    Sample size calculations can be performed using a range of computer soft-            Risk Factor)           Dichotomous            Continuous
    ware. The program Epi-Info can be used to perform sample size calcula-              CATEGORICAL
    tions as well as other statistical functions and is available at no charge from
    the Centers for Disease Control and Prevention (www.cdc.gov/epiinfo/).              Dichotomous              Chi-square test        Student t-test (parametric)
       Ensuring an adequate sample size is particularly important for stud-                                      Fisher exact test      Wilcoxon rank sum test
    ies attempting to prove equivalence or noninferiority of a new treatment                                                               (nonparametric)
    compared with standard therapy. Food and Drug Administration guid-
                                                                                        >2 categories            Chi-square test        Analysis of variance
    ance recommends that noninferiority trials adopt a null hypothesis that a                                                             (parametric)
    difference exists among treatments; this hypothesis is rejected if the lower
    95% confidence limit for the new treatment is within a specified margin                                                             Kruskal-Wallis test
    of the point estimate for standard therapy. Because the null hypotheses                                                                (nonparametric)
    can never be proven or accepted, the failure to reject a null hypothesis of         CONTINUOUS               Logistic regression    Linear regression
    no difference among treatments or exposure does not prove equivalence.
                                                                                                                                        Correlation (Pearson:
    The importance of this distinction is illustrated by an analysis of 25 stud-                                                          parametric; Spearman:
    ies claiming equivalence of therapies for pediatric bacterial meningitis.                                                             nonparametric)
    Twenty-three studies claimed equivalence based on a failure to detect a
6
                                                                                                             Principles of Epidemiology and Public Health         1
for by randomization, restricting sampling to one category of the variable         confidence limit that is >0% indicates statistically significant protection.
or by frequency matching to obtain similar proportions of cases and con-           However, the expected lower confidence limit often is much >0 to be con-
trols in each stratum of the variable. A more extreme form of matching is          sistent with meaningful levels of protection. The most important com-
to select control subjects who are similar to individual cases for extrane-        ponent of a case-control effectiveness study is selecting control subjects
ous variables (e.g., age, sex, underlying disease) and to analyze whether          who have the same opportunity for immunization as do cases. If cases
exposures are concordant or discordant within matched sets. A newer                have less opportunity to be immunized, results will be biased toward
approach to study design is the case-crossover14 or case series15 analysis.        showing protection. Factors such as low socioeconomic status, which can
In this method, exposures occurring in a defined risk period before the            increase the risk of disease and decrease the chance of being immunized,
outcome are compared with exposures occurring outside the risk window              are potential confounding variables and can be controlled for by match-
for the same individual subjects. This approach has been adapted to the            ing control subjects to cases for those factors.
study of adverse events after vaccination. If the vaccine causes the event,           Cohort studies also can be used to determine vaccine effectiveness after
the rate of the event will be greater within a defined risk window than pre-       licensure. A study design called the indirect-cohort method was devel-
dicted by chance alone based on the expected distribution of the event.16          oped by researchers at the Centers for Disease Control and Prevention
The strength of this approach is that each subject, or case, serves as his or      to evaluate the effectiveness of the pneumococcal polysaccharide vaccine
her own control, thereby decreasing confounding.                                   by using data collected by disease surveillance.18 In this study, the cohort
   At the analysis stage, the impact of confounding variables and effect           included persons identified with invasive pneumococcal infections. The
modifiers can be limited by performing a stratified analysis or using a            study hypothesis was that if pneumococcal vaccines were protective, the
multivariable model. In a stratified analysis, the possible association            proportion of vaccinated persons infected with pneumococcal serotypes
between a risk factor and an outcome is determined separately within               that are included in the vaccine formulation would be less than the pro-
different categories, or strata, of the extraneous variable. If the extraneous     portion of unvaccinated persons infected with vaccine-type strains. Vac-
variable is a confounder, the stratum-specific estimates should be similar         cine effectiveness was calculated from the relative serotype distributions
to each other and can be combined into a single estimate using an appro-           overall and for each individual serotype. The point estimate of vaccine
priate statistical test (e.g., a Mantel-Haenszel odds ratio). If a stratifica-     effectiveness for preventing invasive infection was 57% (95% confidence
tion variable is an effect modifier, the relative risk or odds ratio will differ   interval, 45%–66%)19; this estimate is similar to that obtained in a case-
substantially among the strata; for example, an exposure can be a strong           control effectiveness study.20
risk factor in one age group but not another. In this setting, a summary
statistic should not be presented, and results for each stratum should be          DISEASE CONTROL AND PUBLIC HEALTH POLICY
presented separately. When the extraneous variable is confounding, strat-
ifying the analysis by the confounding variable may eliminate an appar-            Outbreak Investigations
ent association between the exposure and the outcome in the unstratified
analysis and indicate that the exposure is not an independent risk factor          Outbreak investigations require knowledge of disease transmission and
for disease.                                                                       use of descriptive and analytic epidemiologic tools. Possible outbreaks
   Because stratified analyses become confusing rapidly as the number of           can be identified from surveillance data showing an increased rate of an
strata increases, techniques of statistical modeling have been developed           infection or an unusual clustering of infection by person, place, or time.
that permit simultaneous control of multiple variables. Significant risk           Comparing the incidence rate of disease with a baseline rate from a previ-
factors determined in a multivariable model are interpreted as each con-           ous period is helpful in validating the occurrence of an outbreak. Other
tributing independently and significantly to the outcome, as the model             explanations for changes in the apparent rate of disease occurrence, such
controls for potential confounding from each included variable. Effect             as diagnostic error, seasonal variations, and changes in reporting, must
modification can be taken into account by including terms expressing the           be considered.
interaction between a risk factor and effect modifier in the model. Vari-             After identifying a potential outbreak, the next steps of an investiga-
ous multivariable models are appropriate for discrete, continuous, and             tion are to develop a case definition, identify cases, and characterize the
time-dependent outcomes.                                                           descriptive epidemiology of the outbreak. An epidemic curve depicts
   A limitation of multivariable modeling is multicollinearity, which              the number of cases over time and can provide information on possible
occurs when 2 or more explanatory variables of interest are highly corre-          transmission (Fig. 1.1). In an outbreak with a point source exposure, an
lated and can result in inaccurate measures of association and decreased           index case may be identified, with other cases occurring after an incuba-
statistical power. The risk of multicollinearity can be reduced by assessing       tion period or at multiples of an incubation period. Plotting the location
correlations among potential risk factors and selecting which variables to         of cases on a spot map can also be helpful in determining possible expo-
include in the model. Various methods to identify and minimize multi-              sures. Describing patients’ characteristics can be important in identify-
collinearity have been developed.17                                                ing at-risk populations for further investigation or targeting of control
                                                                                   measures, as well as for developing hypotheses that can be investigated
VACCINE EFFICACY AND EFFECTIVENESS STUDIES                                         in an analytic study.
                                                                                      Not all outbreaks can be traced to a point source exposure. Outbreaks
Most prelicensure efficacy studies are experimental, randomized, dou-              and epidemics also can result from increased transmission of an endemic
ble-blind, controlled trials in which vaccine efficacy (VE) is calculated          disease (i.e., a disease or condition that normally occurs in a specific
by comparing the attack rates (AR) for disease in the vaccinated and               population or area). In this situation, it can be challenging to determine
unvaccinated groups: VE (%) = (AR unvaccinated − AR vaccinated) / AR               when an increase in disease constitutes an outbreak rather than a nor-
unvaccinated) × 100; or (1 − RR) × 100.                                            mal fluctuation in disease incidence. To determine whether an outbreak
   After licensure, conducting controlled studies, which requires with-            is occurring, the current incidence of the disease must be compared with
holding vaccine from a control group, is no longer ethical. Therefore,             the baseline disease incidence in that area. Often, no standard definition
further studies must be observational rather than experimental, by                 exists for when an increase in endemic disease incidence constitutes an
comparing persons who have chosen to be immunized with those who                   outbreak or epidemic. For instance, in US pertussis epidemics the thresh-
have not. Such observational studies are said to assess vaccine effective-         old for declaring an epidemic has varied by state. In California in 2010
ness rather than vaccine efficacy; however, the two measures use the same          and 2014, an epidemic was declared when the statewide case counts had
abbreviation, VE. In case-control vaccine effectiveness studies, vaccina-          reached 5 times the number of cases observed in a year with baseline per-
tion status of persons with disease is compared with vaccination status            tussis incidence. By contrast, in Washington State in 2012, an epidemic
of healthy control subjects in a real-world setting. The number of vac-            was declared when the incidence of pertussis reached 2 standard devia-
cinated and unvaccinated cases and controls is included in a 2×2 table,            tions above the statewide 10-year average.
and vaccine effectiveness is calculated as 1 minus the odds ratio: VE (%)             Cohort studies are optimal for investigating outbreaks that occur in
= (1 − OR) × 100. When the proportion of cases vaccinated is less than             small, well-defined populations, including in schools, childcare settings,
the proportion of vaccinated controls, the odds ratio is <1, and the point         social gatherings, and hospitals. In populations that are large and/or
estimate for VE indicates that immunization is protective. The precision           not well defined, a case-control study is the most feasible approach. It is
of the estimate is expressed by the 95% confidence interval. A lower 95%           important to select control subjects who had an opportunity equal to that
                                                                                                                                                                  7
Another random document with
 no related content on Scribd:
   “He’s my brother,” I said.
   “I don’t care. He wanted to drown me; he didn’t know I can’t die by
water.”
   “Can’t you?” I said.
   “Of course not. I’m a changeling!”
   She said it with a childish seriousness that confounded me.
   “What made you one?” I asked.
   “The fairies,” she said, “and that’s why I’m here.”
   I was too bewildered to pursue the subject further.
   “How did you fall in there?” I asked.
   “I saw some little fish, like klinkents of rainbow, and wanted to
catch them; then I slipped and soused.”
   “Well,” I said, “where are you going now?”
   “With you,” she answered.
   I offered no resistance. I gave no thought to results, or to what my
father would say when this grotesque young figure should break into
his presence. Mechanically I started for home and she walked by my
side, chatting. Jason strode in our rear, whistling.
   “What a brute he must be!” she said once, jerking her head
backward.
   “Leave him alone,” I said, “or we shall quarrel. What’s a girl like
you to him?”
   I think she hardly heard me, for the whistle had dropped to a very
mellow note. To my surprise I noticed that she was crying.
   “I thought changelings couldn’t cry?” I said.
   “I tell you water does not affect me,” she answered, sharply. “What
a mean spy you are—for a boy.”
   I was very angry at that and strode on with black looks, whereupon
she edged up to me and said, softly: “Don’t be sore with me, don’t.”
   I shrugged my shoulders.
   “Let’s kiss and be friends,” she whispered.
   For the first time in my life I blushed furiously.
   “You beast,” I said, “to think that men would kiss!”
   She gave me a sounding smack on the shoulder and I turned on
her furiously.
   “Oh, yes!” she cried, “hit out at me, do! It’s like you.”
   “I won’t touch you!” I said. “But I won’t have anything more to do
with you,” and I strode on, fuming. She followed after me and
presently I heard her crying again. At this my anger evaporated and I
turned round once more.
   “Come on,” I said, “if you want to, and keep a civil tongue in your
head.”
   Presently we were walking together again.
   “What’s your home, Renny?” she asked, by and by.
   “A mill,” I answered, “but nothing is ground there now.”
   She stopped and so did I, and she looked at me curiously, with her
red lips parted, so that her teeth twinkled.
   “What’s the matter?” said I.
   “Nothing,” she said, “only I remember an old, old saying that the
woman told me.”
   “What woman?” I asked, in wonder, but she took no notice of my
question, only repeated some queer doggerel that ran somewhat as
follows:
                   “Where the mill race is
                   Come and go faces.
                   Once deeds of violence;
                   Now dust and silence.
                   Thither thy destiny
                   Answer what speaks to thee.”
                          CHAPTER III.
                   THE MILL AND THE CHANGELING.
   The outer appearance of the old mill in which we lived and grew
up I have touched upon; and now I take up my pen to paint in black
and white the old, moldering interior of the shell.
   The building stood upon a triple arch of red brick that spanned the
stream, and extended from shore to shore, where, on each side, a
house of later date stood cheek to jowl with it. It looked but an
indifferent affair as viewed from the little bridge aforesaid, which was
dedicated to St. Swithun of watery memory, but in reality extended
further backward than one might have suspected. Moreover, to the
east side a longish wing, with a ridged roof of tiles, ran off at right
angles and added considerably to the general dimensions. To the
west stood a covered yard, where once the mill wagons were packed
or unloaded; but this, in all my memory of it, yawned only a dusty
spave, given over to the echoes and a couple of ancient cart wheels
whose rusty tires and worm-pierced hubs were mute evidence of an
inglorious decay.
   These were for all to see—but behind the walls!
   Was the old mill uncanny from the first, or is it only the ghosts with
which our generation of passions has peopled it that have made it
so? This I can say: That I never remember a time when Jason or I,
or even Zyp, dared to be in the room of silence alone—and in
company never for more than a few minutes. Modred had not the
same awe of it, but Modred’s imagination was a swaddled infant. For
my father I will not speak. Maybe he was too accustomed to specters
to dread them.
   This room was one on the floor above the water, and the fact that
it harbored the mill wheel, whose booming, when in motion, shook
the stagnant air with discordant sounds, may have served as some
explanation of its eeriness. It stood against the east wing and away
from the yard, and was a dismal, dull place, like a loft, with black
beams above going off into darkness. Its only light came from a
square little window in front that was bleared with dust and stopped
outside with a lacework of wire. Against its western wall was reared
a huge box or cage of wood, which was made to contain the upper
half of the wheel, with its ratchet and shaft that went up to the great
stones on the floor above; for the mill race thundered below, and
when the great paddles were revolving the water slapped and rent at
the woodwork.
   Now it behooves me to mention a strange fancy of my father’s—
which was this, that though no grain or husk in our day ever
crumbled between the stones, the wheel was forever kept in motion,
as if our fortunes lay in grinding against impalpable time. The custom
was in itself ghostly, and its regularity was interrupted only at odd
moments, and those generally in the night, when, lying abed
upstairs, we boys would become conscious of a temporary cessation
of the humming, vibrating noise that was so habitual to the place. To
this fancy was added a strange solicitude on the part of my father for
the well-being of the wheel itself. He would disappear into the room
of silence twice or thrice a day to oil and examine it, and if rarely any
tinkering was called for we knew it by the sound of the closing of the
sluice and of the water rush swerving round by another channel.
   Now, for the time I have said enough, and with a sigh return to that
May afternoon and little Zyp, the changeling.
   She followed me into the mill so quietly that I hardly heard her step
behind me. When I looked back her eyes were full of a strange
speculation and her hands crossed on her breast, as if she prayed.
She motioned me forward and I obeyed, marveling at my own
submission. I had no slightest idea what I was to say to my father or
what propose. We found him seated by the table in the living room
upstairs, a bottle and glass before him. The weekly demon was
beginning to work, but had not yet obtained the mastery. He stared
at us as we entered, but said nothing.
   Then, to my wonder, Zyp walked straight up to the old man, pulled
his arms down, sat upon his knee and kissed his rutted cheek. I gave
a gasp that was echoed by Jason, who had followed and was
leaning against the lintel of the open door. Still my father said nothing
and I trembled at the ominous silence. At last in desperation I
stammered, and all the time Zyp was caressing the passive face.
   “Dad, the girl fell into the water and I pulled her out, and here she
is.”
   Then at length my father said in a harsh, deep voice:
   “You pulled her out? What was Jason there doing?”
   “Waiting for her to drown,” my brother answered for himself,
defiantly forestalling conviction.
   My father put the girl from him, strode furiously across the room,
seized Jason by one arm and gave him several cruel, heavy blows
across his shoulders and the back of his head. The boy was half
stunned, but uttered no cry, and at every stroke Zyp laughed and
clapped her hands. Then, flinging his victim to the floor, from which
he immediately rose again and resumed his former posture by the
door, pale but unsubdued, my father returned to his seat and held
the girl at arm’s length before him.
   “Who are you?” he said.
   She answered, “A changeling,” in a voice soft as flowers.
   “What’s your name?”
   “Zyp.”
   “Your other name?”
   “Never mind; Zyp’s enough.”
   “Is it? Where do you come from? What brings you here?”
   “Renny brought me here because I love him.”
   “Love him? Have you ever met before?”
   “No; but he pulled me out of the water.”
   “Come—this won’t do. I must know more about you.”
   She laughed and put out her hand coaxingly.
   “Shall I tell you? A little, perhaps. I am from a big forest out west
there, where wheels drone like hornets among the trees and black
men rise out of the ground. I have no father or mother, for I come of
the fairies. Those who stood for them married late and had a baby
and they delayed to christen it. One day the baby was gone and I
was there. They knew me for a changeling from the first and didn’t
love me. But I lived with them for all that and they got to hate me
more and more. Not a cow died or a gammer was wryed wi’ the
rheumatics but I had done it. Bit by bit the old man lost all his trade
and loved me none the more, I can tell you. He was a Beast Leech,
and where was the use of the forest folk sending for him to mend
their sick kine when he kept a changeling to undo it all? At last they
could stand no more of it and the woman brought me away and lost
me.”
   “Lost you?” echoed my father.
   “Oh,” said Zyp, with a little cluck, “I knew all along how the tramp
was to end. There was an old one, a woman, lived in the forest, and
she told me a deal of things. She knew me better than them all, and I
loved her because she was evil, so they said. She told me some
rhymes and plenty of other things.”
   “Well?” said my father.
   “We walked east by the sun for days and days. Then we came to
the top of a big, soft hill, where little beetles were hopping among the
grass, and below us was a great town like stones in a green old
quarry, and the woman said: ‘Run down and ask the name of it while
I rest here.’ And I ran with the wind in my face and was joyful, for I
knew that she would escape when I was gone, and I should never
see her again.”
   “And then you tumbled into the water?” said my father.
   Zyp nodded.
   “And now,” she said, “I belong to nobody, and will you have me?”
   My father shook his head, and in a moment sobs most piteous
were shaking the girl’s throat. So forlorn and pretty a sight I have
never seen before or since.
   “Well,” he said, “if nobody comes to claim you, you may stop.”
   And stop Zyp did. Surely was never an odder coming, yet from
that day she was one of us.
   What was truthful and what imaginative in her story I have never
known, for from first to last this was the most we heard of it.
   One thing was certain. Zyp was by nature a child of the open air
and the sun. Flowers that were wild she loved—not those that were
cultivated, however beautiful, of which she was indifferent—and she
had an unspeakable imagination in reading their fanciful histories
and a strange faculty for fondling them, as it were, into sentient
beings. I can hardly claim belief when I say that I have seen a rough
nettle fade when she scolded it for stinging her finger, or a little
yellow rock rose turn from the sun to her when she talked to it.
   Zyp never plucked a flower, or allowed us to do so if she could
prevent it. I well remember the first walk I took with her after her
establishment in the mill, when I was attracted by a rare little
blossom, the water chickweed, which sprouted from a grassy trench,
and pulled it for her behoof. She beat me savagely with her soft
hands, then fell to kissing and weeping over the torn little weed,
which actually appeared to revive a moment under her caresses. I
had to promise with humility never to gather another wild flower so
long as I lived, and I have been faithful to my trust.
   The afternoon of her coming old Peg rigged her up some
description of sleeping accommodation in a little room in the attic,
and this became her sanctuary whenever she wished to escape us
and be alone. To my father she was uniformly sweet and coaxing,
and he for his part took a strange fancy to her, and abated somewhat
of his demoniacal moodiness from the date of her arrival.
   Yet it must not be imagined, from this description of her softer side,
that Zyp was all tender pliability. On the contrary, in her general
relations with us and others as impure human beings, she was the
veritable soul of impishness, and played a thousand pranks to prove
her title to her parentage.
   At first she made a feint of distributing her smiles willfully, by turn,
between Modred and me, so that neither of us might claim
precedence. But Jason was admitted to no pretense of rivalry;
though, to do him justice, he at once took the upper hand by meeting
scorn with indifference. In my heart, however, I claimed her as my
especial property; a demand justified, I felt no doubt, by her manner
toward me, which was marked by a peculiar rebellious tenderness
she showed to no other.
   The day after her arrival she asked me to take her over the mill
and show her everything. I complied when the place was empty of all
save us. We explored room by room, with a single exception, the
ancient building.
   Of course Zyp said: “There’s a room you haven’t shown me,
Renny.”
   “Yes,” said I; “the room of silence.”
   “Why didn’t we go there?”
   “Never mind. There’s something wicked in it.”
   “What? Do tell me! Oh, I should love to see!”
   “There’s nothing to see. Let it alone, can’t you?”
   “You’re a coward. I’ll get the sleepy boy to show me.”
   “Come along then,” I said, and, seizing her hand, dragged her
roughly indoors.
   We crossed a dark passage, and, pushing back a heavy door of
ancient timber, stood on the threshold of the room of silence. It was
not in nature’s meaning that the name was bestowed, for, entering,
the full voice of the wheel broke upon one with a grinding fury that
shook the moldering boards of the floor.
   “Well,” I whispered, “have you seen enough?”
   “I see nothing,” she cried, with a shrill, defiant laugh; “I am going
in”—and before I could stop her, she had run into the middle of the
room and was standing still in the bar of sunlight, with her arms
outspread like wings, and her face, the lips apart, lifted with an
expression on it of eager inquiry.
   What happened? I can find an image only in the poison bottle of
the entomologist. As some shining, flower-stained butterfly, slipped
into this glass coffin, quivers, droops its wings and fades, as it were,
in a moment before its capturer’s eyes, so Zyp faded before mine.
Her arms dropped to her sides, her figure seemed as if its whole
buoyancy were gone at a touch, her face fell to a waxen color and
“Oh, take me away!” she wailed in a thin, strangled voice.
   I conquered my terror, rushed to her, and, dragging her stumbling
and tripping from the room, banged to the door behind us and made
for the little platform once more and the open air.
   She revived in a wonderfully short space of time, and, lifting up her
head, looked into my eyes with her own wide with dismay.
   “It was hideous,” she whispered; “why didn’t you stop me?”
   Zyp, it will be seen, was not all elf. She had something in common
with her sex.
   “I warned you,” I said, “and I know what you felt.”
   “It was as if a question was being asked of me,” she said, in a low
voice. “And yet no one spoke and there was no question. I don’t
know what it wanted or what were the words, for there were none;
but I feel as if I shall have to go on thinking of the answer and
struggling to find it forever and ever.”
  “Yes,” I whispered, in the same tone; “that is what everybody
says.”
  She begged me not to follow her, and crept away quite humbled
and subdued, and we none of us saw more of her that day. But just
as she left me she turned and whispered in awe-stricken tone,
“Answer what speaks to thee,” and I could not remember when and
where I had heard these words before.
                          CHAPTER IV.
                           ZYP BEWITCHES.
   Zyp had been with us a month, and surely never did changeling
happen into a more congenial household.
   Jason she still held at arm’s length, which, despite my admiration
of my brother, I secretly congratulated my heart on, for—let me get
over it at the outset—from first to last, I have never wavered in my
passion of love for this wild, beautiful creature. The unexpectedness
of her coming alone was a romance, the delight of which has never
palled upon me with the deadening years. Therefore it was that I
early made acquaintance with the demon of jealousy, than whom
none, in truth, is more irresistible in his unclean strength and
hideousness.
   Zyp and I were one day wandering under the shadow of the
mighty old cathedral of Winton.
   “I don’t like it, Renny,” she said, pressing up close to me. “It’s awful
and it’s grand, but there are always faces at the windows when I look
up at them.”
   “Whose?” I said, with a laugh.
   “I don’t know,” she said; “but think of the thousands of old monks
and things whose home it was once and whose ghosts are shut up
among the stones. There!” she cried, pointing.
   I looked at the old leaded window she indicated, but could see
nothing.
   “His face is like stone and he’s beckoning,” she whispered. “Oh,
come along, Renny”—and she dragged me out of the grassy yard
and never stopped hurrying me on till we reached the meadows.
Here her gayety returned to her, and she felt at home among the
flowers at once.
   Presently we wandered into a grassy covert against a hedge on
the further side of which a road ran, and threw ourselves among the
“sauce alone” and wild parsley that grew there. Zyp was in one of
her softest moods and my young heart fluttered within me. She
leaned over me as I sat and talked to me in a low voice, with her fair
young brow gone into wrinkles of thoughtfulness.
   “Renny, what’s love that they talk about?”
   I laughed and no doubt blushed.
   “I mean,” she said, “is it blue eyes and golden hair or brown eyes
and brown hair? Don’t be silly, little boy, till you know what I mean.”
   “Well, what do you mean, Zyp?”
   “I want to know, that’s all. Renny, do you remember my asking to
kiss and be friends that day we first met, and your refusing?”
   “Yes, Zyp,” I stammered.
   “You may kiss me now, if you like,” and she let herself drop into my
arms, as I sat there, and turned up her pretty cheek to my mouth.
   My blood surged in my ears. I was half-frightened, but all with a
delicious guilt upon me. I bent hastily and touched the soft pink curve
with my trembling lips.
   She lay quite still a moment, then sat up and gently drew away
from me.
   “No,” she said, “that isn’t it. Shall I ever know, I wonder?”
   “Know what, Zyp?”
   “Never mind, for I shan’t tell you. There, I didn’t mean to be rude,”
and she stroked the sleeve of my jacket caressingly.
   By and by she said: “I wonder if you will suffer, Renny, poor boy? I
would save you all if I could, for you’re the best of them, I believe.”
   Her very words were so inexplicable to me that I could only sit and
stare at her. I have construed them since, with a knife through my
heart for every letter.
   As we were sitting silent a little space, steps sounded down the
road and voices with them. They were of two men, who stopped
suddenly, as they came over against us, hidden behind the hedge,
as if to clinch some argument, but we had already recognized the
contrary tones of my father and Dr. Crackenthorpe.
   “Now, harkee!” the doctor was saying; “that’s well and good, but
I’m not to be baffled forever and a day, Mr. Ralph Trender. What
does it all amount to? You’ve got something hidden up your sleeve
and I want to know what it is.”
   “Is that all?” My father spoke in a set, deep manner.
   “That’s all, and enough.”
   “Then, look up my sleeve, Dr. Crackenthorpe—if you can.”
   “I don’t propose to look. I suggest that you just shake it, when no
doubt the you-know-whats will come tumbling out.”
   “And if I refuse?”
   “There are laws, my friend, laws—iniquitous, if you like; but, for
what they are, they don’t recognize the purse on the highway as the
property of him that picks it up.”
   “And how are you going to set these laws in motion?”
   “We’ll insert the end of the wedge first—say in some public print,
now. How would this look? We have it on good authority that Mr.
Trender, our esteemed fellow-townsman, is the lucky discoverer of
——”
   “Be silent, you!” My father spoke fiercely; then added in a low tone:
“D’ye wish all the world to know?”
   “Not by any means,” said the other, quietly, “and they shan’t if you
fall in with my mood.”
   “If I only once had your head in the mill wheel,” groaned my father,
with a curse. “Now, harken! I don’t put much value on your threat; but
this I’ll allow that I court no interference with my manner of life. Take
the concession for what it is worth. Come to me by and by and you
shall have another.”
   “A couple,” said the doctor.
   “Very well—no more, though I rot for it—and take my blessing with
them.”
   “When shall I come?” said the doctor, ignoring the very equivocal
benediction.
   “Come to-night—no, to-morrow,” said my father, and turning on his
heel strode heavily off toward the town.
   I heard the doctor chuckling softly with a malignant triumph in his
note.
   I clenched my teeth and fists and would have risen had not Zyp
noiselessly prevented me. It was wormwood to me; the revelation
that, for some secret cause, my father, the strong, irresistible and
independent, was under the thumb of an alien. But the doctor walked
off and I fell silent.
   On our homeward way we came across Jason lying on his back
under a tree, but he took no notice of us nor answered my call, and
Zyp stamped her foot when I offered to delay and speak to him.
Nevertheless I noticed that more than once she looked back, as long
as he was in view, to see if he was moved to any curiosity as to our
movements, which he never appeared to be in the least.
    Great clouds had been gathering all the afternoon, and now the
first swollen drops of an advancing thunderstorm spattered in the
dust outside the yard. Inside it was as dark as pitch, and I had
almost to grope my way along the familiar passages. Zyp ran away
to her own den.
    Suddenly, with a leap of the blood, I saw that some faintly pallid
object stood against the door of the room of silence as I neared it. It
was only with an effort I could proceed, and then the thing detached
itself and was resolved into the white face of my brother Modred.
    “Is that you, Renny?” he said, in a loud, tremulous voice.
    “Yes,” I answered, very shakily myself. “What in the name of
mystery are you doing there?”
    “I feel queer,” he said. “Let’s get to the light somewhere.”
    We made our way to the back, opened the door leading on to the
little platform and stood looking at the stringed rain. Modred’s face
was ghastly and his eyes were awakened to an expression that I had
never thought them capable of.
    “You’ve been in there?” I said.
    “Yes,” he whispered.
    “More fool you. If you like to tempt the devil you should have the
brass to outface him. Why, you’ve got it!” I cried, for he suddenly let
fall from his trembling hand a little round glittering object, whose
nature I could not determine in the stormy twilight.
    He had it in his clutch again in a moment, though I pounced for it,
and then he backed through the open doorway.
    “It’s naught that concerns you,” he said; “keep off, you beast!”
    “What is it?” I cried.
    “Water-parings,” said he, and clapped to the door in my face as I
rushed at him, and I heard him scuttle upstairs. The latch caught me
in the chest and knocked my breath out for a bit, so that I was unable
to follow, and probably he ran and bolted himself into his bedroom. In
any case, I had no mind for pursuit, my heart being busy with other
affairs; and there I remained and thought them out. Presently, being
well braced to the ordeal, I went indoors and upstairs to the living
room, where I was persuaded I should find my father. And there he
sat, pretty hot with drink and with a comfortless, glowering devil in
his eyes.
   “Well!” he thundered, “what do you want?”
   I managed to get out, with some firmness, “A word with you, dad,”
though his eyes disquieted me.
   “Make it one, then, and a quick one!”
   “Zyp and I were sitting behind a hedge this afternoon when you
and Dr. Crackenthorpe were at words on the other side.”
   His eyes shriveled me, but the motion of his lips seemed to signify
to me that I was to go on.
   “Dad, if he has any hold over you, let me share the bother and
help if I can.”
   He had sat with his right hand on the neck of the bottle from which
he had been drinking, and he now flung the latter at me, with a snarl
like that of a mad dog. Fortunately for me, in the very act some flash
of impulse unnerved him, so that the bottle spun up to the ceiling and
crashed down again to the floor, from which the scattered liquor sent
up a pungent, sickening odor. Then he leaped to his feet and yelled
at me. I could make nothing of his words, save that they clashed into
one another in a torrent of furious invective. But in the midst his
voice stopped, with a vibrating snap; he put his hand to his forehead,
which, I saw with horror, was suddenly streaked with purple, and
down he sunk to the floor in a heap.
   I was terribly frightened, and, running to him, endeavored in a
frantic manner to pull him into a sitting posture. I had half succeeded,
when, lying propped up against the leg of the table, he gave a groan
and bade me in a weak voice to let him be; and presently to my joy I
saw the natural color come back to his face by slow degrees. By and
by he was able to slide into the chair he had left, where he lay
panting and exhausted, but recovering.
   “Renalt, my lad,” he said, in a dragging voice, “what was that you
said just now? Let’s have it again.”
   I hesitated, but he smiled at me and bade me not to fear. Thus
encouraged, I repeated my statement.
   “Ah,” he said; “and the girl—did she hear?”
   “She couldn’t help it, dad. But she can’t have noticed much, for
she never even referred to it afterward.”
   “Which looks bad, and so much for your profound knowledge of
the sex.”
   He looked at me keenly for some moments from under his matted
eyebrows; then muttered as if to himself:
   “Here’s a growing lad, and loyal, I believe. What if I took him a
yard into my confidence?”
   “Oh, yes, dad,” I said, eagerly. “You can trust me, indeed you can.
I only want to be of some use.”
   He slightly shook his head, then seemed to wake up all of a
sudden.
   “There,” he said; “be off, like a good boy, and don’t worry me a
second time. You meant well, and I’m not offended.”
   “Yes, dad,” I said a little sadly, and was turning to go, when he
spoke to me again:
   “And if the girl should mention this matter—you know what—to
you, say what I tell you now—that Dr. Crackenthorpe thinks your
father can tell him where more coins are to be found like the one I
gave him that night; but that your father can’t and is under no
obligation to Dr. Crackenthorpe—none whatever.”
   So I left him, puzzled, a little depressed, but proud to be the
recipient of even this crumb of confidence on the part of so reserved
and terrible a man.
   Still I could not but feel that there was something inconsistent in
his words to me and those I had heard him address to the doctor.
Without a doubt his utterances on the road had pointed to a certain
recognition of the necessity of bribing the other to silence.