Term
|
Definition
Answers: does the clinciain want to know how to initiate therapy?
Avoids ALPHA errors
any treatment changes after randomization are ignored and all outcomes are charged to the orignially assigned therapy |
|
|
Term
|
Definition
answers: does the clinician want to know what to expect after therapy has begun?
Avoids BETA errors
Subjects with treatment changes are removed or excluded |
|
|
Term
|
Definition
E/C
Will overestimate or underestimate impact of therapy when adverse events in untreated patients are very rare |
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
| Harm: relative risk value |
|
Definition
|
|
Term
| Therapy: relative risk value |
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
range withing which the true effect plasibly lies greater
the greater the width of CI, the less precise the observed estimate
(The width also increases if we raise from 90% to 95% to 99% so don't be fooled by smaller CI!) |
|
|
Term
|
Definition
| evalutes the point estimate that provides the single most likely estimate of the underlying true treatment |
|
|
Term
|
Definition
| leads to either an underestimate or overestimate of the underlying beneficial or harmful effect |
|
|
Term
|
Definition
| Proportion of patients in a group in whom an event is observed |
|
|
