Pray for me Epi exam 2

quick and easy *fingers crossed

Case Control study design

you start with the disease ([blanks] and controls) and do a retrospective study to find exposure

How do you find the odds ratio?

(AD/BC) also known as the odds of exposure among cases/ odds of exposure among controls.

• A = the number of times the event of interest occurred in the exposed group (or treatment group).

• B = the number of times the event of interest did not occur in the exposed group.

• C = the number of times the event of interest occurred in the unexposed group (or control group).

• D = the number of times the event of interest did not occur in the unexposed group.

Where is Odds Ratio Used?

In Case Control Studies!

Odds Ratio interpretation

Subjects with (disease) were (insert term) times more likely to have a history of (exposure) compared to controls

Cohort Studies

It starts with exposure status (exposed, unexposed) —> prospective study to see if disease develops

Percent Relative Effect

Take the rate of case subject - rate of controls = answer and then take that answer and divide by rate of case subjects and then turn to absolute value and X 100 to find the percent

What calculations are associated with Cohort Studies?

Relative Risk, attributable risk, and attributable risk percent

Relative Risk formula

[A/A+B] / [C/C+D] ~AKA incidence in the exposed / incidence in unexposed.

Find incidence rate in the exposed group (number in exposed group who got disease over total exposed) and incidence rate in unexposed groups (number in exposed group who got disease over total unexposed)

And then take exposed incidence and divide by unexposed incidence and BAM answer

Interpretation of relative risk

The risk of (disease) was (risk ratio) times greater among (exposed) compared with (unexposed)

Attributable Risk

[A/A+B] -[C/C+D] ~ incidence in expose - incidence in unexposed

Attributable Risk %

[A/A+B] -[C/C+D] / [A/A+B] X 100 ~ so basically take your attributable risk divide by incidence in exposed and multiple by 100 and BAM

What calculation is used in clinical trials

Rate ratio and percent relative effect!

Rate Ratio

(A/A+B) / (C/C+D) ~ incidnece of the outcome in the treatment group / incidence of the outcome in the control/ placebo group (blank over however many person years/ blank over however many person-years)

Percent Relative Effect

(Rate Ratio - 1) X 100

Percent Relative Effect Interpretation

Overall, experimental/treatment group had a (percentage) reduction in the incidence of (disease ) compared with control group

Types of Random Errors (non systematic)

sampling error = sample isn’t representative of population

poor precision (measurement error) = study factor not measured sharply (avoid this by increasing sample size)

Variability in Measurement

Types of Systematic Errors

Selection Bias

Information/Observation Bias

Confounding

Selection Bias: Self Selection

Participates or refuses to participate based on disease or exposure status (avoid by increasing the response rate (over 80%) and check characteristics of subjects

Selection Bias: Control Selection

Distortion of the exposure-disease association by the control group selection (this bias will either weaken or strengthen the association between exposure and disease. to AVOID select controls from the population/community where cases arose)

Selection Bias: Differential Surveillance

treating patients differently based on exposure history

Selection Bias: Loss to Follow-Up

matters when the association between risk factor and disease outcome differs among dropouts compared with study participants who complete the study (reduces power of the study and may bias overall study results)

Information/Observation Bias: Recall Bias

better recall of exposure among cases among controls

( case control mostly likely to have this bias )

Information/Observation Bias: Interviewer Bias

occurs when interviewers probe more thoroughly for an exposure in a case than in a control

Information/Observation Bias: Prevarication (lying) bias

occurs when participants have ulterior motives for answering a question and thus may underestimate or exaggerate an exposure

Confounding

stems from the natural mixing of effects between exposure and disease and a third variable called a confounder. Distorts the relationship between exposure and disease

i. Positive or negative confounders 1. Direction of confounding

1. Exaggerate true association: POSITIVE ~ if the crude rate is higher

2. Hide a true association: NEGATIVE ~ if the crude rate is lower

ii. Confounding most common in ecological studies

Ecological Studies: Strengths

Inexpensive, quick

Can use secondary data

Population level trends

Ecological Studies: Limitations

Ecological Fallacy: observations made at group level may not reflect relationship

confounding

Cross Sectional Study: Strengths

Can measure disease burden in a population sample

Quick, cheap, no follow up

formulate a hypothesis

Cross Sectional Study: Limitations

You cannot tell whether exposure or disease came first

Case Control: Strengthes

Good for rare diseases (start with disease)

Etiological hypothesis - start to see how exposure may affect disease (casuality)

Quick (1 data collection point) and there’s no follow-up

cheaper than a cohort

good for diseases with long incubation periods

Case Control: limitations

may have survivor effect (people who surived disease may have different risk factors/ exposures from people who had the worst disease and died)

Not efficient for rare exposures

more prone to bias than a cohort study (especially recall, interviewer and control selection bias )

Cohort (prospective or retrospective): Strengths

direct measure of risk (very important pointt)

rare exposures (can use occupational cohorts) -because you’re starting with subjects classified by exposure

Can determine temporality between exposure and disease

Can study multiple outcomes (diseases)

Cohort (prospective or retrospective):Limitations

Expensive

Loss to follow up (retention bias - a form of selection bias)

not good for rare diseases

Experimental (Intervention study): Strengths

Researchers have greatest control

cause and effect — going directly from exposure to disease

minimize risk of confounding (researchers have control)

Experimental (Intervention study): Limitations

Ethical issues in human subjects research

Hawthorne Effect

Artificial setting - ie may not be able to replicate results in clinical practice

Selection bias is common in which 2 study designs

Case Control and Cohort

Selection bias in Case-Control Studies

Control Selection Bias

Self-Selection Bias

Selective Survival/loss to follow up

differential surveillance, diagnosis or referral

Selection Bias in Cohort study

subject selection bias (same as self selection)

selective survival/loss to follow up

Healthy worker effect

Information Bias is found in which two study designs

Case Control and Cohort

Information Bias in Case Control and Cohort Studies

Non-Differential Misclassification of Exposure

Differential misclassification of Exposure OR outcome

• recall bias

• interviewer bias

• prevarication/lying bias

non-differential misclassification of Outcome