An Introduction to Epidemiology for Health Professionals

Therearemanygoodepidemiologytextbooksonthemarket,butmostoftheseare addressedtostudentsofpublichealthorpeoplewhodoclinicalresearchwithe- demiologicmethods. Thereisaneedforashortintroductiononhowepidemiologic methodsareusedinpublichealth,geneticandclinicalepidemiology,becausehealth professionalsneedtoknowbasicepidemiologicmethodscoveringetiologicaswell asprognosticfactorsofdiseases. Theyneedtoknowmoreaboutmethodologythan introductorytextsonpublichealthhavetooffer. Insomehealthfaculties,epidemiologyisnotevenpartoftheteachingcurri- lum. Webelievethistobeaseriousmistake. Medicalstudentsarestudentsofall aspectsofdiseasesandhealth. Withoutknowingsomethingaboutepidemiologythe cliniciansandotherhealthprofessionalscannotreadagrowingpartofthesci- ti cliteratureinanyreasonablycriticalwayandcannotnavigateintheworldof “evidence-basedmedicineandevidence-basedprevention. ”Withoutskillsine- demiologicmethodologytheyareinthehandsofexpertsthatmaynotonlyhavean interestinhealth. Some health professionals may believe that only common sense is needed to conductepidemiologicalstudies,butthescienti cliteratureandthepublicdebate onhealthissuesindicatethatcommonsenseisofteninshortsupplyandmaynot thrivewithoutsomeformaltraining. Epidemiologic methods play a key role in identifying environmental, social, and genetic determinants of diseases. Clinical epidemiology addresses the tr- sitionfromdiseasetohealthortowardmortalityorsocialormedicalhandicaps. Publichealthepidemiologyaddressesthetransitionfrombeinghealthytobeingnot healthy. Descriptiveepidemiologyprovidesthediseasepatternthatisneededtolook athealthinabroadperspectiveandtosettheprioritiesright. Epidemiologyisabasic scienceofmedicinewhichaddresseskeyquestionssuchas“Whobecomesill?”and “Whatareimportantprognosticfactors?”Answerstosuchquestionsprovidethe basisforbetterpreventionandtreatmentofdiseases. Many people contributed to the writing of this book: medical students in Denmark,studentsofepidemiologyattheIEAEEPEsummercourseinFlorence, Italy,andstudentsofpublichealthinLosAngeles. Withouttechnicalassistance v vi Preface fromGitteNielsen,JenadeShelley,NinaHoheandPamMasangkaythebookwould neverhavematerialized. LosAngeles,California JørnOlsen Odense,Denmark KaareChristensen IowaCity,Iowa JeffMurray Stockholm,Sweden AndersEkbom Contents Part I Descriptive Epidemiology 1 Measures of Disease Occurrence . . . . . . . . . . . . . . . . . . . 3 IncidenceandPrevalence . . . . . . . . . . . . . . . . . . . . . . . . 4 Incidence. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 RatesandDynamicPopulations . . . . . . . . . . . . . . . . . . . . . 7 CalculatingObservationTime. . . . . . . . . . . . . . . . . . . . . . 9 Prevalence,Incidence,Duration . . . . . . . . . . . . . . . . . . . . . 10 MortalityandLifeExpectancy . . . . . . . . . . . . . . . . . . . . . 11 LifeExpectancy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2 Estimates of Associations . . . . . . . . . . . . . . . . . . . . . . . 15 3 Age Standardization. . . . . . . . . . . . . . . . . . . . . . . . . . 19 4 Causes of Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 5 Descriptive Epidemiology in Public Health. . . . . . . . . . . . . . 29 GraphicalModelsofCausalLinks . . . . . . . . . . . . . . . . . . . 33 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 6 Descriptive Epidemiology in Genetic Epidemiology. . . . . . . . . 37 OccurrenceDatainGeneticEpidemiology . . . . . . . . . . . . . . . 37 ClusteringofTraitsandDiseasesinFamilies . . . . . . . . . . . . . . 38 TheOccurrenceofGeneticDiseases . . . . . . . . . . . . . . . . . . 40 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 7 Descriptive Epidemiology in Clinical Epidemiology. . . . . . . . . 43 SuddenInfantDeathSyndrome(SIDS). . . . . . . . . . . . . . . . . 44 CytologicalScreeningforCervixCancer . . . . . . . . . . . . . . . . 45 ChangesinTreatmentofJuvenileDiabetes . . . . . . . . . . . . . . . 46 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 vii viii Contents Part II Analytical Epidemiology 8 Design Options. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 CommonDesignsUsedtoEstimateAssociations. . . . . . . . . . . . 51 EcologicalStudy . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Case–ControlStudy. . . . . . . . . . . . . . . . . . . . . . . . . . 54 CohortStudy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 ExperimentalStudy . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 9 Follow-Up Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 TheNon-experimentalFollow-Up(Cohort)Study . . . . . . . . . . . 59 StudyingRiskasaFunctionofBMI . . . . . . . . . . . . . . . . . . 60 LongitudinalExposureData. . . . . . . . . . . . . . . . . . . . . . . 62 DifferentTypesofCohortorFollow-UpStudies. . . . . . . . . . . . 63 10 Case–Control Studies. . . . . . . . . . . . . . . . . . . . . . . . . . 67 Case–CohortSampling . . . . . . . . . . . . . . . . . . . . . . . . . 69 DensitySamplingofControls. . . . . . . . . . . . . . . . . . . . . . 69 Case–Non-caseStudy. . . . . . . . . . . . . . . . . . . . . . . . . . 71 PatientControls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 SecondaryIdenti cationoftheSourcePopulation . . . . . . . . . . . 74 Case–ControlStudiesUsingPrevalentCases. . . . . . . . . . . . . . 74 WhentoDoaCase–ControlStudy? . . . . . . . . . . . . . . . . . . . 77 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 11 The Cross-Sectional Study. . . . . . . . . . . . . . . . . . . . . . . 79 12 The Randomized Controlled Trial (RCT) . . . . . . . . . . . . . . 81 Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .