Chapter 5. Introduction to Clinical Experimentation

research class

Evidence-Based Medicine

The practice of medicine that involves making clinical decisions based on the best available, current, valid, and relevant evidence.

Clinical Study

Any attempt to learn more about a disease and its manifestations, causes, or outcomes, varying in size from small case descriptions to large trials.

Clinical Trial

A specific type of clinical study that evaluates the effects of a specific medical or health-related intervention in humans.

Quantifiable Outcome

An outcome in a clinical trial that can be measured in a way that quantifies the effectiveness of an intervention.

Control Group

A group in a clinical trial that does not receive the treatment being tested, used as a benchmark to compare the effects of the experimental treatment.

Randomization

The process of assigning trial participants to treatment or control groups randomly to minimize bias.

Blinding

A method used in clinical trials to prevent bias by ensuring that participants and/or investigators do not know which treatment participants are receiving.

Null Hypothesis

The hypothesis that there is no effect or no difference, which is tested in a clinical trial.

Type I Error

The error that occurs when the null hypothesis is incorrectly rejected, indicating a false positive.

Type II Error

The error that occurs when the null hypothesis is not rejected, indicating a false negative.

Phase I Trial

The first stage in clinical drug development, focusing on safety, dosage, and pharmacokinetics, often involving healthy subjects.

Phase II Trial

The stage that evaluates the effectiveness of a drug while establishing short-term side effects and risks.

Phase III Trial

A large-scale trial conducted to confirm effectiveness, monitor side effects, and compare the drug to commonly used treatments.

Phase IV Trial

Post-marketing studies done after a drug has been approved to evaluate long-term effectiveness and safety.

Statistical Significance

A statistical result that is unlikely to have occurred by chance, often indicated by a p-value less than a specified alpha level, usually 0.05.

P Value

The probability of obtaining a test statistic at least as extreme as the one observed, assuming the null hypothesis is true.

Surrogate Outcome

A measure used as a substitute for a clinically meaningful endpoint, often used in Phase II studies to infer actual clinical benefit.

scientific method

construction of a hypothesis, design of an experiment, analysis of the data, and communication of the results

retrospective studies

collect and analyze data from events that have already occurred

prospective studies

identify a population of subjects and follow them into the future from a specified time point, collecting data on events that occur as time passes

types of clinical studies

relevancy of outcome measures is also known as

content validity

content validity

a measurement validity that assesses how well a test or assessment represents a defined construct or domain of knowledge

The basis for inference from a clinical trial is

the observed difference in the summarized outcome measure(s) between a group that received an experimental treatment and a group that received the control treatment

treatment effect

a measurement that compares the outcomes of a treatment to the outcomes of not receiving the treatment

commonly used stratification factors are

sex, age, race, as well as known measures of risk observable at baseline

randomization

randomly sorting participants into groups

assures that the treatment a pt will receive is determined by chance and is unrelated to characteristics of the specific pt

blinding

when investigators and pts are unaware of which treatment is received

reduces the potential for bias in the reporting of measures taken during the study

null hypothesis

assumption that there is no difference btwn the test and control

delta=0

in a positive trial, the observed delta. . .

would be so different from 0 that we would not believe that the null hypothesis would be true

so you reject the null hypothesis

if the observed delta is not far enough from 0 . . .

then it is a negative trial

aka a failure to reject the null hypothesis

types of error

statistical significance

is reported when the computed test statistic is so large that, if the test and control arms were not really different and we were to repeat our clinical trial many times, no more that 100 x a% of the repeats would produce a test statistic as large as the one we observed

the P value is

the computed probability of observing a test statistic at least as large as the one we observed, assuming the null hypothesis of no difference between the treatment arms

statistical significance is claimed if

the P value is less than the prespecified α

clinical significance

assess the apparent net benefit of a tx for individual pts taking into account other factors such as the risk of adverse events or cost

random assignment of pts and type I error

randomization reduces the potential for bias in the treatment comparison by balancing prognostic factors between treatment groups

randomization introduces the element of chance necessary to interpret the P value as a probability statement

the value of a

the level of significance

is specified in the design of the clinical trial

a is usually set to

0.05

when a=0.05, the type I error is really

0.025 (0.05/2)

bc cases in which the control tx is significantly better than the test tx do NOT count as positive trials

to control type II error (β)

we consider a hypothetical nonzero difference between the test and control treatment arms (ΔA = μT – μC) and the anticipated variation in ΔA (sdiff = standard deviation of ΔA)

The suggestion that ΔA > 0 is called the

alternative hypothesis

power of the trial is

1 – β and is the smallest probability of a positive trial, assuming that the true average difference between test and control treatment arms is at least ΔA

ΔA is

the smallest detectable difference between treatment arms, assuming power of at least 1 – β

The required sample size for a trial ___________ with increasing ΔA

decreases

power is chosen to be at least _______, so that Type II error is less than ______

0.8; 0.2

Smaller Type I error, smaller Type II error, and smaller ratios require

larger sample sizes

relationship between ΔA/s and sample size

possible effects of sample size on Type II error

types of trials designed to show that tx arms are not different are

equivalence trials and noninferiority trials

equivalence trials

designed to show that two treatments give similar results

noninferiority trials

where the objective is to show that the test treatment is not worse than the control by more than a specified amount

Pharmacokinetic studies often do not have a

specified hypothesis test, as the objective of these trials is to estimate variables related to the pharmacology of the treatment

clinical development for new drugs often have _______ phases

4 phases

phase I

the drug is first introduced to humans

usually as a single dose or injection given to a small number of healthy subjects

why is knowing a drug’s half-life important?

because longer half-lives can be associated with the possibility of accumulation and increased risk of side effects

comparison of clinical trial phases

later phase I studies might explore

ADME of the drug by evaluating multiple-dose regimens, bioavailability of oral forms, and comparison of pharmacokinetics between fed and fasting conditions

phase II studies

begin to establish evidence of drug effect for its intended indication

phase IIa should identify a

“target” dose at which maximal efficacy is expected to be observed

phase IIb trials

are randomized and have sample sizes large enough to obtain more precise estimates of efficacy and adverse event rates

where the target dose is tested

An important goal of drug development is to show

that the test treatment is statistically significantly better than the control with respect to outcome measures reflecting clinical benefit, such as survival, time to adverse outcome, disease symptom score, or functional status

surrogate outcomes

a physical sign or laboratory measurement that can be used to predict a clinically important outcome

Surrogate outcomes are used in clinical trials when it's not possible or practical to measure the clinical outcome directly

phase III

where demonstration of clinical benefit is accomplished

phase IV testing obtains

further evidence regarding safety, such as effects of chronic treatment or adverse event profiles in subgroups