N236: Concepts & Methods in Infectious Disease Research

Scientific method steps

Observation, Question, Hypothesis, Experiment, Data Collection, Conclusion, and Retest (if needed)

What is the goal of a scientific study?

Describe the relationship between an exposure (E) and a disease (D), after taking into account the effects of other factors (C, control variables)

How to calculate association

1. calculate measures of disease frequency among “treatment” and “control” groups

2. Compare the measures of disease frequency to obtain a measure of association

3. Interpret the measure of association

Rate ratio

rate in treatment/rate in control (no units)

Rate difference

rate in treatment - rate in control (units)

Risk ratio (aka relative risk)

Compares the risk of contracting a disease between the exposed population and the unexposed population

Risk ratio formula

Risk of exposed/risk of unexposed (OVER TIME)

What does a relative risk of > 1 indicate?

The exposure is harmful.

What does a relative risk of < 1 indicate?

The exposure is protective.

What does a relative risk of = 1 indicate?

The exposure is neither harmful nor protective

Risk difference formula

Risk exposed - risk unexposed

What does a relative difference of > 0 indicate?

The exposure is harmful.

What does a relative difference of < 0 indicate?

The exposure is protective

What does a relative difference of = 0 indicate?

There is no association between the exposure and the outcome

Descriptive studies

seek to describe the distribution factors or protective factors in populations, hypothesis-generating

Analytic studies

seek to identify possible risk factors or protective factors using smaller samples, hypothesis-testing

2 types of analytic studies

experimental and observational

Experimental studies

investigator manipulates the exposure status and follows participants to observe the outcome of interest

Observational studies

investigator simply observes the pre-existing exposure status and the outcome of interest

"Crude" analysis

Does not take into account other factors that may influence the relationship between an exposure and a disease

"Adjusted analysis"

Takes into account other factors (control variables) that may influence the relationship between an exposure and a disease

Study design

A set of procedures that are used to collect and analyze data on variables

Randomized clinical trial

An experimental study designed to test the efficacy of a treatment

Preventive trial

A type of randomized clinical trial that evaluates whether an agent reduces the risk of developing a disease (ex: COVID-19 vaccine trials)

Therapeutic trial

A type of randomized clinical trial that evaluates whether a treatment was effective in treating a disease (ex: Paxlovid treatment trials)

Randomized clinical trial study design

Subjects are randomly and blindly assigned by the investigator to either a "treatment" group or a "control" group

What is the point of randomization?

used to put participants in groups that are comparable on all factors except for exposure factor (treatment or placebo)

Unblinded study

Both subjects and investigators are aware of the exposure assignment

Single-blinded study

Investigators are aware but subjects are unaware of the exposure assignment

Double-blinded study

Both subjects and investigators are unaware of the exposure assignment

Cohort studies

Observational study in which subjects who do not have the disease of interest are followed in time to observe if they develop the disease

Cohort study design

Investigator begins with an "exposed" group of subjects and an "unexposed" group of subjects --> then follows them in time to see who develops the disease

Strengths of cohort studies

allows for establishing the temporal sequence of events because it works forwards from exposure to disease, can be used to study several diseases simultaneously, can be used to study rare exposures

Limitations of cohort studies

usually time-consuming and more expensive to conduct than other observational study designs, loss of participants due to withdrawal, migration, or death could result in errors, not efficient for studying rare diseases

Case-control studies

Observational study in which subjects are selected based on their current disease status and then assess them for exposure history

Case-control study design

Compares a group of people with disease (cases) to a group without disease (controls) --> then looks backward in time for prior exposure

Can you measure risk or rate in a case-control study?

No, because you cannot measure incidence (new cases)

Strengths of case-control studies

quicker and relatively less expensive to conduct than cohort studies, can be used to study several exposures simultaneously, can be used to study rare diseases

Limitations of case-control studies

only one disease can be evaluated at a time (in contrast to cohort studies), does not allow for the direct assessment of disease risk because it works backwards from disease to exposure, not efficient for studying rare exposures

Odds ratio

The likelihood of a disease among individuals exposed to a risk factor compared to those who have not been exposed

Odds (cross product) ratio formula

ad/bc

What does a odds ratio of > 1 indicate?

The exposure is harmful

What does a odds ratio of < 1 indicate?

The exposure is protective

What does a odds ratio of = 1 indicate?

There is no association between the exposure and the outcome

Cross-sectional studies

An observational study in which you ask a group of people about their disease status and exposure at the same time

Cross-sectional study design

Investigator interviews group of individuals from a population, and obtains data on their disease status and exposures at the same time (a snapshot in time)

What association can you measure in cross-sectional studies?

Prevalence!

Prevalence ratio

Prevalence (exposed) / Prevalence (unexposed)

What does a prevalence ratio of > 1 indicate?

The prevalence in exposed is higher.

What does a prevalence ratio of < 1 indicate?

The prevalence in exposed is lower

What does a prevalence ratio of = 1 indicate?

The prevalence in exposed and unexposed is similar.

Prevalence difference

Prevalence (exposed) - Prevalence (unexposed)

What does a prevalence difference of > 0 indicate?

The prevalence in exposed is higher

What does a prevalence difference of < 0 indicate?

The prevalence in exposed is lower.

What does a prevalence difference of = 0 indicate?

The prevalence in exposed and unexposed is similar.

Strengths of cross-sectional studies

convenient, quicker, and relatively less expensive to conduct than other study designs, can be used to evaluate several exposures and several diseases at the same time, can help generate clues about potential associations

Limitations of cross-sectional studies

cannot help establish whether the exposure preceded disease or whether the disease influenced exposure, can only identify prevalent (not incident) cases, may miss diseases with shorter durations