Presentation on theme: "CPH EXAM REVIEW– EPIDEMIOLOGY"— Presentation transcript:
1 CPH EXAM REVIEW– EPIDEMIOLOGY Lina Lander, Sc.D.Associate ProfessorDepartment of Epidemiology, College of Public HealthUniversity of Nebraska Medical CenterJanuary 24, 2014
2 Review of basic topics covered in the epidemiology section of the exam Materials covered cannot replace basic epidemiology courseThis review will be archived on the NBPHE website under Study Resources
3 Outline Overview Terminology Study design Causation and validity Screening
4 Populations Group of people with a common characteristic Place of residence, age, gender, religionPeople who live in Omaha, Nebraska in January, 2014Occurrence of a life event (undergoing cancer treatment, giving birth)This is different from the study of disease in individuals, and populations are at the heart of epidemiology. What you will find in epidemiology is that everything that involves measurement must be defined as precisely as possible. This includes populations. In general, a population is a group of people with a common characteristic. This common characteristic could be as broad as place of residence or age or gender, or as specific as people who live along the Missouri River, or within one mile of a nuclear power plant. Populations can also be defined by occurrence of a life event, such as a medical procedure, giving birth, or entering college. So, for example, you might be interested in evaluating a treatment in men who have been diagnosed with cardiac arrhythmia. The important aspect of defining the population is to identify men that have been diagnosed with cardiac arrhythmia. That is the population in this example.
5 Populations Membership can be permanent or transient Population with permanent membership is referred to as “Fixed” or “Closed”People present at HiroshimaPassengers on an airplanePopulation with transient membership is referred to as “Dynamic” or “Open”Population of OmahaBecause populations can be defined by just about anything, a unifying concept for identifying a population is whether the membership can be permanent or transient. When a population has a permanent membership, we call it a fixed or closed population. An example of this is the Japanese who were present at Hiroshima when the atomic bomb was dropped. No one else can ever be added to this population because this was an historical event. Populations with transient membership are referred to as dynamic or open populations. An example of this is the population of Omaha, where people are migrating in, immigrating out, and people are being born and dying. So new members can be added. Key concepts in the transient population is migration, birth and death.
7 Measures of Frequency “Count” - the most basic epidemiologic measure Expressed as integers (1, 2, 3, …)Answers the question, “How many people have this disease?”Often the numerator of many measuresImportant to distinguish between incident (new) and prevalent (existing) casesCounts is the most basic epidemiologic measure. It is never negative and can reach infinity. This measure answers the question “How many people have this disease”? And constitutes the numerator of most epidemiologic measures.
8 Forecast of cancer deaths This graph shows the number of cancer deaths in the United States from 1900 to 2000, and you can tell that the number of deaths are increasing monotonically over time. Again, at first glance, you may think that life in the US has become more dangerous and that more people are dying every day. What is missing from this information?
9 Ratio One number (x) divided by another (y): 𝑥 𝑦 Range: zero (0) to infinity (∞)(x) and (y) may be related or completely independentSex of childrenAttending a clinicA ratio is simply one number divided by another. So, x divided by y. You can see that since most epidemiology measures are of this form, most epidemiologic measures are RATIOS. Interestingly, though, only a few epidemiologic measures are actually CALLED ratios. Two examples of ratios are shown at the bottom of the slide. For example, of all children coming into the clinic, what is the ratio of females to males? The numerator and denominator are clearly not related in this measure as they are in the second measure, which is the number of females divided by all the children coming into the clinic. What is this measure? Yes, this is a proportion. So you can see that a proportion is always a ratio, but a ratio is not always a proportion.
10 Which of the following terms is expressed as a ratio (as distinguished from a proportion)? (A) Male Births / Male + Female Births (B) Female Births / Male + Female Births (C) Male Births / Female Births (D) Stillbirths / Male + Female Births
11 ProportionRatio in which the numerator (x) is included in the denominator (x+y):Range: zero (0) to one (1)Often expressed as percentage ( e.g., Among all children who attended a clinic, what proportion was female?We saw proportion in the last slide, but here is the example again. And you can see that it is different from a ratio in two ways. One – the upper range stops at 1 or 100% if you are expressing it as a percentage, and the numerator is included in the denominator.
12 RateCan be expressed as (a/T) where (a) = cases and (T) involves a component of timeRange: zero (0) to infinity (∞)Measures speed at which things happenRate of at-risk females coming to a clinic (time)Rate is the most misused measure of frequency. This is likely due to the use of the term in many science disciplines. Rate is expressed as the number of cases divided by a unit of time, rather than a number of people in a population. The often confusing aspect of rate is that a population of risk is used in the denominator, but it is multiplied by the amount of time the population is being followed. Rate measures how quickly things happen.
13 Prevalence Proportion Not a rate – no time component in the calculationMeasures proportion of existing disease in the population at a given time“Snapshot”Dimensionless, positive number (0 to 1)
14 Prevalence proportion Where:A = number of existing casesB = number of non-casesN = total population
15 IncidenceMeasures the occurrence of new cases in a population at risk over timeCan be measured as a proportion or a rateThe most fundamental epidemiologic indicatorMeasures force of morbidity (as a rate)Measures conversion of health status (proportion /rate)In addition to the form of the measure, we also need to distinguish whether we are talking about incident or prevalent disease. Incidence measures the occurrence of new cases in a population over a specified period of time. Incidence is usually considered the most fundamental epidemiologic indicator because it measures the conversion of health state. As a rate, incidence measures the force of morbidity (or mortality if we’re measuring death) over time.
16 Incidence proportion Synonyms: incidence, cumulative incidence, risk Measures probability (risk) of developing disease during period of observationDimensionless, positive number (0 to 1)Incidence proportion, then, measures the probability of developing a disease during a period of observation. This is what we measure when we talk about RISK of disease. Risk is the only measure that can be interpreted on an individual level. So, when you go to the doctor, you could actually ask, what is my risk of getting this disease.
17 Incidence Proportion Where: a = number of new onset cases (events) The basic measure of incidence is incidence proportion, which is simply the number of new cases of disease divided by the Number of people in a population at risk at the beginning of a time period.Where:a = number of new onset cases (events)N = population-at-risk at beginning
18 Incidence ProportionAppropriate for fixed (closed) populations and short follow-upMust specify time period of observation because risk changes with timeNot appropriate for long-term follow-up due to potential loss of subjectsAssuming: complete follow-up, same risk over time
19 Follow 2000 newborns at monthly intervals to measure development of respiratory infection in the first yearSuppose 50 infants develop respiratory infection in first year of lifeThe risk (probability) of developing a respiratory infection in the first year of life is ~ 2.5%25 of 1000 infants in this population or 1 in 40 will develop infection in the first year of life.
20 Incidence RateMeasures how rapidly new cases develop during specified time periodCases per person-timeSynonyms: incidence, incidence density, rateFollow-up may be incompleteRisk period not the same for all subjects
21 Incidence Rate Where: a = number of new onset cases T = person-time at risk during study period (follow-up)
22 Person-timeAccounts for all the time each person is in the population at riskThe length of time for each person is called person-timeSum of person-times is called the total person-time at risk for the population
23 Person-time 1 2 3 4 5 Died T = person-time at risk during study period = = 16 person-years
24 Person-time Assumption 100 persons followed 10 years = 1000 person years1000 persons followed for 1 year = 1000 person years
25 Follow 2000 newborns at monthly intervals to measure development of respiratory infection in the first year50 infants develop respiratory infection1900 complete the first year disease free25 complete 3 months (0.25 years) before infection25 complete 6 months (0.5 years) before infectionCalculate incidence rate:= 2.6 per 100 person-years
26 Incidence, Prevalence, Duration Prevalence increases as new cases added to the existing cases (i.e., incidence)Prevalence decreases as people are cured or diePrevalence = Incidence * Duration
28 Mortality Measures the occurrence death Can be measured as a proportion or a rateCan measure disease severity or effectiveness of treatment
29 Mortality RateMeasures rate of death in the population over a specified amount of timePositive number (0 to ∞)Can be a measure of incidence rate (risk) when disease is severe and fatal, e.g. pancreatic cancerSynonym: fatality rate
30 Mortality Rate Where: d = number of deaths N = total population at mid-point of time periodT = follow-up time (usually one year)
31 Cancer Death Rates*, for Men, US, 1930-2003 *Age-adjusted to the 2000 US standard population.Source: US Mortality Public Use Data Tapes , US Mortality Volumes ,National Center for Health Statistics, Centers for Disease Control and Prevention, 2002.
32 Case Fatality Rate This is not a rate, this is a proportion Proportion of deaths from a specific illnessCase Fatality RateWhere:a = Number of deaths from an illnessN = Number of people with that illnessWhat percentage of people diagnosed as having a disease die within a certain time after diagnosis?
33 Case-fatality rateCase-fatality – a measure of the severity of the diseaseCase-fatality – can be used to measure benefits of a new therapyAs therapy improves - the case-fatality rate would be expected to declinee.g. AIDS deaths with the invention of ARVs
34 Proportionate Mortality Of all deaths, the proportion caused by a certain diseaseCan determine the leading causes of deathProportion of cause-specific death is dependent on all other causes of deathThis does not tell us the risk of dying from a diseaseProportionate mortality from Cardiovascular Disease in the U.S, in 2013= # of U.S deaths from cardiovascular diseases in 2013 x 1,000Total deaths in the U.S. for 2013
35 Which measure of mortality would you calculate to determine the proportion of all deaths that is caused by heart disease? (A) Case fatality (B) Cause-specific mortality rate (C) Crude mortality rate (D) Proportionate mortality ratio (E) Potential years of life lost
36 Other Mortality Rates Crude Mortality Rate Includes all deaths, total population, in a time periodCause-Specific Mortality RateIncludes deaths from a specific cause, total population, in a time periodAge-Specific Mortality RateIncludes all deaths in specific age group, population in the specific age group, in a time period
38 Mortality Rates (Year = 2000) PanamaSwedenPopulation = 2,899,513Deaths = 13,483Mortality Rate =4.65 per 1000 per yearPopulation = 8,923,569Deaths = 93,430Mortality Rate =10.47 per 1000 per yearWhy do you think Sweden has almost a 2x higher mortality rate?
39 Differences in Mortality Rates Crude mortality rates do not take into account differences between populations such as age
40 Can we remove this confounding by age? Separate (stratify) the population into age groups and calculate rates for each ageCompare age-specific mortality ratesIf two different populations, adjust (standardize) the mortality rates of the two populations, taking into account the age structuresResults in comparable rates between populations or in the same population over time
41 Direct Standardization If the age composition of the populations were the same, would there be any differences in mortality rates?Direct age adjustment is used to remove the effects of age structure on mortality rates in two different populationsApply actual age-specific rates to a standard population (US population 2000)
42 Indirect Standardization When age-specific rates are not available – use age-specific mortality rates from the general population to calculate expected number of deathsStandardized mortality ratios (SMR)= observed deaths/ expected deathsIf the age composition of the populations were the same, would there be any differences in mortality rates?
44 Study DesignExperimental studies (Clinical Trial, Randomized Controlled Trial)Observational studiesCohortCase-controlCross-sectionalEcological
45 Experimental studies are characterized by: The population under study: who is eligible for study entry?The intervention(s) being used or compared: what treatment(s) are being used? (Therapeutic (e.g., drug) or preventive (e.g., education)The method of treatment assignment: how are subjects assigned to intervention(s)?The outcomes of interest: how will success be measured?
46 Randomized Controlled Trials A randomized controlled trial is a type of experimental research design for comparing different treatments, in which the assignment of treatments to patients is made by a random mechanism.Customary to present table of patient characteristics to show that the randomization resulted in a balance in patient characteristics.
48 Steps in carrying out a clinical trial Select a sample from the populationEthical considerationsMeasure baseline variablesRandomizeApply interventionsFollow up the cohortsMeasure outcome variables (blindly, if possible)
49 Use of “Blinding”Important when knowing treatment could influence the interpretation of resultsEspecially important when outcomes are subjective (pain, functional status) and/or when placebo is employed (either alone or to mask actual treatment)Placebo- ensure control and treatment group have same “experience”May not be necessary if the outcome is an object measure (death, blood glucose)
50 TreatBlinding of the participants to which treatment was used ensures that bias is avoidedSingle-blind: patient does not know what treatment they are receivingDouble–blind: patient and investigator do not know what treatment (cannot be used for some treatments, e.g. surgery)Once we have comparable groups, we treat in a double-blind fashion and according to standard practice.
51 It All Comes Down to…Obtaining groups that are comparable for everything except the treatment…So that differences in outcome can fairly be ascribed only to the difference between the groups (i.e., to the treatment).
53 Cohort StudiesDefinition: groups, defined on the basis of some characteristics (often exposure and non-exposure) are typically prospectively followed to see whether an outcome of interest occurs (may also be retrospective)Comparison of interest: Compare the proportion of persons with the disease in the exposed group to the proportion with the disease in the unexposed group.Motivation: If the exposure is associated with the disease, we expect that the proportion of persons with the disease in the exposed group will be greater than the proportion with disease in the unexposed group.
54 Cohort Studies Not Exposed Exposed Not diseased Not diseased Diseased ProspectiveRetrospectivepastnowNot ExposedDiseasedNot diseasedExposedDiseasedNot diseasedfuturenow
55 Prospective cohort studies Define sample free of the disease/outcome of interest, measure the exposure and classify to exposed vs unexposed at time zero, then follow up at fixed time point to ascertain outcomeMeasure the ratio of outcome between the exposed and unexposed (Relative Risk)
56 Retrospective cohort studies Synonyms: historical cohort study, historical prospective study, non-concurrent prospective studyDo not design retrospective cohort studies a priori – question always in retrospectExposures and Outcomes have already occurred - data on the relevant exposures and outcomes must have been collected
57 Cohort study strengths May be used to define incidence / natural historyKnown temporal sequenceEfficient in investigating rare exposuresPermits study of multiple exposures AND outcomesFewer biases or bias can be limited or evaluatedHomogenous sample populationAccurate measurement of important variablesNo bias in ascertainment of outcome
58 Cohort study limitations Expensive and inefficient – especially for rare diseases or outcomes (large sample size, long-term follow up)Associations may be due to confoundingCan adjust statistically for any measured potential confoundersMust exclude subjects with outcome present at onsetDisease with long pre-clinical phase may not be detectedSensitive to follow-up bias (loss of diseased subjects)
60 Case-control StudiesDefinition: compare various characteristics (past exposure) for cases (subjects with disease) to those of controls (subjects without the disease)Comparison of interest: Compare the proportion with the exposure in the cases to the proportion with the exposure in the control group.Motivation: If the exposure is associated with the disease, we expect that the proportion of persons with the exposure in the cases will be greater than the proportion with the exposure in the control group.
61 Case-control Studies Cases with disease Exposed in past Not Exposed in pastControls withoutdiseaseExposed in pastNot Exposed in past
62 Case-Control Studies Efficient for rare diseases Efficient for diseases with long latencyCan evaluate multiple exposuresLower costs of exposure assessment relative to cohort studiesIncident cases preferable for causal researchChallenges of control selectionChallenges of retrospective exposure assessment
63 Case-control studies are among the best observational designs to study diseases of: (A) High prevalence (B) High validity (C) Low case fatality (D) Low prevalence
64 Nested Case-Control Studies Case-Cohort Study: A case control study nested in a cohort studyControls selected either at baseline (case-cohort) or at the time the case occurs (nested)AdvantageData on exposure are obtained before disease developsPossibility of recall bias is thus eliminated.Less expensive than expanding the analysis to include the entire cohortHere the OR is a statistically unbiased estimate of relative risk
65 Example Is smoking associated with coronary heart disease (CHD)? 3000 smokers and 5000 nonsmokers were followed to see if they developed CHDCase control or cohort study?ExposureDeveloped CHDNo CHDTotalSmokerNon-smoker8487171291649137829300050008000
66 Cross-Sectional Studies Prevalence studiesAll measurements of exposure and outcome are made simultaneously (snapshot)Disease proportions are determined and compared among those with or without the exposure or at varying level of the exposureExamine association – determination of associations with outcomes; generates hypotheses that are the basis for further studiesMost appropriate for studying the associations between chronic diseases and and chronic exposureSometimes useful for common acute diseases of short duration
67 Cross-Sectional Studies TimeDefinedPopulationGather Data on Exposure (Cause) and Disease (Effect / Outcome)T0Exposed:Have DiseaseExposed:No DiseaseNot Exposed:Have DiseaseNot Exposed:No disease
68 Ecological The unit of observation is the population or community Disease rates and exposures are measured in each of a series of populationsDisease and exposure information may be abstracted from published statistics and therefore does not require expensive or time consuming data collection
69 Study Design Category Type Subtype Characteristics Analytic ExperimentalInvestigates prevention and treatmentObservationalInvestigates causes, prevention and treatmentCohortInvestigates health effects of exposureCross-SectionalExamines exposure-disease associations at a point in timeCase-ControlInvestigates risk factors for diseaseEcologicalExamines exposure-disease association at the population levelDescriptiveDescribes health of populations
71 ProbabilitiesDenote the probability of an event by p, where p ranges from 0 to 1.Notation:p = probability that event occurred1-p = probability that event did not occur
72 Example What is the probability of CHD? 171/8000 = 0.02 Exposure Developed CHDNo CHDTotalSmokerNon-smoker8487171291649137829300050008000
73 Relative Risk RR = incidence among exposed incidence among unexposed Approximates how much more likely it is for the outcome to be present among a certain group of subjects than another groupRR = 1 implies that the risk is the same in the two groupsRR < 1 implies that the risk is higher in the unexposedRR > 1 implies that the risk is higher in the exposed
74 Example Sleeping Position and Crib Death Crib Death Usual sleeping positionYESNOTOTALProneOther9683717558461761Total15259226071-year cumulative incidence prone = 9/846 = per 10001-year cumulative incidence other = 6/1761 = per 1000Risk Ratio = per = 3.13.41 per 1000
75 Odds RatiosRelative risk requires an estimate of the incidence of the diseaseFor case control studies, we do not know the incidence of disease because we determine the number of cases and controls when the study is designedFor case control studies, use the odds ratio (OR)
76 Odds Odds are another way of representing a probability The odds is the ratio of probability that the event of interest occurs to the probability that it does not.The odds are often estimated by the ratio of the number of times that the event occurs to the number of times that it does not.
78 Odds Ratio Example CHD Cases Controls Total Smokers 112 176 288 Nonsmokers88224312 Total200400600Case control study of 200 CHD cases and 400 controls to examine association of smoking with CHD(Note: now we are examining the probability of exposure)What is the probability of smoking among CHD cases?p = 112/200=0.56What is the odds of smoking among CHD cases?p/(1-p) = 0.56/0.44 = 1.27
79 Odds Ratio Example CHD Cases Controls Total Smokers 112 176 288 Nonsmokers88224312 Total200400600What is the probability of smoking among controls?p = 176/400=0.44What is the odds of smoking among controls?p/(1-p) = 0.44/0.56 = 0.79The odds ratio is / 0.79 = 1.62Interpretation: The odds of smoking is 1.62 times higher for CHD cases compared with controls
80 Odds Ratio vs. Relative Risk DiseaseNo DiseaseTotalExposedaba+bNot Explosedcdc+da+cb+da+b+c+d
81 Odds ratio Odds ratio = odds of exposure in case odds of exposure in controlsOR=1 exposure is not associated with the diseaseOR>1 exposure is positively associated with the diseaseOR<1 exposure is negatively associated with the disease
82 Odds Ratio vs. Relative Risk Both compare the likelihood of an event between two groupsOR compares the relative odds of an event in each groupRR compares the probability of an event in each groupMore ‘natural’ interpretation because risk measured in terms of probabilityCannot always be computedCan lead to ambiguous interpretations
83 A case-control study comparing ovarian cancer cases with community controls found an odds ratio of 2.0 in relation to exposure to radiation. Which is the correct interpretation of the measure of association?Women exposed to radiation had 2.0 times the risk of ovarian cancer when compared to women not exposed to radiation(B) Women exposed to radiation had 2.0 times the risk of ovarian cancer when compared to women without ovarian cancer(C) Ovarian cancer cases had 2.0 times the odds of exposure to radiation when compared to controls(D) Ovarian cancer cases had 2.0 times the odds of exposure to radiation when compared to women with other cancers
84 Odds Ratios vs. Relative Risks In general, odds ratios summarize associations from case-control studies and cross-sectional studies, and relative risks can be used to summarize associations in cohort studies.Odds ratio can be used to estimate the relative risk when in a case control study when:1. Cases are representative of people with the disease in the population with respect to history of exposure ANDThe controls are representative of people without the disease in the population with respect to history of exposure ANDThe disease is rare
85 Odds ratio estimates relative risk when disease is rare When the disease is rare, the number of people with the disease (a and c) is small so that a+b≈b and c+d≈d
86 Odds Ratios for matched case control studies Often, cases are matched with a control based on age, sex, etc.For a matched study, describe the results for each pairConcordant pairs: both case and control exposed or both not exposedDiscordant pairs: Case exposed/control unexposed or case unexposed/control exposed
87 Odds Ratios for matched case control studies ExposedUnexposedabcdCasesOR is based on the discordant pairs:OR = b/c
88 Cohort study is to risk ratio as: (A) Ecologic fallacy is to cross-sectional study (B) Case-control study is to odds ratio (C) Genetics is to environment (D) Rate ratio is to ecologic study
91 Risk RR and OR measure strength of the association How much of the disease can be attributed to the exposure? How much of the CHD risk experienced by smokers can be attributed to smoking?OR and RR do not address this.
92 Measures of Association Contrast measure of occurrence in two populationsCancer incidence rates in males and females in CanadaIncidence rate of dental caries in children within a community before and after fluoridationBoth of these are measures of association
94 Risk Difference Most often referred to as “attributable risk” Refers to the amount of risk attributable to the exposure of interestFor example, in the birth cohort analysis, where exposure = prenatal care in the first 5 monthsRD = R1 – R0 = Excess risk of preterm birthattributable to prenatal care
95 Absolute Excess Measures Incidence proportion (or rate)Absolute Excess MeasuresIncidence due to exposureExcess risk (or rate) in the exposedIncidence notdue to exposureBackground risk –incidence rate in unexposedUnexposed ExposedIf E is thought to cause D: Among persons exposed to E, what amount of the incidence of D is E responsible for?
96 Example Sleeping Position and Crib Death Crib Death Usual sleeping positionYESNOTOTALProneOther9683717558461761Total15259226071-year cumulative incidence prone = 9/846 = per 10001-year cumulative incidence other = 6/1761 = per 1000Risk difference = per 1000 – 3.41 per 1000 = 7.23 per 1000Added risk due to exposure
97 Attributable Risk Percent (Risk difference / Risk in Exposed) x 100What proportion of occurrence of disease in exposed persons is due to the exposure?
98 Example Sleeping Position and Crib Death Usual sleeping position YES NOTOTALProneOther9683717558461761Total15259226071-year cumulative incidence prone = 9/846 = per 10001-year cumulative incidence other = 6/1761 = per 1000Risk difference = per 1000 – 3.41 per 1000 = 7.23 per 1000Attributable risk percent = per 1000 – 3.41 per 1000 x = 68.0%10.64 per 1000
99 Population Attributable Risk What is the excess risk in the populationcaused by exposure E?
100 Population Attributable Risk Percent What proportion of occurrence of disease in the population is due to the exposure?
101 Population Attributable Risk Incidence proportion (or rate)Unexposed Exposed PopulationShould resources be allocated to controlling E or, instead, to exposures causing greater health problems in the population
102 Example Sleeping Position and Crib Death Crib Death Usual sleeping positionYESNOTOTALProneOther9683717558461761Total15259226071-year cumulative incidence total = 15/2607 = 5.75 per 10001-year cumulative incidence other = 6/1761 = per 1000Population attributable risk (PAR) = 5.75 per 1000 – 3.41 per 1000 == 2.35 per 1000
103 Example Sleeping Position and Crib Death Crib Death Usual sleeping positionYESNOTOTALProneOther9683717558461761Total15259226071-year cumulative incidence total = 15/2607 = 5.75 per 10001-year cumulative incidence other = 6/1761 = per 1000Population attributable risk percent (PAR) == 5.75 per 1000 – 3.41 per 1000 x 100 = 40.8%5.75 per 1000
104 Population Attributable Risk Percent Affected by the prevalence of exposure in thepopulation and the relative risk
105 Absolute Measures Measure Abbrev. Formula Helps answer the question Risk difference (attributable risk to the exposed)RDARI1 – I0If E is thought to cause D: Among persons exposed to E, what amount of the incidence of D is E responsible for?Attributable risk percentAR%[(I1 – I0)/I1] X 100What proportion of occurrence of disease in exposed persons was due to the exposure?Attributable risk to the populationPARIT – I0Should resources be allocated to controlling E or, instead, to exposures causing greater health problems in the population?Attributable risk to the population (%)PAR%[(IT – I0)/IT] X 100What portion of D in the population is caused by E? Should resources allocated for D be directed toward etiologic research or E?
106 Summary of MeasuresAbsolute measures address questions about public health impact of an exposureExcess risk in the exposed or population attributable to the exposureRelative measures address questions about etiology and relations between exposure and outcomeRelative difference in risk between exposed and unexposed populations
109 Definition of a cause“That which produces an effect, result or consequence or the one such as a person, event or condition that is responsible for an action or result” American Heritage DictionaryImplies reason and occasion
110 Key Characteristic of a Cause Essential attributes: association, time order and directionCauses include:host and environmental factorsactive agents and static conditionsCauses may be either positive (presence induces disease) or negative (absence induces disease)
112 Factors involved in the Natural History of Disease AgentVectorIn the case of many communicable diseases, such as malaria, the agent can only reach the host via a third party, called the vector. The vector is animate. For example, the vector for malaria is the female anopheles mosquito. She can convey the malaria parasite to a susceptible host when she consumes a blood meal.EnvironmentHost
113 Risk factors vs. causesRisk factors often used in epidemiology instead of causesA cautious way of making causal inferenceRisk factors are not direct causes of diseaseServe to identify proximate causes
114 Causal InferenceDuring 1950s -1960s epidemiologists developed a set of postulates for causal inferences regarding non-infectious diseases of unknown etiologyResponse to the discovery of association between smoking and lung cancerDebates by many epidemiologists yielded 5 criteria in the 1964 Report of the Advisory committees to the US Surgeon general on Smoking and HealthSir Austin Hill came up with the best known criteria or guidelines in 1965In 1976 Rothman presented a view of causations now known as the “Sufficient-Component Theory of Causation”
115 Hill Criteria 1. Strength of Association 2. Consistency 3. Specificity of the Association4. Temporal relationship5. Biological gradient6. Biologic plausibility7. Coherence8. Experiment9. Analogy
116 Sufficient-Component Cause Model Sufficient cause is a complete causal mechanism that inevitably produces diseaseSufficient cause is not a single factor but rather a minimal set of factors that inevitably produce diseaseSufficient cause for AIDS may include: exposure - HIV infection, susceptibility, lack of preventive exposures-absence of ARVsEach participating factor in a sufficient cause is termed a component cause
117 Disease Causation – 2 components Sufficient Causeprecedes the diseaseif the cause is present, the disease always occursNecessary Causeif the cause is absent, the disease cannot occur1. Using these 2 components of Causation, we can produce a unified framework of causation that will encompass all Disease Processes
118 Disease causation: Types of causal relationships Necessary and sufficient: Without that factor, the disease never develops, and in the presence of that factor, the disease always developsNecessary but not sufficient: Without that factor, the disease never develops but need other factors as wellSufficient but not necessary: The factor can produce the disease, but so can other factor.Neither sufficient not necessary: The factor itself cannot cause the disease but plays a role—multiple factors interact to cause the disease
119 Sufficient-Component Cause Model - attributes Blocking the action of a single component stops the completion of the sufficient cause, thus prevents the disease from occurring by that pathwayCompletion of a sufficient cause is synonymous with the biologic onset of diseaseComponent causes may be distant causes and others may be proximate causesmany causal components remain unknown
120 From Study to Causation Associations between ‘exposures’ and outcomes identified in observational studies may or may not be ‘causal’There is need to pay attention to valid assessment of exposure and outcome in order to think about causalityReliabilityValidityExternal validityInternal validity – three concepts are consideredBiasConfoundingChance (Random error)
122 ValidityImplies that a measure purports to measure what it is expected to measure:AppropriateAccurate (has same numerical value as the phenomenon being investigated, i.e. free of systemic error or bias)Precise (minimal variations are only because of chance or random error)Validity of a study implies that the findings are the “truth”The degree to which a measurement or study reaches a correct conclusionTwo types of validity: Internal validity, External validity
123 External validity: generalizabilty The extent to which the results of a study are applicable to the general populationDo the study results apply to other patients?A representative sample is drawn from the population (usually randomly)Individuals have equal chance to participate in the studyUsually involves a sampling frameInference is made back to the population
124 Internal validityIs the extent to which the results of the study accurately reflect the true situation of the study populationIs influenced by:ChanceThe probability that an observation occurred unpredictability without discernible human intention or observable causeBiasAny systemic error (not random or due to chance) in a study which leads to an incorrect estimate of the association between exposure and diseaseConfoundingThe influence of other variables in a study which leads to an incorrect estimate of the association between exposure and disease
125 Random errorChance“That part of our experience that we cannot predict” (Rothman and Greenland)Usually most easily conceptualized as sampling variability and can be influenced by sample sizeThe term “random error” is widely employed yet less widely understood. Rothman and Greenland (Modern epidemiology, p78) define it as “that part of our experience that we cannot predict” (some might say the part that we do not attempt to predict).The concept can easily be demonstrated in the form of sampling variability, where we can readily show that a series of samples from a single population will in general differ from one another and from the population, at least slightly.
126 Random error can be problematic, but . . . Influence can be reduced– increase sample size – change design of sampling – improve precision of instrumentProbability of some types of influence can be quantified (e.g., confidence interval width)Random error can certainly be problematic, but it is generally not as problematic as systematic error. For one, we know a great deal about random error and how to reduce it. We can, for one, increase sample size. It may also be possible to reduce random error by changing the design of our sampling plan or using more precise measurement instruments. Moreover, the probability of various degrees of influence from random error can be quantified, so that we can use confidence intervals to express the uncertainty inherent in our estimates due to sampling variability.The precision of an estimate is defined as the reciprocal of the standard error of the estimate, but the width of the confidence interval is often used in epidemiology. The difference between the upper and lower 95% confidence limits for a rate or proportion is approximately four (2 x 1.96) times the standard error of the point estimate. The standard errors for many ratio measures (e.g., , risk ratio, rate ratio, odds ratio) are obtained with a log transformation, so the width of their confidence intervals is quantified as the ratio of the upper to the lower confidence limit (called the confidence interval ratio, or CLR). The natural log of the CLR is 2 x 1.96 times the log of the standard error.
127 I. Bias - DefinitionAny systemic error (not random or due to chance) in a study which leads to an incorrect estimate of the association between exposure and disease or outcomeTherefore:Bias is a systematic error that results in an incorrect (invalid) estimate of the measure of association
128 I. Bias - DefinitionCan create spurious association when there is none (bias away from the null)Can mask an association when there is one (bias towards the null)Bias is primarily introduced by the investigator or study participantsBias does not mean that the investigator is “prejudiced”Can occur in all study types: experimental, cohort, case-controlOccurs in the design and conduct of a studyBias can be evaluated but not “fixed” in the analysis phaseTwo main types are selection and observation bias
129 Direction of biasBias towards the null – observed value is closer to 1.0 than is the true valueBias away from the null – observed value is farther from 1.0 than is the true valueObservedTrueNullTrueObserved 2NullObserved 1
130 Types of bias Selection bias Refusals, exclusions, non-participants Failure to enumerate the entire populationLoss to follow upObservation/Information biasDiagnostic (lead time) surveillance biasInterviewer biasRecall biasClassification of exposure and outcomeMisclassification bias (is part of information bias)Non-differentialDifferential
131 II. Selection biasAny systematic error that occurs in the process of identifying study populationsThe error that occurs whenever the identification and selection of individual subjects for inclusion into study is not independent of outcome (cohort) or exposure (case-control)Error due to systematic difference between those selected for study versus those not selected for the study
132 II. Selection BiasResults from procedures used to select subjects into a study that lead to a result different from what would have been obtained from the entire population targeted for the studyMost likely to occur in case-control or retrospective cohort because exposure and outcome have occurred at the time of study selectionSelection bias can also occur in prospective cohort and experimental studies form differential loss to follow-up- impacts which subjects are “selected” for the analysis
133 II. Selection biasOccurs when there is a systematic difference between those selected for study versus those who were notRefusers, non-participants, non-response, exclusionsFailure to enumerate the entire populationThose lost to follow up if related to exposure or outcomeDifferential selection of exposed/unexposed groups, or cases and controlsVolunteersHealthy workers - example
134 II. Selection bias- cohort study Selection bias occurs when selection of exposed and unexposed subjects is not independent of the outcome (so this type can only occur in retrospective cohort study)Examples:A retrospective study of an occupational exposure to asbestos and lung disease in a factory settingThe exposed and unexposed groups are enrolled on the basis of prior employment recordsThe records are old and many are lost, so the complete cohort working in the plant is not available for study.If people who did not develop the disease and were exposed were more likely to have their records lost, then there will be an overestimate of association between the exposure and disease
135 II. Selection bias- cohort study Solutions:Increase participationGet relevant information on refusersDevelop follow-up mechanismsUse comparable populationsValid assessment of outcome in prospective cohort studies (for example: blinding)
136 II. Selection bias: case-control study Sources of selection biasDecisions about selecting incident or prevalent (survival) casesWhen controls do not reflect the population that gave rise to the casesThe selection of cases and controls must be independent of the exposure statusDo controls in the study have higher or lower prevalence of exposure than controls not selected for the study?Cases and controls should have the same exposure opportunities (e.g., welders and general population)
137 II. Selection bias: case-control study Occurs when controls or cases are more or less likely to be included in a study if they have been exposed –inclusion in the study is not independent of exposureResults: relationship between exposure and disease observed among study participants is different from relationship between exposure and disease in eligible individuals who were not includedThe odds ratio from a study that suffers from selection bias will incorrectly represent the relationship between exposure and disease in the overall study population
138 II. Selection bias: cross-sectional study Selection bias can occur whenSampling frame does not represent the true underlying population of interestVoter registration listsDriver’s license recordsTelephone lists
139 II. Selection bias: cross-sectional study Estimates of association do not take into account the sampling structureThere are sufficient numbers of refusers that the underlying sampling structure is compromisedWhen the “sample” is a convenience sample (sampling fraction)
140 II. Selection Bias: solutions? Little or nothing can be done to fix this bias once it has occurred in cross-sectional studiesNeed to avoid it during design and implementation:using the same criteria for selecting cases and controlsobtaining all relevant subject recordsobtaining high participation ratestaking in account diagnostic patterns of disease
142 III. Observation/information bias An error that arises from systematic differences in the way information on exposure or disease is obtained from the study groupsResults in participants who are incorrectly classified as either exposed or unexposed or as diseased or not diseasedOccurs after the subjects have entered the studySeveral types of observation bias: recall bias, interviewer bias, and differential and non-differential misclassification
143 III. Observation/Information bias Recall biasPeople with disease remember or report exposures differently (more/less accurately) than those without diseaseDifferential ability of subject to remember previous activities and exposures, e.g. in serious diseasesCases search their memory to understand their illnessE.g., birth defectsCan result in over-or under-estimation of measure of association
144 III. Observation/Information bias Recall biasSolutions:Use controls who are themselves sickUse standardized questionnaires that obtain complete informationMask subjects to study hypothesis
145 III. Observation/Information bias Interviewer biasSystematic difference in soliciting, recording, interpreting informationCan occur whenever exposure information is sought when outcome is known (as in case-control) or when outcome information is sought when exposure is known (as in cohort study)
146 III. Observation/Information bias Interviewer biasThe way interviewer asks questions, and there is possibility of probing e.g. in-person interviews and telephone interviews, especially where the outcome has already occurred (case-control, and retrospective cohort studies)Solutions:Mask interviewers to study hypothesis and disease or exposure status of subjectsUse standardized questionnaires, or standardized methods of outcome or exposure ascertainmentUse biomarkers to compare when possibleSurrogates tend to underreport exposures
147 III. Observation/Information bias Classification of exposure and outcomeLeads to misclassification biasIf exposure status is known in cohort studies, or outcome status is known in case-control studiesSolution- blinding of data collectors to exposure/outcome status
148 III. Observation/Information bias – Misclassification bias A type of information biasError arising from inaccurate measurement or classification of study subjects or variablesSubject’s exposure or disease status is erroneously classifiedHappens at the assessment of exposure or outcome in both cohort and case-control studiesTwo types: non-differential and differential
149 A. Non-differential misclassification Inaccuracies with respect to disease classification are independent of exposureInaccuracies with respect to exposure are independent of disease statusThe probability of exposure (or of outcome) misclassification is the same for cases and controls (or in study/comparison groups)Bias results towards the null - if the exposure has two categories, will make groups more similar (Type II error)Solution: Use multiple measurements, most accurate sources of information
150 B. Differential Misclassification Probability of misclassification of disease or exposure status differs for exposed and unexposed persons (cohort) or presence of absence of exposure (case-control)Probability of misclassification is different for cases and controls or for levels of exposure within cases and controlsDirection of bias is unknown, i.e. overestimation or underestimation of the true riskKnow that the observed OR deviates from truth, but direction is unknown
151 B. Differential Misclassification Also known as systematic misclassification:The probability of misclassification of disease or exposure status is correlated with presence or absence of characteristic in the study or control group.Thus, misclassification of presence or absence of disease differs for exposed and unexposed persons, or of presence or absence of exposure in cases and controls: it is differential
153 Definition and ImpactAn alternate explanation for the observed association between exposure and disease“A mixing of effects”: the association between exposure and disease is distorted because it is mixed with the effects of another factor that is associated with the diseaseResult of confounding is to distort the true association toward the null (negative confounding) or away from the null (positive confounding)
154 ConfounderA confounder is associated with the exposure and independently of that exposure is a risk factor for the diseaseA confounder has effect on the outcome which can be:overestimating (positive), orunderestimating (negative) or evenchange the direction of the observed effect, a spurious relationship
155 Criteria for a variable to be a confounder The third variable must not be an intermediate link in the causal chain between exposure and outcome (i.e., is not an intermediate or intervening variable)The third variable must cause the outcome event (i.e., must be an independent predictor of disease with or without exposure)The third variable must be associated (correlated) with exposure (but not caused by the exposure)
156 Example:smoking is a confounder of effect of occupational exposures (to dyes) on bladder cancerage is confounder of effect of DDT pesticide exposure and breast cancer
157 Opportunities for confounding In an experimental designs:Participation differs in study and control groupsThere is no evidence for randomizationThere is evidence of residual confoundingIn cohort and case-control studies:When selection of comparison group differs by subject characteristicsWhen risk factors other than the exposure are distributed differently between the exposed and unexposed groups
158 Controlling for confounding In the design phase:Goal is to eliminate or reduce variation in the level of the confounding factor between compared groupsRemember, a variable can only be a confounder if it is different between compared groups.
159 Control for confounding- design phase RandomizationWith sufficient sample size, randomization is likely to control for both known and unknown confounders- but not guaranteedRestrictionRestrict admissibility criteria for study subjects and limit entrance to individuals who fall within a specified category of the confounderMatchingSelect study subjects so that the potential confounders are distributed in an identical manner among the exposed and unexposed groups (cohort study) or among the cases and controls (case-control study)
160 Control for confounding- analysis phase Standardization: by age, race, gender, or calendar time in order to make fair comparisons between populationsStratified analysis: test for homogeneity between strata to check if it is confounding or effect modification, only pool for summary measure if evidence of homogeneityMatched analysis: implement analysis of matched designRestriction: Restrict during data analysisMultivariate analysis: To enable controlling for several potential confounders simultaneously
161 Effect modification Interaction The strength of the association between an exposure and disease differs according to the level of another variable.Modification of the relationship between exposure and a disease by a third variable.If the association changes according to the level of the third variable, then effect modification is present
164 MeasurementMeasurement of exposure, outcome, and other relevant characteristics are a key part of epidemiologic studiesAlmost all tests and measures are imperfect! Knowledge of how well a measure performs helps to:Choose alternative measuresInterpret results of studies using a specific measure
165 ReliabilityHow closely do duplicate measurements of the same characteristic agree with each otherExamples:Test-retest reliability: agreement between responses on a questionnaire that is administered two (or more) times to the same personIntra-observer: agreement of a given interpreter with him/herselfInter-observer: agreement among different interpretersReliability is usually higher with more standardized or automated measurement procedures, lower when more complex judgments are required by human observers (e.g., x-ray reading)
166 ValidityThe degree to which an instrument measures what it sets out to measureFor many epidemiologic applications, the underlying characteristic being measured is dichotomous (e.g., diseased vs. non diseased; exposed vs. not exposed). Then, validity can be regarded as having two components:SensitivitySpecificity
167 Results of Screening Test SensitivityThe probability of testing positive if the disease is truly presentSensitivity = a / (a + c)True Disease Status+-Results of Screening Testabcd
168 Results of Screening Test SpecificityThe probability of screening negative if the disease is truly absentSpecificity= d / (b + d)True Disease Status+-Results of Screening Testabcd
169 Test Disease PRESENT (+) ABSENT (-) Test positive (+) TP FP Test negative (-)FNTNTestSens= TP /(TP+FN)Spec= TN /(TN+FP)PPV= TP /(TP+FP)NPV= TN /(TN+FN)
170 Relationship between Sensitivity and Specificity Lowering the criterion of positivity results in an increased sensitivity, but at the expense of decreased specificityMaking the criterion of positivity more stringent increases the specificity, but at the expense of decreased sensitivityThe goal is to have a high sensitivity and high specificity, but this is often not possible or feasible
171 Relationship between Sensitivity and Specificity The decision for the cut-point involves weighing the consequences of leaving cases undetected (false negatives) against erroneously classifying healthy persons as diseased (false positives)Sensitivity should be increased when the penalty associated with missing a case is highWhen the disease can be spreadWhen subsequent diagnostic evaluations are associated with minimal cost and riskSpecificity should be increased when the costs or risks associated with further diagnostic techniques are substantial (minimize false positives)Example: positive screen requires that a biopsy be performed
172 A screening test is used in the same way in two similar populations, but the proportion of false-positive results among those who test positive in population B is higher than that among those who test positive in population A. What is the most likely explanation for this finding? (A) The specificity of the test is higher in population A (B) The specificity of the test is lower in population A (C) The prevalence of disease is higher in population A (D) The prevalence of disease is lower in population A
173 Performance Yield Predictive Value Positive (PV+) Individuals with a positive screening test results will also test positive on the diagnostic testPredictive Value Negative (PV-)Individuals with a negative screening test results are actually free of disease
174 Results of Screening Test Performance YieldPredictive Value Positive (PV+)The probability that a person actually has a disease given that he/she tests positivePV+ = a / (a + b)Predictive Value Negative (PV- )The probability that a person is truly disease free given that he/she tests negativePV- = d / (c + d)True Disease Status+-Results of Screening Testabcd
175 Performance Yield Factors that influence PV+ and PV- The more specific the test, the higher the PV+The higher the prevalence of preclinical disease in the screened population, the higher the PV+The more sensitive the test, the higher the PV-
176 Sampling errorsRealityTreatments not differentHo trueTreatments are differentHo falseConclude treatments are not differentFail to reject HoCorrect decisionType II errorβConclude treatments are differentReject HoType I errorαPowerProbability =1- βDecisionCorrect :Reject the null hypothesis when it is falseDo not reject the null hypothesis when it is trueErrors:Reject the null hypothesis when it is true ( Type I error =a)Do not reject the null hypothesis when it is false( Type II error =ß)Power = probability of detecting a difference if one truly exists, i.e. probability that a study will find a statistically significant difference, when a difference of a given magnitude truly exists.Power = 1- ß, where beta is the probability of declaring a difference not statistically significant, when a difference truly exists.
178 OutbreakEndemic: Habitual presence of disease within a given geographic areaEpidemic: Occurrence in a community of a group of illnesses of similar nature in excess of what would normally be expected. Amount of disease depends on number susceptible (at risk) and number not susceptible by way of immunization, or genetics (immune).Pandemic: worldwide epidemic
179 Outbreak Types Common source: group of persons exposed to common agent Point-exposed over a brief period of timeIntermittent-exposed over a long period of timePropagated: spreads gradually from person to personMixed epidemic: common source and from person to person
180 How do you find outbreaks? SurveillanceTrack disease/injury rates over timeOnly for reported diseasesTime delay: primary purpose is to examine trends over timeLaboratory reportsHealthcare institutionsPublic health officeObservant healthcare personnelKey Points:Surveillance is what you use to collect the data.Lab reports are a good method of identifying an outbreak. Laboratories can test for resistance factors.If there is an outbreak at another nearby healthcare institution, it might be a good idea to review your own situation and make sure that everything is as it should be.Public Health officials can help to identify an outbreak by contributing lab resources.It is usually an astute clinician or infection preventionist that identifies the outbreak.
181 Epidemic Investigation Establish the presence of an epidemic – case definitionCommunicate/ControlAnalyze the outbreakForm a hypothesisTest the hypothesisComplete the investigation