RCT- Dr.Z

FYI on what to focus on for each section for Dr. Z:

• green section- characteristics of RCTs

• 2 Exam Q

FYI on what to focus on for each section for Dr. Z:

• green section- characteristics of RCTs

• 2 Exam Q

FYI on what to focus on for each section for Dr. Z:

• blue section- internal v external validity

• 2 Exam Q

FYI on what to focus on for each section for Dr. Z:

• purple section- randomization, blinding, and sample size

• 2 Exam Q

FYI on what to focus on for each section for Dr. Z:

• orange section- different equations dealing w/ risk

• 2 Exam Q

FYI on what to focus on for each section for Dr. Z:

• pink section- 3 strengths and weaknesses of RCTs

• 1 Exam Q

What are the 10 characteristics of RCTs?

• 2 EXAM Q from this LIST!!!!

1. Primary study

2. Near Perfect Conditions

3. Level 1 Evidence

4. Experimental

5. Strongest type of study design

6. strong internal validity

7. Prospective

8. Unfiltered

9. Best Research

10. GOLD STANDARD

A characteristic of RCT is strong internal validity. If a RCT has strong internal validity, it indicates what? Additionally, to minimize threats to Internal validity, what can we do?

Strong Internal Validity= Causal Relationship

To minimize threats: control group, randomization, blinding, etc.

What is bias and how is it minimized?

bias- systematic error in study design or implementation that can lead to incorrect findings

---

to minimize- use the right study design and implement it

What’s validity and the 4 subtypes of validity?

Validity- degree to which findings represent TRUTH (I think we know what the word valid means)

---

1. face (are they testing what they say they are going to test)

2. content- (is it correct to experts?)

3. criterion- (Does it match up w/ a gold standard)

4. construct- (Does it relate to theoretical concepts)

What is internal validity? How is internal validity strengthened?

Internal Validity- Degree to which the study outcome (efficacy or safety) can be explained by the differences in the intervention (treatment)

---

Strengthened: by having a control group and randomization

List and briefly describe the 8 key biases that threaten internal validity?

1. selection Bias- putting pts. in specific groups

2. history Bias- study outcome may be bc of External events

3. maturation Bias- normal change in pts. over time

4. Attribution Bias/ Mortality- “Drop-Out” (pt. leaves)

5. Testing Bias- pts. taking tests repeatedly

6. Instrumentation Bias- sensitivity of instrument, improved technology, changes in measurement

7. Investigator Bias- error by investigator

8. Detection/Surveillance Bias- difference btwn groups in how outcomes are determined

Which of the 8 key biases can be minimized by randomization?

• selection bias

• maturation bias

Which of the 8 key biases can be minimized by control group?

• history bias

• maturation bias

• testing bias

Which of the 8 key biases can minimized by blinding?

• investigator

• detection/surveillance bias

What is external validity and what criteria does it relate to?

indicates that the findings of a given study can be generalized to other settings

---

inclusion and exclusion criteria

What 5 key factors that threaten external validity?

1. subject selection

2. treatment of patients

3. study location/setting

   • HAWTHORNE EFFECT

4. historical factors

5. multiple treatments

What is the Hawthorne Effect?

study subjects modify their behaviors due to longitudinal learning over the course of a study or because they know they are being observed

• basically: subjects act different because they know ppl are watching them

What is randomization and what are the 3 methods of randomization?

process of assigning pts. to a treatment or control group based on chance alone

---

1. simple- random # generator

2. stratified- by characteristics

3. block- divided into blocks before, then blocks randomized

List and describe the 4 types of blinding:

1. single blind- pt/investigator unaware

2. double blind- both pt/investigator unaware

3. triple blind- everyone unaware (pts, investigator, data collectors)

4. open label- everyone IS AWARE

Sample size is the number of subjects that are needed to test the primary hypothesis… WHAT 2 KEY FACTORS does it take into consideration.

• PLEASE KNOW THIS

1. EFFECT SIZE

2. POWER

What are the 3 most common types of control groups and when are they used?

1. placebo control- when no known effective therapy

2. active control- when KNOWN effective therapy

3. historical (external) control- pts. compared to an old control group

Describe each of the following RCT designs:

1. parallel

2. crossover

3. superiority

4. noninferiority

1. parallel- each subject either tx or control group

   • ex: drug A or placebo

2. crossover- each subject receives all txs

   • ex: drug A and drug B

3. superiority- one better than the other

4. noninferiority- sees whether two interventions are equivalent

2 disadvantages of crossover RCT design:

1. carryover effect

2. washout period

What is intention-to-treat analysis?

• analyzes patients AS IF THEY COMPLETED the study in their originally assigned group

• preferred in superiority trials

Describe on-treatment (per-protocol) analysis:

• we can see the true effect of study drug

• preferred in noninferiority trials

What is risk and how do we calculate it?

probability of an event when an intervention is given

---

What is relative risk and how do we calculate it?

• difference between groups in PROSPECTIVE STUDIES

• RR < 1 means lower risk/negative association

  ---

  RR= risk in tx group/ risk in control group

What is relative risk reduction and how do we calculate it?

how much the risk is reduced in the treatment group compared to the control group

---

RRR- 1- RR

What is absolute risk reduction and how do we calculate it?

• true difference in risk

• increase utility

---

ARR= (risk in control group) - (risk in treatment group)

What is the “number needed to treat” and how do we calculate it?

• effectiveness of intervention

• want a LOW number

• always ROUND UP

---

NNT= 1/ ARR

What is the “number needed to harm” and how do we calculate it?

• risk to pt. from an intervention

• ROUND DOWN

---

NNH= 1/ ARR

What are the 3 strengths of RCTs?

1. can determine UNBIASED EFFICACY of an intervention

2. PROSPECTIVE DESIGN allows for control

3. Methods and protocols allow ability to IDENTIFY CAUSAL RELATIONSHIPS

What are the 3 weaknesses of RCTs?

1. limitations on generalizability

2. potential HAWTHORNE EFFECT

3. takes lots of time/expensive