Midterm 1 Study Guide PDF

Preview

Summary

Midterm study guide...

Description

Terms: Know it Learn it Quizlet with terms: https://quizlet.com/_65w54g Epidemiology - Distribution and determinants of health and disease in human populations Core Concepts of Epidemiology - Health and disease; data collection from individuals; inferences to populations; probability and statistics Observational study - Observe risk of outcome associated with exposure in a population, most frequently used study type in Epidemiology John Snow - Controlled cholera epidemic by figuring out that the source of the water resulted in people getting sick (natural experiment) Five Major Areas of Epidemiology - Descriptive, etiological, evaluative, health services, clinical Descriptive Epidemiology - Describes health and disease and their trends over time in specific populations Etiological Epidemiology - Risk factors; searches for hazardous or beneficial factors influencing health conditions (i.e. toxic pollutants, inappropriate diet, deadly microorganisms; beneficial diets, behavioral habits to improve fitness) Evaluative Epidemiology - Program evaluation; evaluates the effects of preventative interventions; quantitatively estimates risks of specific diseases for persons exposed to hazardous factors Health Services Epidemiology - Describes and analyses the work of health services; comparative effectiveness Clinical Epidemiology - Describes the natural course of a disease in a patient population and evaluates the effects of diagnostic procedures and of treatments; diagnosis, prognosis, treatment Population of Interest - Collection of individuals defined by at least one organizing characteristic Dynamic Population - Allows movement in and out of population Stationary Population - Does not allow movement in and out of population Population-at-risk - Population free of disease Cases - People diagnosed with a disease or outcome Variable - Any measured characteristic that differs amongst individuals Prevalence - Function of incidence and duration; # of diseased persons/# of persons in the population (at a specific point in time); burden of disease in a population Incidence Rate - Makes an assumption about how much time people experience risk; # of new cases/person-time at risk Cumulative Incidence (Risk) - Probability of the event in a specified interval of time; # of new cases/# of persons in population at risk @ the beginning of the time period Probability - Proportion of events in the population at risk; ranges between 0-1 (in terms of proportion) or 0-100% (in terms of percentage) Risk Factor - Variable that increases or decreases the probability of the event

Case-Fatality - Number of people who die from a disease from the population-at-risk Health Indicators - Measures of the occurrence of infection, syndrones, symptoms, biological, or subclinical markers; binary, ordinal, continuous Binary - Takes on two values: present or absent Ordinal - Takes on multiple graded values; examples: health rating, symptom frequency, ability to perform health-related activity Continuous - Variables with continuous response options, equal intervals between categories; examples: age, weeks of pregnancy, viral load, cholesterol level, blood pressure Exposure - Any measurable variable that affects or is associated with health; types: acute, chronic, time-varying Acute - Occurs for a relatively short duration, doesn’t repeat; examples: natural disasters, motor vehicle accidents Chronic - Stable over time; examples: pollution, poverty, sex, race and ethnicity, DNA sequence Time-varying - Vary across life course of an individual; examples: diet, exercise, smoking, alcohol consumption Characteristics of Exposure - Duration of exposure, latency and critical windows, timing of exposure Critical window - when exposure occurs Latency - when the disease occurs; gap between exposure and disease Age adjusted rates - adjusted rates so that differences between comparison groups are taken into account Standardized morbidity ratio - uses weights from “exposed” group; observed/expected Necessary - variable is required for outcome to occur Sufficient - if variable is present, outcome will occur

Concepts Core Concepts ● Health and disease ○ Factors associated w/ good health ○ Disease causation and prevention ● Data collection from individuals ● Inferences to populations (rather than individuals) ○ Apply data collected from individuals to populations studied ○ Can't say anything about specific individuals w/in populations ■ Can only apply inference from population as a whole, not definitive declaration ● Probability and stats ○ Very few things are definitive ■ Probability isn't 100% ○ "it depends" ○ Correlation does not necessarily equal causation 7 Steps to Conduct an Epi Study 1. Define pop of interest

2. Conceptualize and create measures of exposures and health indicators 3. Take a sample of pop 4. Est measures of association between exposures and outcomes (health indicators of interest) 5. Evaluate whether the association measured is causal 6. Assess the evidence for causes working together (two or more exposures working together to increase risk of outcome in pop) 7. Assess the extent to which the result matter -> applicable to other pops ● First four: descriptive ● 5: causal inference ● Last two: interaction Evolution of Epi ● Hippocrates - interaction between environment and lifestyle factors ● John Graunt - analysis of mortality data by sex and disease ● William Farr - developed a more sophisticated life table ● Pierre-Charles Alexandre Louis - effect of any potentially beneficial treatment can only be assessed by a comparison of closely similar subjects receiving and not receiving it; numerical method ● John Snow - controlled cholera epidemic by figuring out that the source of the water resulted in people getting sick (natural experiment) ○ Didn't need to worry about differences in populations that could account for different rates of infection - same population was being exposed to different water sources (people right next door to each other) ● Rudolf Virchow - medicine and PH as applied social sciences, founder of cellular pathology ● Joseph Goldberger - importance of epi in non-infectious disease ● 19th century: infectious disease ● 20th century: shifted to noncommunicable diseases Epi Today ● Applied objective: prevention and treatment of disease and promotion of health ● Biomedical and social science ● Epi is to a community as a doctor is to a patient ● Observational vs experimental studies Epi Research ● Flip-flop on causal associations ○ An exposure could be a cause of an outcome in some populations and not others (reason for varying results of different studies) ○ Interactions of exposure w/ unknown variables (i.e. a gene present in some populations that isn't present in others) ○ Try to figure out what is valid for the pop we're focusing on ○ V important to have clear definition of exposure, pop, outcome - valid estimate in study ● Some that we know (i.e. cigarettes cause cancer), but a lot that we don't know ● Individual vs population

○ ○

As an individual, cigarettes will not necessarily cause cancer At the pop level, cigarette smokers have a greater probability of getting cancer than non-smokers; know that it is causal through multiple studies of multiple pops

2x2 Table Diseased? Exposed?

Yes

No

Yes

A

B

No

C

D

Total Exposed: A+B Total Unexposed: C+D Total Diseased: A+C Total Disease-Free: B+D Total Sample: N Population Eligibility criteria: ● Geographic space and time period ● Characteristics of persons, events, or exposures for which health-related factors are of interest ● Factors that promote successful study completion ○ Good responders to surveys ○ Likely to continue to attend the health facility for follow-up visits ○ Individuals who are healthy so that they are unlikely to die during the course of the study outside scope of inquiry ○ Affects generalizability - people who respond to surveys or continue to visit a health facility are likely healthier than general population ■ Have qualities others may not have ● Ex: Baby Boomers ○ Time - year of birth (1946-1964) ○ Geographic location - US ○ Dynamic population - people are moving out (dying) ■ If they had been observed in 1965 (no one died yet) then it could be considered stationary 2x2 PT (person-time) Table - for Dynamic Populations Diseased? Exposed?

Yes

Person-time

Yes

A

PT Exposed

No

C

PT Unexposed

Total diseased

PT Total

Prevalence vs. Incidence

Incidence

Prevalence

Numerator

# of new cases of disease during a specified period of time

# of existing cases of disease at a given point in time

Denominator

Population at risk

Population at risk

Focus

Whether the event is a new case Time of onset of disease

Presence or absence of disease Snapshot in time

Uses

Estimates the risk of becoming ill Used to measure both chronic and acute diseases Important in causal inference

Estimates the probability of the population being ill at the time of the “snapshot” Useful in the study of the burden of chronic diseases Implications for health services

Factors influencing Prevalence ● Anything that increases incidence or lengthens duration - increases prevalence ● Increases: ○ Longer duration of disease ○ Prolongation of life of patients w/out cure ○ Increase in new cases ■ Increase numerator ○ In-migration of cases ■ Increase numerator ○ Out-migration of healthy people ■ Decrease denominator ○ In-migration of susceptible people ○ Improved diagnostic facilities (better reporting) ● Decreases: ○ Shorter duration of disease ○ High case-fatality rate ○ Decrease in new cases ■ Decrease numerator ○ In-migration of healthy people ■ Increase denominator ○ Out-migration of cases ■ Decrease numerator ○ Improved cure rate of cases P/(1-P) = I x D - P - prevalence, I - incidence, D - duration - If disease is rare (low prevalence): P = I x D

CI = 1 - exp(-I x t) - CI - cumulative incidence, exp - natural exponential function (2.72), t - length of risk period - Works if mortality from other diseases is ignored and incidence remains constant - If disease is rare (low incidence) or risk period is short: CI = I x t Measurement example: ● Research question: are individuals who have depression more likely to be overweight than individuals w/out depression? ● Measuring depression ○ Constellation of symptoms ○ Characterized by disabling feelings of hopelessness, sadness, and loss of interest in activities ● Measuring overweight ○ BMI > 30 ● Clarity ○ Be clear about construct being measured ○ If measurements include respondent-answered questions, make sure questions are easily interpretable, concise, specific Age Adjusted Rates (Ratio Measure) ● Adjust rates so that differences between comparison groups are taken into account ● Select "weights" and apply them to each group ○ Can use any weights, so long as you use the same weights for both groups ○ Ex: weight of 0.5 ■ Normal: (0.5 x 5.2) + (0.5 x 11.4) = 8.3 ■ Overweight: (0.5 x 10.4) + (0.5 x 26.5) = 18.45 ■ Natural weights would be the percent that makes up the population ○ Ex: 20% 40-49 and 80% 50-59 ■ Normal: (0.2 x 5.2) + (0.8 x 11.4) = 10.16 ■ Overweight: (0.2 x 10.4) + (0.8 x 26.5) = 23.28 ● Select weights that correspond to the population you would like to make an inference to ○ Could pick natural weights (i.e. age distribution) ○ If you want to make an inference about the "exposed" population -> weights should come from exposed age distribution (19% and 81%) ○ If you want to make an inference about the "unexposed" population -> weights should come from unexposed age distribution ● Standardized morbidity ratio (SMR) - uses weights from "exposed" group ○ Ex: 19% 40-49 and 81% 50-59 ■ Normal: (0.19 x 5.2) + (0.81 x 11.4) = 10.222 ■ Overweight: (0.19 x 10.4) + (0.81 x 26.5) = 23.441 ■ SMR = 23.4/10.2 = 2.2941 ■ Rate of diabetes is 2.3 times greater amongst exposed group than unexposed group, adjusted for age ○ Observed/expected ■ Expected = age-specific rates of unexposed group

■ 23/[(0.0052 x (0.019 x 984)) + (0.0114 x (0.81 x 984))] = 2.3 Causal Models ● Necessary-sufficient model ○ Variable X may cause Y ○ Necessary - X is required for effect to occur ○ Sufficient - if X is present, outcome will occur ○

○ ○ ○

X is necessary

X is sufficient

1.

+

+

2.

+

-

3.

-

+

4.

-

-

Ex (1): given an unvaccinated population, the measles virus Ex (2): X - HIV, Y - AIDS ■ Most infectious diseases ■ Need to be exposed to infectious agent to get the disease, but being exposed doesn't mean you are going to get the disease ○ Ex (3): X - fall, Y - head trauma ■ If X happens, Y is guaranteed ■ Y has multiple causes/sources ○ Ex (4): X - obesity, Y - diabetes ■ Basically everything that doesn't fall into the first 3 categories ● Causal pie model ○ Implications: ■ Multicausality ■ Strength of causes ■ Interaction between causes ■ Sum of attributable fractions ■ "Natural" history of disease (ex: induction time) Koch's postulates (1882) - worked somewhat well for infectious diseases, not at all for chronic diseases ● Agent must be shown to be present in every case of disease (necessary cause) ● Agent must not be found in cases of other disease (specific cause) ● Agent must be capable of reproducing the disease in experimental animals (sufficient cause) ● Agent must be recovered from the experimental disease produced Hill's criteria of causation ● Did the exposure precede the disease? ○ Necessary for causation ● How strong is the association?

○

● ● ● ● ● ●

Use ratios of incidence rates or risks to estimate strength/magnitude of association ■ Risk ratio: risk of exposed/risk of unexposed ■ Incidence rate ratio: rate of exposed/rate of unexposed Does the association become stronger w/ increasing exposure? Is the association consistent? Is the association specific? Is the association consistent w/ other bio evidence? Has the association any analogue? Is the association coherent across different studies?

Practice Problems Quizzes Quiz 1 1. What is epidemiology? 2. List one core epi concept? 3. If A is correlated with B, then A necessarily caused the occurrence of B - True or False? 4. Epidemiology focuses on epidemics and only epidemics - True or False? 5. Epidemiology is primarily concerned w/ making inferences to individuals - True or False? 6. What is the (a) exposure and what is the (b) outcome in the following example? a. A study focused on 225 men and women who underwent bypass surgery between 1987 and 1990. They were asked about their marital status and how happy they were in those marriages. After following the study participants for 15 years after surgery, researchers from the University of Rochester in New York found that those who were married were 2.5 times more likely to be alive following bypass surgery than those who were not married. b. (c) What is the denominator (“population at risk”)? c. (d) What is the numerator? Quiz 2 1. According to a new study published in the Annals of Internal Medicine, sitting can kill you. The study found that sedentary behaviors - like sitting for eight to 12 hours or more a day - increase one's chances of getting a disease or condition that kills prematurely even if you exercise. Sedentary behavior leads to cardiovascular issues, cancer and Type 2 diabetes. According to the World Health Organization, physical inactivity has been identified as the fourth-leading risk factor for death for people all around the world. In the above example: a. What is the exposure? b. What is the outcome? c. What is the population at risk? 2. Who was the epidemiologist in London who figured out how Cholera was spread and subsequently prevented further spread of disease? John Snow

3. Experimental studies are more common than observational studies in Epidemiology? True or False? 4. Which of the following is NOT one of the 5 major areas of epi? a. Cancer epi b. Descriptive epi c. Etiological epi d. Evaluative epi e. Health services epi f. All of the above are listed as the 5 major areas of epi 5. Why is John Snow’s “Grand Experiment” often referred to as a “natural experiment”? Naturally randomized Quiz 3

1. Populations are defined by: # of individuals, eligibility criteria, characteristics, geographic area, time frame Give 3 examples that can define a population: age, exposure, an event Briefly explain the difference between a dynamic and stationary population Prevalence deals with new cases of disease - True or False? Incidence deals with current cases of disease - True or False? List two factors that can contribute to the increase AND decrease in prevalence (2 of each) - increase: in migration of cases, out migration of healthy people, longer duration of disease - decrease: shorter duration of disease, high case fatality rate from disease, decrease in new cases 7. What is the difference between a cumulative incidence and incidence rate? 8. Would you use incidence or prevalence to draw causal inferences? Explain 9. When you want to see how many people in a given population have a disease at a specific point in time, would that be a measure of prevalence or incidence? 10. If the annual incidence of Swine flu is 1% in a population of 5 million and the average duration of Swine flu in the population is 3 months, what is the point prevalence of Swine flu and the prevalent number of Swine flu cases? Assume that Swine flu is a rare disease and that incidence is constant over time

2. 3. 4. 5. 6.

Quiz 4 1. 2. 3. 4. 5.

List 2 of 3 health indicators: Give 2 examples of each of the 2 health indicators you listed List 2 of the 3 types of exposures: time varying, acute, chronic Briefly define the 2 types of exposures you listed Briefly describe the difference between exposure duration and exposure timing

Self-Assessment Quiz 1. Prevalence is increased by all of t? a. Improved cure rate of cases

b. Improved diagnostic facilities c. In-migration of cases d. All of the above are factors that increase prevalence 2. Give one example for each of the following health indicators and “make up” at least 3 response options for each example a. Ex: Binary - Male, Female b. Ordinal c. Continuous 3. What is the difference between acute and chronic exposures? Give an example of each 4. Define latent time exposure and give one plausible example of a latent time exposure variable to explain differences in exam scores among university students 5. Which exposure type (acute, chronic, or time-variable) is the most challenging to measure? Give one reason why 6. Incidence or prevalence? a. Proportion of campers who developed food-borne disease within a few days of eating chicken salad at the dining hall. b. Proportion of persons who reported having arthritis as part of the 2013 National Health Interview Survey. c. Proportion of children who have immunity to mumps, either because they had the disease or because they received the vaccine. d. Proportion of premature babies who contract infections this year in the Kapiolani Hospital NICU. e. Proportion of US 12- to 16-year-olds who have tried marijuana. f. Proportion of the US population that contracted influenza during the 2010-2011 flu season. g. Proportion of breast cancer patients who have BRCA mutations. h. Proportion of 3-6 year-olds currently diagnosed with autism spectrum disorder in the US. i. Proportion of 3-6 year-olds who received a new diagnosis of autism spectrum disorder in the US in 2012. j. Proportion of women age 15 to 25 who tested positive for HIV for the first time in South Africa in 2010. 7. To enable early discovery and treatment of cervical cancer, regular gynecological checkups were carried out on women betw...