STAT 509 Lesson 6 - Objectives, Endpoints, and Estimation

Vocabulary terms and definitions covering clinical trial objectives, various endpoint types, censoring methods, surrogate outcomes, dose-finding strategies, statistical risk measures, and survival analysis based on STAT 509 Lesson 6.

Clinical trial objective

Must be stated in specific terms; success should depend on obtaining a valid result rather than observing one predetermined numerical outcome.

Endpoint (outcome)

A quantitative measurement determined for each study participant that is required by the scientific objectives.

Primary objective / primary endpoint

The principal objective or outcome of a clinical trial; it is the basis for the trial's sample-size calculation.

Secondary objective / secondary endpoint

Additional objectives or outcomes beyond the primary one; their statistical power is constrained by the sample size selected for the primary objective.

Hard endpoint

An endpoint that is well-defined in the protocol, definitive with respect to the disease process, and requires no subjectivity.

Soft endpoint

An endpoint that does not relate strongly to the disease process or requires subjective assessment by investigators and/or patients.

Continuous measurement

A numerical measurement on a continuous scale, such as blood pressure, weight, or blood chemistry variables.

Event time

Time until an event occurs, such as time to cancer recurrence or survival time.

Count

Number or frequency of occurrences, such as migraine-headache frequency or number of rescue-medication uses.

Binary endpoint

An outcome with exactly two categories, such as recurrence/no recurrence or major cardiac event yes/no.

Ordered categories

Categories with a meaningful order, such as absent, mild, moderate, or severe pain.

Unordered categories

Categories without a natural ranking, such as adverse-experience categories like GI and cardiac.

Repeated measurements

Endpoints that are assessed multiple times during the course of the study.

Event-time endpoint

Measures the time from study entry until a specified event occurs.

Censoring

Occurs when some follow-up is available but the exact event time for a participant is not fully observed.

Right censoring

Occurs when the event has not been observed by the last known follow-up time due to study termination, loss to follow-up, or death from an unrelated cause.

Left censoring

Occurs when the initiation time for the subject, such as the exact time of diagnosis, is unknown.

Interval censoring

Occurs when a subject is not observed for a period and it is unknown exactly when within that interval the event occurred.

Type I right censoring

Censoring where all subjects begin at the same time and the study ends at the same fixed time.

Type II right censoring

Censoring where all subjects begin at the same time and the study terminates when a predetermined proportion have experienced the event.

Type III right censoring

Censoring that is random, commonly occurring in clinical trials due to staggered entry and unequal follow-up.

Independent censoring assumption

The assumption in event-time methods that the censoring mechanism and the event occurrence are independent.

Informative censoring

Occurs when patients are censored before the event occurs for reasons related to the event, potentially leading to an underestimation of the event rate.

Surrogate endpoint

An outcome measured in place of a biologically definitive or clinically meaningful endpoint; it usually tracks disease progress or extent.

Surrogate causal pathway

$$\text{Disease} \rightarrow \text{surrogate endpoint} \rightarrow \text{definitive endpoint}$$

Treatment Mechanism study

An early developmental trial, such as a pharmacokinetic study, investigating the mechanism of treatment effect like absorption and elimination.

Dose-escalation

A design component specifying how doses are increased for subsequent groups of subjects.

Dose-ranging

A design testing some or all of a prespecified set of fixed dose levels.

Dose-finding

A design that titrates the dose toward a prespecified optimum based on biological or clinical considerations.

Optimal Biological Dose (OBD)

The goal of a Phase I dose-finding study; the dose identifying high efficacy potential while meeting safety criteria.

Therapeutic index criterion

A method to define an optimum dose as the dose with the maximal separation between risk and benefit.

Minimum Effective Dose (MED)

Example: For a pain medication, the dose that eliminates mild-to-moderate pain in $80\%$ of participants.

Maximum Nontoxic Dose (MND)

Example: The highest dose associated with serious side effects in no more than $1$ of $20$ patients.

Maximum Tolerated Dose (MTD)

Example: For a cytotoxic cancer drug, a dose yielding serious but reversible toxicity in no more than $30\%$ of patients.

Dose-limiting toxicity (DLT)

A specific toxicity definition used to govern dose-escalation and de-escalation decisions.

Continual Reassessment Method (CRM)

An algorithm that fits a mathematical model to accumulating trial data to repeatedly update the best current guess of the optimal dose.

Fibonaccisequence

$1, 1, 2, 3, 5, 8, 13, 21, 34,...$; each new number is the sum of the two previous numbers.

Risk Difference (RD) formula

$RD = p_T - p_P$, where $p_T$ and $p_P$ are event probabilities in the treatment and placebo groups respectively.

Relative Risk (RR) formula

$$RR = \frac{p_T}{p_P}$$

Odds formula

$$\text{Odds} = \frac{p}{1-p}$$

Odds Ratio (OR) formula

$$OR = \frac{\frac{p_T}{1-p_T}}{\frac{p_P}{1-p_P}}$$

Estimated Relative Risk (2x2 table) formula

$\hat{RR} = \frac{\frac{a}{a+b}}{\frac{c}{c+d}}$ based on successes (a, c) and failures (b, d) in Group 1 and 2.

Estimated Odds Ratio (2x2 table) formula

$$\hat{OR} = \frac{ad}{bc}$$

SE of log odds ratio

$$SE[\ln(\hat{OR})] = \sqrt{\frac{1}{a} + \frac{1}{b} + \frac{1}{c} + \frac{1}{d}}$$

Mantel-Haenszel assumption

The assumption that the odds ratio is equal across different strata even if event rates differ among those strata.

Dose-response / trend analysis

Tests whether increasing dose is associated with an ordered increase or decrease in response.

Jonckheere-Terpstra (JT) trend test

A nonparametric trend test based on the sum of Mann-Whitney-Wilcoxon comparisons across ordered pairs of dose groups.

Cochran-Armitage (CA) trend test

A trend test for binary data that can utilize actual numerical dose levels.

Kaplan-Meier estimator

A nonparametric method for estimating survival probability over time in the presence of censoring.

d_k

The number of failures or events occurring at the distinct observed failure time $t_k$.

n_k

The number of patients at risk immediately before failure time $t_k$, excluding those censored before that time.

Kaplan-Meier formula

$$\hat{S}(t) = \prod_{t_k \leq t} (1 - \frac{d_k}{n_k})$$

Logrank test

A Mantel-Haenszel-type nonparametric test used to compare survival curves across treatment groups, assuming proportional hazards.

Generalized Wilcoxon test

A nonparametric survival-curve comparison that extends the Wilcoxon rank-sum test to censored data; it does not assume proportional hazards.

Bootstrap

A resampling method used to estimate the variability or variance of an estimator by drawing samples with replacement from the original dataset.

Bootstrap variance formula

$$Var_{boot}(\hat{\theta}) = \frac{1}{B-1} \sum (\hat{\theta}_b - \bar{\theta}_{boot})^2$$

Exploratory analysis

An analysis focusing on objectives not specified in the primary protocol; used to generate hypotheses rather than confirm them.

Multiple-testing problem

Increasing the number of hypothesis tests inflates the Type I error rate, raising the chance of finding a significant result when no true effect exists.

Subset analysis

Comparing treatment groups within specific subgroups (e.g., gender, age, cholesterol levels); remains exploratory unless planned in advance.