The objective of a clinical trial is the research question phrased in concise quantitative terms.
So-called endpoints are obtained and used to answer the research question of interest.
Example: asthma trial
objective: to improve lung function
endpoint: forced expiratory volume in one second - FEV1
Piantadosi (1997), p 127
‘Endpoints are the quantitative measurements implied or required by the objectives. An endpoint is determined in each study subject, whereas the objectives are met by the aggregate of endpoints.’
The choice of primary endpoint is crucial and should be clearly stated in the research protocol. Secondary endpoints often feature too.
Sample size calculations are based upon the primary endpoint (efficacy measure) - more on this later.
Estimation of an important quantity is often phase specific. For example:
maximum tolerated dose (MTD) in phase I trials
dose-response relationship (function) in phase II
treatment differences in phase III ( in means/proportions)
event rates of adverse events in phase IV,
or when selecting a treatment regime.
For example:
“optimal dose”
testing procedures
time.
“hard”: preferred!
Well defined and reliable
no subjectivity, objectively measured
examples: death, presence of ulceration, laboratory measurements (e.g weight, blood pressure, serum cholesterol).
“soft”
difficult to define, subjective measures
examples: satisfaction with treatment, pain (Visual Analogue Scales), depression etc (i.e. subjective measures).
Between the extremes: e.g. pathologic classification depending on expert’s opinion and experience, but reliable.
discrete: e.g. length of stay in hospital[days]; number of seizures
continuous: e.g. laboratory parameters; blood pressure; serum cholesterol.
ordered: e.g. pain: minimal/ moderate/ severe/ unbearable
nominal:
e.g. juvenile-onset diabetes/ maturity-onset diabetes/ non-diabetic
special case: binary
e.g. diabetic / non-diabetic.
Analysis methods are dependent upon the scale of measurement.
from baseline.
no. of events within certain time interval.
follow-up time and indicator variable (event/censored).
correlated observations within a patient.
psychological and functional features.
We shall not cover these in any more detail here as each is a large topic in its own right.
Also called “surrogate markers”, “intermediate”, or “replacement endpoint”.
ICH E9 (glossary)
“Surrogate variable: A variable that provides an indirect measurement of effect in situations where direct measurement of clinical effect is not feasible or practical.”
Often measures extent or progress of disease.
Association with clinically meaningful outcome.
Further reading: Prentice RL (1989) Stat Med 8: 431-440.
| Disease | Definitive | Surrogate |
|---|---|---|
| HIV inf. | AIDS (or death) | CD4+ count |
| cancer | mortality | tumor size reduction |
| cardiovascular | stroke, MI | BP, cholesterol |
| glaucoma | vision loss | intraocular pressure |
Taken from Piantadosi (1997), Table 6.3.
Prentice’ Criterion, 1989
“ a surrogate endpoint should capture the dependence between the true endpoint and the treatment arm.”
The true endpoint represents the gold-standard/