Epi2 Midterm
What calculation should you do for the following: " maternal alcohol consumption and the development of fetal alcohol syndrome in babies born in Oslo Norway, 1990-1993. ? Why? Exposed Not Exposed 60 A 30 B Diseased PERSON TIME 6000 C 15000 D
(60/6000) / (30/15000)= 5.0 "Mothers who consume alcohol during pregnancy are 5 times as likely to have a child with Fetal Alcohol Syndrome as those who do not consume alcohol during pregnancy."
What type of study is this? What else can be said about it? The UK Biobank cohort is a population-based cohort of 500,000 participants recruited in the United Kingdom (UK) between 2006 and 2010. Approximately 9.2 million individuals aged 40-69 years who lived within 25 miles (40 km)of one of 22 assessment centers in England, Wales, and Scotland were invited to enter the cohort, and 5.5% participated in the baseline assessment. The representativeness of the UK Biobank cohort was investigated by comparing demographic characteristics between nonresponders and responders. Sociodemographic, physical, lifestyle, and health-related characteristics of the cohort were compared with nationally representative data sources. UKBiobank participants were more likely to be older, to be female, and to live in less socioeconomically deprived areasthan nonparticipants. Compared with the general population, participants were less likely to be obese, to smoke, and to drink alcohol on a daily basis and had fewer self-reported health conditions. At age 70-74 years, rates of all-cause mortality and total cancer incidence were 46.2% and 11.8% lower, respectively, in men and 55.5% and18.1% lower, respectively, in women than in the general population of the same age. UK Biobank is not representative of the sampling population; there is evidence of a"healthy volunteer" selection bias. Nonetheless, valid assessment of exposure-disease relationships may be widely generalizable and does not require participants to be representative of the population at large.
(Source: Suggested Reading)
Incidence Density Method, Example: ID = 2 deaths = 0.242 deaths/ person year, and we know person time is 8.25. Then, the 3 year risk of death post MI survival,using the incidence density method is:
1 - e -[0.242*3] = 0.513
OBSERVATIONAL studies (how many types??)
1) COHORT 2) CASE-CONTROL 3) ECOLOGIC 4) CROSS-SECTIONAL 5) HYBRIDDDD
Describe types of case-control studies
1) Case-based case-control study (traditional case control study) 2) Case-control study within a cohort aka a] case cohort study or b]nested case control study
Please calculate the following incidence density and provide interpretation: What is the rate of cancer mortality in Ashville, NC from 2000 - 2005? Given: Number of cancer deaths = 50 Population estimate of Ashville, NC = 50,000:
50/50,000 (5)=.0002 Interpretation: 2 cancer deaths per 10,000 person years
Selection bias in a case-based case-control study
A cross-sectional ascertainment identifies primarily prevalent cases, that is, those with the longest survival. Cases and controls who die before they can be included in the study may have a different exposure experience compared with the rest of the source population. Furthermore, exposures may be related with survival but not risk of disease if using prevalent cases. It is preferable to ascertain cases concurrently, i.e. to identify and obtain exposure information from cases as soon as possible after diagnosis. Same rules apply to controls.
This is an example of: Dates of admission, diagnosis of diabetes, age, and gender were collected for all 522,326 children age 21 or younger, of active duty military personnel admitted to US Army treatment facilitates during fiscal years 1971-1991. Incidence rates were expressed as cases per 100,000 person-years.
A descriptive study
What is Cumulative Incidence?
A measure of risk
Measures of Association calculate:
A relative risk to identify relative difference.
What is a cross-sectional study?
A snapshot in time A survey
Life table Method/Approach
A systematic record of a group's mortality or morbidity experience, over the study period which is broken up into specific time intervals Probabilities of survival and of cumulative incidence (CI) are calculated for each interval, Overall cumulative survival is calculated for the study period, & Lets you calculate CI (risk) over specified intervals The Denominator corrects for the LOST of Followup!! This is assuming that the loss of follow up occurs at the mid interval (that's what the 0.5) represents *Remember that its not beginning of time (its interval)*
Odds
A type of disease frequency Provides the ratio fo the probability of the even/ the probability of no event Can be calculated from both incidence and precalance Odds Equation= Event occuring/ 1- event occuring Prevalence/1 - prevalence Incidence / 1- incidence
Do disease occur at random?
Absolutely NOT!
NHANES (National Health and Nutrition Examination Survey) is an example of....
An Ecologic Study! Why? Because it is studying the survey respones aka its a research study
Describe the case-control study design and the rationale for its use
An alternative to the cohort study for assessing exposure-disease relationships Subjects are selected into the study based on their disease status and assessed for previous exposure Exposure data may be measured at the time of the study or gathered from existing data This design optimizes speed and efficiency - BUT Calculate an odds ratio as an estimate of the RR Is particularly susceptible to certain types of bias
Why might there be an increase in rates due to an increase in the number of overall cases?
An expansion in the definition of the case being researched
Which is better: Human or animal studies?
Animal studies b/c humans are harder to study
Population Based Cohort Studies
Any well-defined population (geographic, occupational, membership in HMOs) Typically evaluate multiple hypotheses Primary Justification : external validity
Counts
Are used for prevalent and incident cases Used to monitor the occurrence of disease
When does selection occur for: Nested Case Control?
At the time disease occurs
types of Attributable Risk
Attributable Risk in Exposed aka (Excess Risk) Amount of risk in excess of the event that we can prevent by eliminating the exposure Percent Attributable Risk in Exposed The percent of the total risk in the exposed group attributable to the exposure, if causality has been established. Population Attributable Risk % (PAR) The excess risk in the total population attributable to the exposure if a causal relationship is assumed.
What is the difference between the 3 types of absolute difference measures?
Attributable Risk in Exposed aka (Excess Risk) Percent Attributable Risk in Exposed Population Attributable Risk % (PAR) : this one is POPULATION SPECIFIC!
Types of Measures of Association
Based on Relative Differences (Ratio Measures) Cumulative Incidence Ratio (CIR, Risk Ratio) Incidence Density Ratio (IDR, Rate Ratio) Odds Ratio (OR) Based on Absolute Differences/attributable risk Attributable risk in the exposed (AR E, % AR E ) Population attributable risk (PAR, % PAR)
Describe the Key issues in Epidemiologic research:
Bias, sampling issues, study design type used, measuring disease frequency, controlling for endogeneity, interpretation of results
Cohorts can be...
Birth cohort: individuals born in the same period (often year) Exposure cohort: individuals sharing a common exposure (often an occupational exposure such as asbestos, etc.)
How do we go about quantifying the distribution and determinants of disease?
By calculating measures of disease frequency and association
What is the Incidence Density Method equation?
CI = incidence rate * specified time period There is another equation
Calculation for incidence density (ID)
CI exposed/ CI unexposed
What are and when should you use Measures of association?
CIR (Cohort) IDR (Cohort w/ Person Time) Odds Ratio (Case-Control)
Difference between Clinical Medicine and Public Health?
CLlnical medicine benefits one person versus Public health is population based and feeds both clinical research and health science
What is the association between taking anti-malarial pills (E) and the development of malaria (D) among Peace Corps volunteers in Kenya, 1997? (1 year of follow-up, no losses). Using the table below, What should you calculate? Provide Interpretation. Diseased Not Diseased 60 A 940 B Exposed Not exposed 10 C 990 D
Calculate Cumulative Incidence Ration (CIR) by the simple method [A/ (a+b) / C / (C+D) ] [60/ (60+940) / 10 / (10+990) ]= [60/1000] /[10/1000] = 60/10=6 Interpretation: "Those not taking anti-malarial pills are 6 times more likely to develop malaria compared to those who do take anti malarial pills."
Measures of Association
Calculated from analytic studies Reflect the strength of the statistical relationship between exposure status and disease occurrence (example: relative risk, odds ratio) Reflect the extra number of cases attributable to or prevented by the exposure in a particular population during a given time. (example: attributable risk) Compares the disease frequency between 2 or more groups at different exposure levels Primary measures: CIR, IDR, OR, AR, %AR, PAR.
Measures of Disease Frequency
Calculated from descriptive studies Used to enumerate the occurrence of disease (or whatever your outcome is) in a specified population in a specified period of time. The frequency of disease can be measured for either incident or prevalent cases The frequency can be expressed as either a count (an absolute measure) or as a relative measure Primary measures: cumulative incidence, incidence density, prevalence rate, prevalence proportion.
Odds ratio (OR)
Calculates odds of exposure or of developing disease compared to the unexposed or diseased Used typically to approximate CIR/OIR in case-control studies when disease is rare in the pop. Odds Ratio Equation: AD/BC OR = RR x 'built in bias' OR= 1- RR Therefore, the OR can be calculated from the RR
2 types of Case-control study within a cohort
Case cohort study: controls are selected from the baseline cohort. Nested case-control study: controls selected at time when each case occurs (incidence density sampling) aka as the cases occur longitudinal
Subject Selection: for case-control studies
Cases must be representative of all cases in the source population - the same ones who would be considered cases if a cohort study was done. Controls selected so that their exposure distribution reflects the exposure distribution among the person time in the source population, [i.e. the same "source cohort (population)"] as the cases. Both cases and controls must be selected independent of exposure status
Weaknesses of Case-Control Design
Causal inference less clear (temporal ambiguity) Often cannot estimate the frequency of disease in a population Insufficient for studying rare exposures Particularly susceptible to both selection and information biases
What are the assumptions of Kaplan Mier Method?
Censored patients are considered to have survival prospects similar to those who remain in the study. The cause of drop out can not be related to the study (called non-informative censoring).
Enternal comparison groups: non exposed in cohort studies
Chosen from a different source population often general population controls from area are use must be susceptible to the same selective factors as the E group less costly if data already available
Selection of Cases
Clearly define the source population Establish strict diagnostic criteria for case definition, independent of exposure (cases really cases) Either incident or prevalent cases, but incident are ideal Can be selected cross-sectionally (at a point in time) or longitudinally - longitudinally is a better choice Can use all cases within the population or a sample from the population
Is this study below well designed? What type of study is it? Is Mold Exposure a Risk Factor for Asthma? A remarkably consistent association between home dampness and respiratory symptoms and asthma has been observed in a large number of studies conducted across many geographic regions. In a recent review of 61 studies, it was concluded that dampness was a significant risk factor for airway effects such as cough, wheeze, and asthma, with odds ratios ranging from 1.4 to 2.2. Positive associations have been shown in infants, children, and adults, and some evidence for dose-response relations has also been demonstrated.
Cohort Study No, because Dampness does NOT = Mold. It is just one risk factor that can lead to mold exposure. The study is not actually testing for mold itself
Which study design goes with which? Cohort Case-Control Cross-Sectional Descriptive or Analytic (assign to above)
Cohort= Analytic or Descriptive Case-Control= Analytic Cross-Sectional= Descriptive
Types of Cohort Studies
Concurrent / prospective Nonconcurrent / retrospective / historical Mixed Design All 3 types have the SAME DESIGN
Case-control Studies within a Cohort
Controls may be selected from the baseline cohort, i.e. "case-cohort" design. Controls may be selected from individuals at risk at time each case occurs, i.e. "nested case-control" design. Likelihood of selection bias diminished with either approach compared to case-based study design. Case-control study conducted within the framework of existing, defined cohort, which becomes the source population Cases are selected from the cohort (all or a sample) as they develop Controls are selected by random sample of the total cohort at baseline Controls have potential to become a case OR ~ CIR, no rarity assumption needed Selection bias is reduced due to control selection within the source population
Types of Outcome Measures
Cross-Sectional or Longitudinal Continuous or Categorical Prevalence or Incidence
Instead of relative risk, what terms should you use?
Cumulative Incidence Ratio (CIR) (Risk Ratio), Incidence Density Ratio (IDR) (Rate Ratio), Odds Ratio
How many types of ratio measures (relative risk) are there? What are they?
Cumulative Incidence Ratio (CIR): (Risk Ratio) Incidence Density Ratio (IDR) : (Rate Ratio) Odds Ratio
Preventive Fraction
Cumulative Incidence of unexposed - Cumulative Incidence of exposed / Cumulative Incidence of unexposed
The probability of an event in a specified time period or interval is...
Cumulative incidence
Exposure: Important Issues are and applies to what study designs?
Definition of exposure: intensity, duration Induction period: duration of time that it takes from exposure to onset of disease Latency Period: duration of time from disease initiation to disease detection Changing exposures Allocation of person-time ( especially given changes in time) Categorizing exposure
Describe the relationship between density and rates.
Density is based on individual data Rate is based on aggregate data Both are the SAME if there is a state population with the same disease occurrence over time
External Validity Issues and Considerations for Cohort Studies
Depends on eligibility criteria for inclusion Initial response of the sample Stability of the cohort during follow-up. Requires variability of exposure and outcome levels Susceptibility of the population to the risk factor Population risk of the outcome of interest
What is the main difference between odds and an odds ratio?
Depends which type of DISEASE FREQUENCY (DF) or DISEASE ASSOCIATION (DA) you want to measure Do you want to know the PROBABILITY (DF) or RATE (DA) of Disease?!
Study Type: Ecologic study. Interpretation: Colon cancer is associated with higher meat consumption.
Describe the study below. Provide Interpretation
Study Type: Ecologic study. Interpretation: Prostate cancer is associated with higher sugar consumption.
Describe the study below. Provide Interpretation of the image below
Primary aims of Epidemiologic Research?
Describe, Explain & Predict the occurrence of disease in pop. Describe (what's going on aka abstract) Explain (why does it matter? What are the risk factors) and Predict (what do you foresee? What is the recommendation of this study?) the occurrence of disease in the population
Primary purposes of a cohort study
Descriptive (measures of frequency) To describe the incidence of an outcome over time or to describe the natural history of disease Analytic (measures of association) To analyze the relation between occurrence of outcomes and risk factors (or predictive factors)
*Distinguishing* Features of: COHORT STUDY
Directionality: forward, incident cases People are recruited into the study based on being non-disease at baseline
Used to enumerate the occurrence of disease (or whatever your outcome is) in a specified population in a specified period of time.
Disease frequency
What are some of the characteristics of the Life Table Method?
Does not require complete follow-up, i.e. it handles censoring. Does not require the exact times at which events and withdrawals occurred. Follow-up period of interest must be split into successive intervals (these don't have to be equal). Intervals should be defined so that the rate of events & withdrawals is uniform within it.
What are the assumptions of Incidence Density Method ?
Each age-specific ID is constant over the entire age interval for which that rate is estimated. Each age-specific ID is constant over calendar time (no secular trends) during relevant time frame. The population is closed. The event under study is inevitable (with no competing risks). The number of events at each event time is a small portion of the number at risk.
What type of study design is this? Homicide on the Job: Workplace and Community Determinants. Purpose: To identify workplace-level predictors of homicide risk-Design: Case-control study of worker killings in N.C. 1994-1998-Unit of analysis: Workplace•Case workplaces - those where a worker was killed during the study period•Control workplaces - density sample of N.C. workplaces, matched on time and industry sector-Data Collection: risk and protective factors assessed in telephone interviews with workplace managers-Data Analysis: O.R. estimated by logistic regression
Ecologic CASE CONTROl Study Remember, ECOLOGIC Studies MUST be Case-control, cohort or cross-sectional
Disadvantages of Ecologic Studies
Ecologic bias Temporal ambiguity uncertainty about exposure preceding disease Collinearity covariates more highly correlated at group level Confounder control Adjusting for confounders may increase bias
The inability of the ecologic inference to accurately estimate the biologic effect at the individual level.
Ecological Fallacy
Define epidemiology & its purpose:
Epidemiology is the study of the distribution and determinants of disease (or other health related states or events) in a specified population over a specific period of time
DESCRIPTIVE studies DESCRIPTIVE studies [WHO, WHAT, WHEN, WHERE] are Used to:
Establish prevalence AKA disease frequency Describe & evaluate trends in the distribution of health & disease Generates hypotheses, calculates measures of disease frequency The basis for planning health services & prevention programming Compare subgroups
Population Based Cohort Studies Advantages
Estimation of distribution and prevalence of relevant exposure variables Calculation of risk factor trends over time Strong internal validity Strong external validity (primary justification)
What does etiologic mean?
Etiology: identifying the factors that lead to the development of disease
What are administrative losses?
Ex. Follow-up period ends before 3 years of individual follow-up occurs
Ecologic study example
Example of what type of study?
Population Based Cohort Studies Disadvantages
Expensive $$$, logistically complicated Often associated with large loss to follow-up If exposure is rare, inefficient
Which is stronger: Experimental studies or observational studies?
Experimental because you can control for unknown confounders
What is the primary *distinction* between observational and experimental designs?
Experimental: can control for exposure [distinguishing factor] & is easier Observational: is more prone to bias
Selecting subjects in a cohort study are affected by...
External validity. It increases when you have study participants that are eligible, from the source, and then target pop. It decreases when you do the opposite way (Lec 6/ slide 28)
True or False: You can project the results from one population to another if the frequency of exposure differs between the population.
FALSE you cannot Pop AR% approximates AR exposed when the Projected exposure is high
Non-ecologic study cannot include ecologic level variables
FALSE! Non-ecologic study CAN include ecologic level variables
True or False: A Study population is chosen based on exposure status for case control studies
FALSE! A Study pop. is chosen INDEPENDENTLY of exposure status
Incidence Density assumes fluctuation in rates do NOT occur over time and therefore does not represent an average rate.
FALSE, it does assume a steady state of disease occurrence over time and therefore represents and average rate
ECOLOGIC Studies can be Case control, cohort or cross-sectional
FALSE, they must be!!! They must fall under one of those study designs. An ecologic study only means that each variable is ecologic.
Descriptive Studies are to analytic studies:
Form the hypotheses that are then tested in analytic studies!
Uses of Ecologic Studies
Generating individual-level etiologic hypotheses Appropriate when broad social or cultural factors are of interest Alternative to collecting sensitive or expensive data from individuals Testing impact of group-wide interventions
What Type of Unit Of Analysis is this??
Group, countries
How is the Kaplan Mier Method SIMILAR to the Life method?
Handles loss to follow-up Assumes independence between withdrawal and survival Assumes lack of secular trends
Limitations/Design Issues/Negatives [-] about OBSERVATIONAL studies?
How the population is observed & the manner in which they are observed Data collection method(s) Study Participant selection Types of Outcome measure
Study Identification STEPS:
Identify the exposure groups (exposed & non exposed) What is the exposure? Identify the disease groups (diseased & non diseased) What is the disease? Is it categorical/Continuous? What is the study design? In what manner/way were the subjects/groups selected? Any {distinguishing factors}?
Odds Ratio (OR) and Relative Risk (RR) interpretations
If RR/OR = 1 There is no difference between groups No association If RR/OR > 1 Risk/Rate/Odds > in the exposed vs non-exposed Positive association b/t risk factor and outcome It could be a causal factor If RR/OR < 1 Risk/Rate/Odds < in the exposed vs non-exposed Negative association b/t risk factor and outcome It could be a protective factor
Case Cohort Design Bonus
If the baseline cohort sample is representative of the source population, risk factor distributions and prevalence rates needed for population attributable risk estimates can be obtained. Can use the same controls to study multiple outcomes (with different cases identified)
Since we cannot measure the instantaneous rate, we estimate an average rate for a given period. This is known as:
Incidence density (Rate)
Cannot calculate CIR because we do not have the population at risk for CIR which is
Incidence/ PAR
Incidence Mortality Rate: correct or incorrect and why?
Incorrect; its technically a risk. This is the proportion of disease in which mortality occurs its NOT a rate (refer to definition)
Incidence density Assumptions
Independence between censoring and survival lack of secular trends Risk of disease is constant over study period ID fluctuates over time therefore provides an average rate Aggregate the data if withdrawals and events occur uniformly
What Type of Unit Of Analysis is this?
Individuals
Units of Analysis can be....
Individuals or Groups!
Identify the issue with the example below: If ID = 5 per 100,000 per year and sample size calculations say you need 100 cases, given an initial cohort of 40,000 subjects follow-up would have to continue for 50 years
Inefficient for studying rare diseases
Challenges with the Selection of Controls
Is critical and can be difficult Controls must come from the same source population that gives rise to the cases Controls must have the same exposure distribution as in the source population of the cases Chosen independent of exposure status, i.e. the same sampling rate for exposed and unexposed controls If sample size is large enough, problems due to sampling error are avoided
What is the Incidence Density Method?
Is the relationship between risk and rate Specifically, it approximates the risk (CIR) using an incidence rate, where the rate is constant over a period of time
What is NHANES?
It Combines biological data w/ physical data. (Surveys with physica examinations). This makes it unique. Also used to find and establish prevalence. A program of studies designed to assess the health and nutritional status of adults and children in the United States. The survey is unique in that it combines interviews and physical examinations." Began in the 1960s as a series of surveys focusing on different population groups or health topics. 1999, NHANES survey became a continuous program. It has a changing focus on a variety of health and nutrition measurements to meet emerging needs. The survey examines a nationally representative sample of ~ 5,000 persons each year. These persons are located in counties across the country, 15 of which are visited each year.
How do you determine which time scales to use for stratification of person-time
It depends on your research question It is also based on the characteristics of the disease/event you are studying Examples: calendar time, follow up time, age, birth cohort
Which Method do you use to calculate cumulative incidence ratio (CIR)?
It depends on your research question AND the data you have available/using
What does the Kaplain Mier Method estimate?
It estimates survival (just like life tables) by plotting survival time and then depicting it as a survival curve (usually going dowards)
What type of study is this? Positives/Negatives? Have smoking's risk been underestimated by US cohort studies that exclude the poor. Presented at APHA, November 2000.
It is a cohort study. Not good because of the data source
Why is the latency period important
It is relevant when considering covariates such as detection bias
Ecologic study
Joint distribution of exposure and disease is unknown
True or False: The proportion of disease is a rate.
LIES! A proportion is the probability of disease in a population vs. the rate at which change or disease occurs over time.
Strengths of Case-Control Design
Less expensive and time consuming than cohort design Good for studying the etiology of rare diseases Good for studying diseases with long induction and/or latency periods Possible to study many different exposures with respect to outcome of interest
Loss to follow up should be....
Less than 20%
The odds approximates (~) the proportion when the proportion of incidence or prevalence is...
Less than <0.1
Nested Case-Control Study Example
Levels of Maternal Serum AFP in Pregnant Women and Subsequent Breast Cancer Risk Univ. of Ca. Berkeley Child Health & Development Studies (CHDS) 1959-1994 Cohort of 12,552 pregnant women Follow-up conducted by using license records from the department of motor vehicles, and review of death certificates Cases women in the CHDS cohort who developed breast cancer, identified through the California Cancer Registry Controls were members of the cohort who had not been diagnosed at that point in time with breast cancer Exposure assessment: Frozen sera accrued between 1959-1966 Data analysis: logistic regression
Methods to measure CIR
Life table Method/Approach Simple cumulative method
Advantages of Ecologic Studies
Low cost and convenience Measurement limitations of individual-level studies Design limitations of individual-level studies; i.e. limited variability in exposure Interested in ecologic effects
Population Subgroups: special/exposed groups of Cohort Studies
May ensure the cohort has exposure of interest Less likely to be lost to follow-up because of lower mobility (military, occupational cohort) May have relevant information on exposure & potential confounders in the medical and employment records Generally, smaller sample sized needed Reduced ability to generalize results to the general population Access to data may be limited
What are the 2 MAIN types of Epidemiologic Measures?
Measures of Disease frequency and Measures of Association
What are the health and nutritional source used to capture the health status of Americans?
NHANES
Does odds ratio approximates the relative risk well?
NO!
Point Prevalence or Period Prevalence?
NONE, its prevalence (keep it simple!)
In what case can hypertension be a case control study?
Never, because it's not rare
Is there follow-up in a case-control study?
No
Simple cumulative method
No (or very few) losses to follow-up (for any reason including study termination or loss from competing outcomes). Generally, the period of risk is short (but this isn't mandatory). Assumes a closed population: All individuals included in the calculation are followed up for the entire period of study. CIR Simple Cumulative Method Equation: # of NEW cases/ total population @ risk
Is cumulative incidence ratio (CIR) realistic? Why or Why Not?
No it is not realistic because we cannot calculate risk in most real world situations due to administrative follow up
How is the Kaplan Mier Method DIFFERENT to the Life method?
No need to categorize follow-up time into intervals; intervals are based on the exact time when the event occurs Risk is estimated for the follow-up time corresponding to each case occurrence Primary distinction: no assumption about uniformity of withdrawals
An attack rate is....
Not a rate! It's a risk
Effect Estimation can be done .....
Not able to calculate rate or risks directly due to lack of information Regress group-specific disease rates on group specific exposure rates Linear, logistic, multi-level analysis (contextual analysis, hierarchical regression), etc.
What should you calculate for this? Please show calculations & Interpretation. If the mortality rate after an initial MI is higher in the 1st year of follow-up than in later years, among 1000 people followed for 1 year and 200 people followed for 5 years, which group will have a higher ID?
Nothing, because time is not a covariate. Example of Incidence Density
If, RR = 6.0 and CI UE = .0030 and CI E = .0180, then OR equals what? Show your work.
OR = 6.0 x [(1-.0030) / (1-.0180)] = 6.09
When the The disease is rare in the population (~ ≤ 0.10) or Controls are selected to represent the same source population that gives rise to the cases, not just the non-cases then....
OR approximates the IDR/CIR
Measure of Association for a Case-Control study: Odds Ratio
OR dis = OR exp There is a built-in bias away from the null The odds ratio (OR) can approximate the Relative risk (RR) in specific situations The odds ratio does not EQUAL the relative risk. The OR ESTIMATES the relative risk.
What are the simple methods for measures of association and frequency?
OR, risk ratio, relative risk
An example of Categorizing exposures in cohort studies
Occupational study where exposure is categorized according to duration of employment in a particular job. Highest exposure category is at least 20 years of employment. If a worker has been employed 30 years, it is a mistake to assign that employee to the 20+ years of employment from the beginning because they only reached that exposure in the last 10 years. That's the time frame relevant to 20+ years of exposure.
Which disease frequency isn't commonly used in epidemiology?
Odds
What can be used to approximate the risk?
Odds ratio
Categorizing exposures in cohort studies
Often there is no guidance on appropriate categories of exposure. Or the line between exposed and unexposed is not defined No problem if your data are continuous If you want to calculate rates directly, you must observe populations within categories of exposure. Common error: apportioning to PT time, the unexposed time of a person who eventually becomes exposed.
What are the differences between Simple cumulative method, Life table method, kaplan mier method, & Incidence density?
Only incidence density calculates a rate
EPIc DIAGRAM:
PAR→ TIME TIME TIME TIME TIME → D Disease/Non-Diseased E Exposed/Non-exposed
Given the following data, what is percent of the total risk of CHD in the population attributable to being a current smoker? Using the table below, What should you calculate? Provide Interpretation. Risk of CHD ExposeD= 0.321 Unexposed= 0.162 TOTAL POP= 0.197
POPULATION attributable RISK Interpretation: "17.8% of the total risk of CHD in the population is attributable to being a current smoker."
If a cohort study what changes in the EPIC diagram?
Participants at risk (PAR) are non-diseased @baseline!
Describe the relationship between power, validity, bias & loss to follow up
Power is increased based on the sample size Validity decreases as loss to follow up increases Bias increases as loss to follow up increases
What measure is best used in cross-sectional studies or at baseline in cohort studies?
Prevalence
Types of Data Sources
Primary (collected by researchers) & secondary (existing dataset)
MATCHING on Case control studies Purpose (+): Main Benefit: Negative (-):
Purpose (+): to reduce confounding Main Benefit: to gain in efficiency (greater than on validity) Negative (-): it creates a sample of controls that is not representative of exposure in the population or the population as a whole. The effect of the matching variable can no longer be studied directly, and the exposure frequency in the control sample will be shifted towards that of the cases Matching in case-control studies also does not completely control for the variable or variables used for matching, in general. This means that researchers who implement matched designs must perform matched or stratified analyses If an unmatched analysis is performed on matched data, the validity of the case-control comparison may decrease
Is epidemiologic qualitative or quantitative>?
Quantitative
Features of epidemiologic research?
Quantitative, probabilistic, comparative (not always, hypothesis driving research with a study protocol
Which is considered stronger between observational and experimental designs and why?
RCT's, they are the gold standard. Why? Because they use randomization which allows for randomizing the unknowns
Induction period Example
Radiation exposure & Leukemia are roughly 3.5 years. Exposure is classified as high, medium and low based on the amount of time working in a job where you are exposed to radiation. Person-time at risk does not equal time during which the exposure occurred. Why? because it takes some time for the exposure to lead to leukemia.
The ID will always be larger for which calculation?
Rate Incidence Density (Rate)
Why does incomplete follow up happen?
Refusal to continue participation Dying from causes other than the outcome under study (called: competing outcome. For example getting hit by a bus) Migration Administrative losses
What is the difference between relative differences and absolute differences?
Relative differences are used most often when evaluating the determinants of disease because they represent the magnitude of the association. This information is critical in the determination of causality. Once causality is assumed, absolute differences are more important measures from the perspective of public health administration and policy.
Measures of Incidence:
Risk & Rate
Basic CIR Formula:
Risk= I/PAR # of new events in a defined population over a specified period of time (incidence) / Population at risk for the event over the specified period of time (PAR)
Draw the diagram illustrating the relationships between determinants/distribution of disease, ratio, prevalence, Ci/ID/Odds, Relative risk, Rates, incidence, analytic/descriptive studies, absolute risk, ratio
See diagram
Fill in the blanks for when you would us the following Methods Simple Cumulative Life Table : Classical or Kaplan-Meier Incidence Density Method of calculating risk
See image
Primary design concerns with the case-control design
Selection Bias - can occur when cases and controls are not selected from the same source population When selective survival occurs Information Bias - can occur when there is bias in the measurement of exposure resulting in misclassification since exposure is ascertained after disease has occurred.
Case-Based (Traditional) Sampling: Issues
Selection bias may occur when cases and non-cases are not selected from the same source population, or populations with similar relevant characteristics. Selection bias may occur if losses that happen before the study groups are selected affect their comparability.
"It is difficult or impossible to select unbiased controls without a well defined source population", is an example of what?
Selection of Controls
What are the 4 Ways you can calculate cumulative incidence?
Simple cumulative method, Life table method, kaplan mier method, & Incidence density
What is the strength or weaknesses of this population? To determine the national incidence of diabetes in children by studying a group representing all parts of the country: the dependent children of U.S. Military personnel.
Strength= the population goes to the hospital and therefore will capture the majority of cases of children with type 1 diabetes Limitation= military children are not representative of general population b/c they have access to healthcare
Cohort Studies Advantages
Study many different outcomes with exposure of interest Temporal relation not in doubt, therefore, preferred for causal inference Less prone to selection bias as D status does not influence selection of subjects with respect to exposure Repeated exposure data can be collected Provides information on changing risks over time Modification of risks by increasing age Quality control can be implemented and designed into data collection and followup
What type of study is this? Positives/Negatives? Source of smoking risk data - based on nurses, Cancer Society or MRFIT volunteers, HMO members.
Study population is Not representative of general population
Advantages of Cross-Sectional Studies
Surveys are useful to administers in planning health facilities, services and programs. Good for hypothesis generating
Types of populations:
Target, source, & eligible population Study participants
Absolute risk
Tells us about the amount of disease that can be attributed to the population.
Relative risk
Tells us about the strength of association is NOT a specific enough term! It means either the risk or rate ratio. must say specifically what you calculated; don't say "this is the relative risk"
Life Course Model shows us... What is the life course model?
That we have sensitive periods of life where exposure may vary. Exposures can occur at a particular period of time in a particular population (& differently in another) Think white health outcomes (generally healthier/better) vs. minority health outcomes. It shows us how physical and social experiences impact health
The 2 method of calculations used to calculate cumulative incidence depend on...
The Features of the cohort Your research study question
What is the difference between the Odds Ratio and Relative Risk for a cohort study w/ count data?
The OR is best used to approximate CIR or IDR in case Control Studies when disease is RARE Used typically to approximate CIR/OIR in case-control studies Why? Because for many it's easier to interpret Because it is impossible to calculate the RR with certain study designs (case-control) It's easy to adjust an OR for confounding and can be derived from modeling (logistic regression) OR (event) is the exact reciprocal of the OR (nonevent) Which is not the case with the RR The OR is easier to interpret within statistical programs
Odds of disease is equal to....
The Odds of exposure
If you had to describe: "The annual rate and risk of heart disease per 1,000" What would you say?
The annual risk of heart disease per 1,000 people is.... The annual rate of heart disease per 1,000 person years is...
Example of a Case Cohort Study
The cases are occurring and are being selected as the cases occur overtime LONGITUDINALLY. Controls are being selected at baseline They are counted twice: As a case and as a control this is controlled for in an analysis
Advantages of Case-Control Studies within a Cohort
The estimated exposure odds ratio is a statistically unbiased estimate of the relative risk since cases are included in the sampling frame for the selection of controls. Exposures are often assessed before disease occurs Efficient when need additional information (particularly detailed exposure information) that is not available for the entire cohort. its very expensive $$$ to do it for the entire cohort
Sampling Frame in a cohort
The extent to which a ______ sample is representative of the total reference population depends on the completeness of the population frame available to sample from as well as participation rates.
Cohort Study
The observation of a cohort(s) over time to measure a stated outcome.
What are the assumptions of the Life Table Method?
The rate of disease (after loss) of withdrawals is the same as the rate for those who remain under observation. Disease and withdrawals occur midway through the interval, on average. Participants survive at risk for entire study period, i.e. risk is conditional on survival. No secular trend in risk.
The Source Population is:
The source of subjects for a particular study Defined by the participant selection methods of your study.
What is the most important relationship between a study factor and an outcome? (Strength of ratio measure)
The strength of the association between a study factor and outcome is the most important thing used to assess causality
trengths of Ratio measures
The strength of the association between a study factor and outcome is the most important thing used to assess causality all ratios compare to null value of 1.0 Comparisons can be made b/c: OR approximates the CIR/IDR IDR approximates CIR
the strengths and weaknesses of case-control studies
They are efficient for rare diseases or diseases with a long latency period between exposure and disease manifestation. They are less costly and less time-consuming; they are advantageous when exposure data is expensive or hard to obtain. They are advantageous when studying dynamic populations in which follow-up is difficult. Disadvantages: They are subject to selection bias. They are inefficient for rare exposures. Information on exposure is subject to observation bias. They generally do not allow calculation of incidence (absolute risk).
Are the factors that influence incident and prevalent diseases the same or different? Why?:
They differ! Factors that influence incident disease: SES, gender, race, geographical location Factors influence prevalent disease: will be quality of treatments, comorbidities, access to treatments
What do Life Tables Look like? What do they Measure? How are they Used?
They estimate the risk of the event over time They measure the distribution of events within a specific time interval The distribution of time is divided into very specific time intervals They are used when there is: loss to follow up occurs An extended period of risk When exact times of disease occurrence and withdrawal are unknown When the rate of disease varies over the follow-up period When you are interested in knowing interval-based risk (*most important one*)
Example of Mixed design
This is a Type of classification for ecologic study designs. Interested in Time & location
Example of multiple group design
This is a Type of classification for ecologic study designs. Meat Consumption and Colon Cancer: A Multi-National study
What type of study is this? Positives/Negatives? Objectives: Examine associations between prenatal manganese concentrations and placental transfer of manganese with neurodevelopment in 224 2-year-old children residing near the Tar Creek Superfund Site. Subjects were participants in a prospective birth cohort study of biologic markers of fetal and early childhood exposure to metals, maternal psychosocial stress, and their impact on neurodevelopment. This research was conducted in the area of the Tar Creek Superfund site in Ottawa County, Oklahoma. This Superfund site, a former lead and zinc mining area, contains numerous piles of mine waste enriched in metals that are dispersed throughout the region. Pregnant women were recruited during prenatal visits or at delivery from the Integris Baptist Medical Center in Miami, Oklahoma. Mothers and offspring were followed until children were 7 y of age. Eligibility criteria included a) giving birth at Integris Hospital; b) intention to live within the study area for the next 2 y; c) not being currently enrolled in the study with another child; and d) having English-language proficiency sufficient to participate in the informed consent process. Cite: Henn et al. Maternal and Cord Blood Manganese Concentrations and Early Childhood Neurodevelopment among Residents near a Mining-Impacted Superfund Site.
This is a cohort study using special/exposed groups. How do you know? The subjects were not selected based one exposure
Loss to Follow-up
Threatens the validity (internal and external) of the study. Can also lead to Type II error, a loss of power. Includes: Dropout and attrition due to missing data Other considerations: Pre-inclusion dropout (those who don't enroll).
True or False Any prospective study with loss to follow up is vulnerable to bias regardless of the method used to calculate disease frequency
True
True or False: Odds Ratio is biased away from the null ( in both directions)
True
*Distinguishing feature or characteristics* of Cross-Sectional Studies
Typically random sampling from a single target population Disease status is ascertained at that time (so the frequency measure = prevalence, the measure of association is the prevalence OR) If exposure is measured, it can be at that point in time or in the past No follow-up
Case-Based (Traditional): Cumulative Sampling
Typically, cases identified as diagnosed during study period from a stated source population Controls (non-cases) identified from the same source population from among the non-cases at the end of the study period (cumulative sampling). Exposure to the risk factor of interest is measured/gathered They are NOT representing that source population (even though they are coming from there). They are representing a subset of the non-diseased. Used when Disease is RARE. Selecting controls from those disease-free at the end of the observation period during which cases are identified.
What matters in analyzing a study is your...
Unit of analysis!! and probably your research question (b/c that always matters!!!)
Is the rate/risk of disease among the exposed different than that among the non-exposed? If yes, in what direction and by how much?
Used to answer the same research question as in cohort studies:
How do we quantify the determinants of data?
Using measures of association
How do we quantify the distribution of data?
Using measures of frequency
How representative is this example below? List Positives/negatives. Study Begun in 1976 ~ 270 published papers Original study question = oral contraceptives and breast cancer Many other areas researched Cohort Married registered nurses Between ages 30-55 (in 1976) Living in 11 most populous states whose nursing boards agreed to supply members' names ~ 122,000 out of 170,000 responded Follow-up to date: 90%
Very, its a Great Cohort! Problem: confounding Note: SES does not matter b/c this is a biologic issue
True or False: We can measure the incidence of cancer, heart attack, depression and/or dementia or asthma
We canNOT measure the incidence of cancer, depression or asthma because we do not know when they initiated or began To measure, we usually define by stages: Ex. Cancer( stage 1, stage 2,etc.) identifying at which stage cancer occurs we cannot identify prevalence for work for all types of cancer We CAN measure heart attacks because we know when they occur
True/False: We are 95% confident we can predict disease in individuals?
We canNOT predict disease in PEOPLE
Stratifying factor: age exposed: cholesterol disease/outcome: coronary heart disease (CHD) 2 different measures of frequency: incidence rate and rate ratio interpret incidence rate first "At every cholesterol level, those with older people, double or triple the frequency of CHD. Also at every cholesterol level, the rate of CHD increases for both young and old, with a marked increase & blah blah blah" Confidence intervals and p-values are missing so we do not know if its statistically different The annual rate of heart disease per 1,000 person years (vs if it were risk it would just say per 1,000 people) The incidence rate represents the frequency The rate ratio represents the measures of association
What is the exposure, outcome or disease of interest, the stratifying factor, measure of association, measure of disease frequency in the image below? Interpret the image.
A Life table Calendar Time
What is this an example of ? & What are you following?
See google doc notes
What is this an example of? Answer the questions below: EXAMPLE: A patient follow-up is broken into 1 year intervals for the calculation of probabilities of death and survival. Solve the following based on the chart: What is the risk of dying in interval 2? What is the probability of surviving interval 2? What is the risk of death during the 3 years post MI? What is the probability of survival during the 2 years post MI?
Ecologic study with population not individuals as unit of analysis
What type of study is in the image below?
This is a time trend design which belongs to the classifications of ecologic studies. It demonstrates to us that there are particular causes that contribute to the disease of interest (anaphylaxis) compared to others over time. Specifically, the proportion attributable to other causes has increased significantly over the years.
What type of study is this? What do you notice? Provide Interpretation
The attributable risk identifies the amount of the exposure that we are able to prevent only if we have causality established.
Whatt does this graph show you?
Incidence Density Sampling
When a case occurs, a control (non-case) is selected (controls selected longitudinally) "Matches" control to case based on time Controls have the potential to later become a case Ensures that controls represent the source population from which cases come Rare disease assumption not needed, OR ~ IDR/CIR for both common and rare diseases using this strategy
Study population → Sources of Data → Unit of Analysis
When selecting a study population, what should they address? How are they observed? Who do they represent?
Can you use prevalence in a case-control study? Yes/No and for what?
Yes but not to identify or answer questions about etiology
Can you stratify person-time covariates? Name the assumptions if any.
Yes, ASSUMING there are: No crossover effects That if exposure status changes, so does the risk TIME is not a covariate
Are EPI studies descriptive?
Yes, Always
How representative is this example below? List Positives/negatives. Oral contraceptives, diet and lifestyle risk factors in a younger cohort Planned to collect better exposure information (type of OC, duration, important covariates) Cohort Registered nurses, 25-42 years, Sample size = 125,000 Recruitment: send a single questionnaire. Enroll all eligible who respond 116,686 out of 517,000 enrolled Follow-up to date: 90%
Yes, it accounts for confounders by using more covariates and includes a large sample size
The risk of disease is approximately constant over the study period in incidence density calculations.
Yes, this is an assumption of incidence density
Calculation for cumulative incidence ratio (CIR)
[ a / (a+b) ] / [ c / (c+d) ]
Cross-Sectional Studies are also referred to as
a survey or a prevalence study
Risk Factor
a variable associated with an increased risk of a disease
In a nested case-control study OR...
approximates the IDR
How an ecologic approach works
attempts to provide a wide perspective on social phenomena that occur on different levels in society. It takes into account the influence of environmental factors at multiple levels (e.g., family, neighborhood, nationality) that shape individual behavior. Each social level is nested and functions within the operation of another one. Thus, the study of the etiology of aggressive discipline needs to also consider social context variables such as the characteristics of the society and the neighborhood in which the family resides.
Excess risk of taking anti-malarial pills attributed to the development of malaria among Peace Corps volunteers in Kenya, 1997. Using the table below, What should you calculate? Provide Interpretation. Diseased Not Diseased 60 A 940 B Exposed Not exposed 10 C 990 D
attributable RISK (60/1000) - (10/1000) =0.05 "5% excess risk of developing malaria among those who take anti-malarial medication versus those who do not take anti-malarial medication."
What is the percent of the total risk of developing malaria among those who take anti-malarial medication attributable to the medication? (Peace Corp Volunteers, 1997). Using the table below, What should you calculate? Provide Interpretation. Diseased Not Diseased 60 A 940 B Exposed Not exposed 10 C 990 D
attributable RISK [ [(60/1000) - (10/1000)] / (60/1000) ]* 100 =0.8333333333333*100 =0.83 "83.3% of the total risk in the group taking anti-malarial medication is attributable to taking the medication, conditional on the establishment of causality."
Case-based case-control study (traditional case control study)
cases and controls are selected at a given point in time typically at the end of "follow-up."
internal comparison groups: non exposed in cohort studies
elect from the same source population as Exposed increases the likelihood that the E and E members are from a similar subgroup of the population generally have the same follow-up procedures and therefore, the same likelihood that D will be detected
external validity
extent to which we can generalize findings to real-world settings
A cohort study can either be...
fixed or dynamic
How the cohort is selected is what differentiates...
internal or external validity
Measures of association
is an effect Determines if there is an association between a risk factor (exposed/unexposed) and an outcome
What is an assumption of loss to follow up?
it does not differ in the rate of those of disease compared to those who stay in the study "The Characteristics of the study population that we lost is equally similar to the study population of characteristics of those who remain in the study. Therefore these study findings are still relevant."
Sampling Strategies to Select Controls
minimize selection bias (covered in previous slides, and maximize the potential for the OR to approximate (~) the RR If a disease is rare, all sampling strategies will give the same result (OR ~ IDR/CIR) If disease is common, different sampling strategies will give different results
Disadvantages of Cross-Sectional Studies
not good for diseases that are rare or a short duration
Calculate an Odds ratio for the following: " A case control study of the association between aspirin consumption and the development of a stomach ulcer" & provide an interpretation: Diseased Not Diseased 150 A 76 B Exposed Not exposed 78 C 69 D
or= ad/bc =1.75 iNTERPRETATION: "People who consume aspirin have a 1.75 increased odds of developing a stomach ulcer."
How do you measure occurrence of disease? What are the 2 ways to measure the occurrence of disease?
sing Prevalence and incidence & which one we use is disease specific!
Ecologic studies have all variables at what level?
the ecological level
Risk
the proportion of new cases OR the probability of developing an event over a specified period of time
rate of change
the rate in which change or a disease occurs in time OR a measure of change in disease occurrence per unit change in time
What is survival analysis?
the study of time between entry into observation and the outcome of interest
True or False: If random sampling is not used, it is hard to justify that the measure is a representation of the target population
true, for ecologic studies
At a particular point in time, in a particular population... SPECIFIED period of time... SPECIFIED population
we can identify or explore XYZ exposures
When are cohort studies done?
when sufficient evidence has accumulated from other studies to indicate a prospective cohort study is warranted a new agent is introduced and requires monitoring for its possible association with adverse health outcomes (levaquin, vioxx) temporal association is unclear from a case-control study impressive results are obtained from a c-c study (either positive or negative) and issues of validity (selection or information bias) are evident or a concern
Limitations/Design Issues/Negatives [-] about EXPERIMENTAL studies?
$$$: expensive & Cost limitations Ethical Issues: You can't give people HIV Validity: External or Internal (ex. people that participate are generally healthier)
Cohort Studies Disadvantages
$$$Costly, resource intensive (manpower, time and money) Inefficient for studying diseases with long latency Inefficient for studying rare diseases "Study effects" Changing exposure Withdrawals/loss to follow-up Basic design allows only 1 risk factor to be studied
Calculation for cumulative incidence (CI)
# of cases/ pop @risk
Lost to follow up should be below what percentage?
20%
Ecologic studies can be what type of study design
Any: Case control, cohort or cross-sectional
Classification of Epidemiologic Studies [aka List the Types of Epidemiologic Studies]
DESCRIPTIVE studies [WHO, WHAT, WHEN, WHERE] ANALYTIC studies
How should you start planning a research study?
Develop a research question, FIRST!
ANALYTIC studies
Establish measures of association (clarifies DETERMINANTS) To test preventive hypotheses Provide suggestions for health promotion or prevention
Our class focuses on...
Hypothesis driven etiologic research!
Example of time-trend design
This is a Type of classification for ecologic study designs. Notice the separation of causes
Why would you use a hybrid study design?
To avoid bias!
A life Table with a SYNTHETIC FIXED COHORT Following time
What is this an example of? What are you following here?
Can you calculate risks from rates?
Yes
Incidence Rate (IR):
is a RATE The occurrence of cases (disease) at a certain time/point in relation to the population at risk Range: 0 to Infinity!
Calculate the odds and interpretation of the following: if the proportion of smokers in a population = 20% then...
0.20/(1-0.20)= 0.25 The odds of being a smoker is 1:4 or .25 to 1.
Types of Sampling strategies:
1) Traditional (cumulative) sampling 2) Density Sampling
3 Types of classification for ecologic study designs
1) multiple group design 2) time-trend design 3) Mixed design
Types of case control studies? (How Many)
1. Traditional Case Control (TCC) 2.Nested Case Control (NCC) 3. Case Cohort, Case-Control (CCCC)
EXPERIMENTAL studies
1clinical trials, intervention studies, community trials
What are the values of a protective odds?
<1, Less than 1
The risk needs to be below what for the Incidence Density Method?
<20%
What is Etiologic Research?
Focuses on the factors that INFLUENCE the CHANGE in disease Status
*Distinguishing* factor of case control studies?
How you select the controls
If you are interested in observing screen time and its relationship with hyperactivity, how would you measure risk?
I would investigate the risk of hyperactivity among people who are EXPOSED to screen time vs. those who are NOT EXPOSED to screen time.
ANALYTIC study types
OBSERVATIONAL studies EXPERIMENTAL studies
Used to estimate the magnitude of a health problem is...
Prevalence
Major Concerns of a Cohort Study
Selecting a cohort (sampling frame, sampling and recruitment, external vs. internal validity) Exposure assessment Follow-up
When exact time of event and censorings is known,
Then we CAN calculate incidence density
What occurs with the Incidence Density Method equation?
There is an overestimate because the formula does not take into account the number of people dying for year. Therefore, it can overestimate the true death rate
What do odds ratios and relative risk have in common?
They both compare the likelihood of an event between two groups
When exact time of event and censorings is unknown,
We cannot calculate incidence density
What do we care about? What should we use to measure that?
We care about etiology! & we should use INCIDENCE to measure etiology
What is the ecologic fallacy?
We do not know the distribution at the individual level. The interpretation may be incorrected
What assumptions are made when you use ID to estimate person time, meaning when the exact time of events and censorings are unknown?
We use the average population and assume that all events and censoring occur at the same time. Anyone who withdraws during interval contributed half the interval
Example of Case-control studies within a defined cohort: Case Cohort
We want to develop cohort of non-diseased women who don't have hypertension and take a sample of 400. the controls can become a case THe controls represent the source population and not just the the study therefore no rarity assumption is needed
It shows you that the survival rate of treatment A is higher than Treatment B and then it eventually levels off At every single time point, treatment A outperforms treatment B
What does this chart show us?
Survival is highest for cancer at stage 1 than any other stage. Cancer levels off eventually for all stages except stage 4
What does this graph shows us?
When would you use the relative risk over the odds ratio?
When disease is not rare. The OR is best when disease is rare!
If you increase sample size what happens?
You increase PRECISION!
In a case cohort study the OR....
approximates the CIR
Ecologic studies
compare group-level variables rather than individual level data
Primary measures of Disease frequency
cumulative incidence, incidence density, prevalence rate, prevalence proportion
ALWAYS READ THE ENTIRE EXAMPLE- why?
do not be mislead by the suggested interpretations or suggested study design.it may be incorrect
Example below is of what?: " An individual-level parent characteristics are important for understanding aggressive discipline. However, a more complete understanding of the etiology of aggressive discipline requires an..."
ecologic approach
Marginal frequencies of exposure and disease are known
ecologic study quality
The frequency of disease can be expressed as:
either a count (an absolute measure) or as a relative measure
The frequency of disease can be measured for:
either incident or prevalent cases
internal validity
extent to which we can draw cause-and-effect inferences from a study
Incidence
frequency of a new event in a population during a specified period of time
Prevalence
frequency of an existing event (old or new) during a given time period
Confidence Interval (CI%) Equation?
incidence/pop. @ risk = CI%
Cumulative Incidence Risk (CIR):
is a PROPORTION / used for cohort study Conditional probability: The probability of disease developing in a person w/o disease over a specific amount of time This is conditional on survival to the interval ---what does this mean? Unconditional Probability: probability of dying over any period of time Range: 0 to 1 (this is how you interpret the CIR risk)
Rare disease in case control studies should always be
less than or equal to 10% (≤ 0.10)
Prevalent cases =
people who have disease at baseline aka cohort study
When does selection occur for: cohort?
selected at baseline
Discuss primary design concerns in case-control studies
selection of cases and controls collection of accurate exposure data control of extraneous factors
If the rate is NOT Constant then....
use a generalized incidence density method (formula # 3)
