Randomized Controlled Trials Questions and answers 2024

What is basic research? Basic research is bench research. What is applied research? Study on people, animals, in a lab, etc. What is the goal of research? To draw inference from the study finding to the general population. What is internal validity? Ability to make an inference from the study findings to the truth in the study (study sample to the population of interest). Difference between study plan, intended samples, and intended variables to actual study, actual subjects, and actual variables. What is external validity? Ability to make an inference from the truth in the study to the general population (population of interest to the general population). Difference between research question, target population, and phenomena of interest to study plan, intended sample, and intended variables. What are systematic errors? Bias and confounding. Disrupt internal validity. What are random errors? Chance. These also disrupt internal validity. What is considered the gold standard of study design? Randomized controlled trials (RCT). What are advantages of RCTs? Determine causality, minimize bias, and eliminate/reduce confounding variables. What are disadvantages of RCTs? Narrow clinical question, more expensive, less generalizable due to control of study and narrow question, take a long time, and issue with recruitment of subjects. What is the goal of a superiority RCT? Show that one treatment is superior to the other. What is the goal of an equivalence RCT? Show that the treatments are equivalent. What is the goal of a non-inferiority RCT? Show that one treatment is not inferior to the other (standard). What are extraneous variables? Variables that need to be accounted for or they become confounding variables. EX: a patient taking antibiotics in an infectious disease study. What are active variables? Manipulated by the investigator (or random). EX: intervention or control, medication or placebo, etc. What are attribute variables? Not able to be manipulated and are observed within groups. EX: patient demographics. What is intervention considered in a RCT? The intervention is a type of active variable. Single intervention is easier to determine causality. What is the control group? Control group is also an active variable against which experimental/intervention group is compared. What are types of control groups? No therapy, standard of care, and placebo. What is the goal regarding the difference of control and intervention groups in RCTs? The goal is to make the only difference between the groups the difference in independent variables. What step is important to remember in crossover trials? Allow the drug to wash out of the subjects' systems before switching. What is a run-in period? Used in RCTs before the actual trial to test compliance. What are surrogate markers and when are they used? Measures in changes in intermediate factors in the main pathway of the clinical outcome that are used when the clinical outcome is difficult to measure and determine. What is efficacy? Benefit of intervention compared to control. What is effectiveness? Benefits under "real world" conditions. Why are analysis considerations important to read about in a study? Explains how the subjects who did not complete the study were analyzed. What is intention to treat (ITT)? People who didn't complete the study did not have the primary outcome. If the primary outcome was cure, the overall cure rate will go down. The cure rate will be a conservative estimate of efficacy. What is per protocol (PP)? Only patients who completed the entire study are analyzed. If the primary outcome was cure, the overall cure rate will be higher than actual. In what situation would PP be preferred? If the outcome of a study is toxicity, PP would more accurately show the people who completed the study and had a toxicity. If ITT was used in this situation, it would deflate the rate of toxicity (assuming anyone who quit the study did not have toxicity). What is a type 1 error? Stating that there is a difference when there really is not. False positive. What is a type 2 error? Stating that there is not a difference when there truly is. False negative. What is beta? Probability of making a type 2 error. What is the power of a study? 1-beta. Usually 80-90% power. Associated with the sample size. What is sample size? Number needed in each group to detect a difference. What is the effect size? Minimum desired difference that the researchers are looking for. Based on current efficacy data to drive the rates and the least difference that is relevant in practice. When is a difference significant/relevant compared to the effect size? If a difference is above the effect size, then it is significant/relevant. EX: effect size is 10% and difference is 15%. What signifies effect size? Delta. What is a confidence interval? Shows the range of results likely to represent the true population value that contains a point estimate (result from the study). Used to draw conclusions from sample values. Generally reported as 95% CI.