EBM madule 3

What is random error

Error due to chance or probability

How is random error measured

Using p-values and confidence intervals

What causes random error

Sampling because only a representative sample of the population can be studied

How does sample size affect random error

Smaller samples have a greater likelihood of error in results

How can random error be reduced

Increase sample size

What is systematic error

Error caused by systematic problems in study design that cause results to deviate from true population statistics

What is another term for systematic error

Bias

How does bias affect internal validity

It undermines internal validity

How can bias be avoided

Through good study design

What are the two main categories of bias

Selection bias and information bias

What is selection bias

Bias that occurs when the people selected for a study or who remain in it result in groups that are not comparable

What is detection bias

Bias caused by different methods of screening for a condition between control and experimental groups

What is chronology bias

Bias caused by using historical controls instead of controls enrolled during the same time period

What is prevalence-incidence bias also called

Neyman bias

What is prevalence-incidence or Neyman bias

Bias occurring when mild self-resolving cases and rapidly fatal cases are not captured

What is information bias

Bias resulting from incorrect determination of exposure outcome or both

What is misclassification bias

Incorrectly classifying a patient as having a disease disorder or outcome without meeting accepted criteria

What is interviewer bias

Bias caused by interviewing people in different groups non-uniformly

What aspects of an interview can cause interviewer bias

Wording or order of questions tone of voice gestures and other differences in interviewing

What is recall bias

Bias that occurs when cases recall past exposures more accurately than controls or are more willing to report them

What is an observational study

A study in which treatment groups and other events are not under the investigator's control

Are observational studies randomized

No observational studies are non-randomized

What can observational studies establish about exposure and outcomes

Association but not necessarily causality

What are the two time orientations of observational studies

Prospective and retrospective

What is a descriptive study

A study with one study group

What is a comparative study

A study with more than one study group

Why are selection bias and confounding concerns in non-randomized studies

Because randomization is absent

What is confounding

A situation in which another variable is associated with the exposure and outcome and can distort their apparent relationship

What is the confounding variable in the example involving sunburn and ice cream sales

Hot weather

Why does the relationship between ice cream sales and sunburn not establish causality

Hot weather can influence both ice cream sales and sunburn

What methods can adjust for confounding in observational studies

Regression and propensity score methods

What are the observational study designs listed in Module 3

Non-randomized intervention studies cohort studies case-control studies and cross-sectional studies

What is the STROBE statement

Guidelines for reporting observational studies

What is a non-randomized interventional study

An experimental study in which the investigator assigns the treatment or intervention without randomization

What is another name for a non-randomized interventional study

A quasi-experimental study

When are non-randomized interventional studies commonly used

When randomization is unethical or infeasible

Why might randomization be unethical in a clinical situation

It may be inappropriate to withhold a necessary treatment such as an anticoagulant in a patient experiencing an intracranial hemorrhage

What RCT-like components may non-randomized interventional studies have

Well-timed and well-defined endpoints and a data adjudication committee

How might outcomes be evaluated in a non-randomized interventional study

By comparing outcomes before and after the intervention

What is an example of a before-and-after intervention comparison

Comparing intracranial hemorrhage volume at baseline and after treatment administration

What is a prospective cohort study

A study that follows people forward in time from exposure to outcome

What can prospective cohort studies measure

Incidence risk and time-to-event

What is an example of a prospective cohort study involving statins

Following adults without a prior myocardial infarction to see who experiences a first myocardial infarction

Why can prospective data collection help identify myocardial infarctions

It allows events to be assessed using guideline criteria and appropriate information

How could a prospective cohort study evaluate chemotherapy-related cardiomyopathy

By following patients receiving chemotherapy with serial echocardiograms and troponin measurements

Why might retrospective data miss chemotherapy-related cardiomyopathy

Serial echocardiograms or troponin measurements may not have been performed or may not have been performed at the needed time

What is a retrospective cohort study

A study that begins with existing records after the exposure has already occurred

What is an example of a retrospective cohort study involving SGLT2 inhibitors

Identifying SGLT2 inhibitor users and nonusers in 2021 and tracking heart failure admissions through 2024

What bias may occur when records do not establish whether a patient actually took an SGLT2 inhibitor

Detection bias or misclassification bias

Why might a retrospective cohort study misclassify a heart failure hospitalization

It may be difficult to determine whether the admission was caused by heart failure or another condition

What should researchers consider before conducting a retrospective cohort study

Whether the data needed for the research question are available

What are common data sources for cohort studies

Electronic health records administrative claims databases and disease-specific registries

What are strengths of cohort studies compared with RCTs

They can be less costly less time-consuming and can demonstrate temporal association

What does temporal association mean in a cohort study

The exposure occurs before the outcome

When might a cohort study be the best or only possible study design

When an RCT would be unethical or infeasible

Which medication-use behaviors are well suited to cohort studies

Medication adherence and persistence

What is a case-control study

A study that starts with outcome status and looks back to compare exposure histories

What does a case-control study investigate

Associations between exposures or risk factors and a disease or condition

Who is a case in a case-control study

A person who has the disease

Who is a control in a case-control study

A person who does not have the disease

When are case-control studies often used

To study rare diseases or outcomes that take a long time to develop

When are case-control studies not suitable

When the exposure of interest is rare

What is one advantage of a case-control study regarding exposures

It can investigate several possible exposures

What are advantages of case-control studies regarding resources

They require smaller sample sizes and are relatively inexpensive and quick

What is a cross-sectional study

A study that measures exposures and outcomes at one point in time

Why is a cross-sectional study described as a snapshot

It captures exposure and outcome information at a single point in time

What are the main uses of cross-sectional studies

Estimating prevalence and examining associations

Why can a cross-sectional study not establish causality well

It cannot clearly determine whether the exposure occurred before or after the outcome

What are advantages of cross-sectional studies

They are quick and low-cost

What type of study are administrative claims database studies usually

Retrospective cohort studies

What is real-world evidence

Evidence collected outside the tightly controlled RCT setting

What patient populations may real-world evidence include compared with RCTs

More diverse patient populations

Can real-world evidence data be collected prospectively or retrospectively

Yes

How are treatment recommendations generally made in real-world settings

Through guidelines and physician discretion without a strict study protocol

What is one limitation of event reporting in real-world evidence

Events may be overreported or underreported

How do RCT populations differ from real-world populations

RCTs use preselected populations and strict inclusion and exclusion criteria

How do RCT protocols differ from real-world practice

RCTs use strict study protocols while real-world treatment decisions are often made at physician discretion

How are event rates typically assessed in RCTs according to Module 3

They are objectively adjudicated

What is a limitation of RCT generalizability

RCTs show how a product performs in a particular study population but that population may not reflect real-world patients

What is pharmacovigilance

Monitoring and evaluating medication safety

What is the SEERS program used for in real-world evidence

Surveillance epidemiology and end-results information

What is the FDA Sentinel program used for

Safety surveillance and analysis using real-world data

What is MedWatch

An FDA system associated with safety communications and analyses

What are REMS programs

Risk Evaluation and Mitigation Strategy programs used to fulfill regulatory obligations

What does PMSS stand for

Post-Marketing Safety Surveillance

What is comparative effectiveness research

Research that compares the effectiveness of treatments or healthcare interventions

How can real-world evidence be used to evaluate medication use

By examining prescribing patterns and medication utilization

How can real-world evidence support quality improvement

Through tools such as disease registries

How does real-world evidence relate to RCT evidence

It complements rigorous but tightly controlled randomized trials

What are examples of real-world evidence study types

Pragmatic RCTs prospective registries prospective cohort studies retrospective observational studies and administrative claims database analyses

What is a pragmatic RCT according to Module 3

An RCT in which participants are randomized to a plan rather than a specific medication

Can internal and external validity differ across real-world studies

Yes validity can vary markedly between and within real-world study designs

What is an administrative claims database

A database containing information needed to facilitate reimbursement for healthcare services

What demographic information may claims databases contain

Age gender race and insurance type

What healthcare encounter information may claims databases contain

Types and counts of hospitalizations office visits and diagnostic or laboratory tests

What is ICD-10 used for in claims databases

Codes depicting diseases present

What is CPT used for in claims databases

Codes used to bill for professional time and procedures in inpatient and outpatient settings

What is a DRG

A diagnosis-related group that aggregates ICD-10 codes to determine reimbursement for a hospitalization

What does a DRG not include

All professional time

What are prescription drug fill records

Records generated by a pharmacy when medications are picked up