Bioequivalence and Bioavailability Forum

Dear Pravin,

searching the forum would have given you the answer in the first thread (which can be found on the forum’s entry page): Power = 1 – β.

All formal decisions are subjected to two types of error:

• α, risk (or error) type I: patient’s risk to be treated with a bioinequivalent formulation (generally 5%).
  
• β, risk (or error) type II: producer’s risk to fail in demonstrating bioequivalence of a “true” bioequivalent formulation (generally 10–20%).

Example from our justice system:
┌────────────────────────────────────────────────┬──────────────────────────────┐
│                                                │          Defendant           │
│                   Verdict                      ├──────────────┬───────────────┤
│                                                │   innocent   │    guilty     │
├────────────────────────────────────────────────┼──────────────┼───────────────┤
│ Presumption of innocence not accepted (guilty) │ Error Type I │    Correct    │
│ Presumption of innocence accepted (not guilty) │    Correct   │ Error Type II │
└────────────────────────────────────────────────┴──────────────┴───────────────┘
In more statistical terms:
┌──────────────────────────────────┬──────────────────────────────┐
│                                  │        Null-Hypothesis       │
│            Decision              ├──────────────┬───────────────┤
│                                  │     True     │     False     │
├──────────────────────────────────┼──────────────┼───────────────┤
│ Null-Hypothesis rejected         │ Error Type I │    Correct    │
│ Failed to reject Null-Hypothesis │    Correct   │ Error Type II │
└──────────────────────────────────┴──────────────┴───────────────┘
In BE-Testing the Null-Hypothesis is that of bioinequivalence!
┌──────────────────────────────────┬────────────────────────────────────┐
│                                  │           Null-Hypothesis          │
│            Decision              ├──────────────────┬─────────────────┤
│                                  │       True       │      False      │
├──────────────────────────────────┼──────────────────┼─────────────────┤
│ Null-Hypothesis rejected         │  Patient's Risk  │  Correct (BE)   │
│ Failed to reject Null-Hypothesis │ Correct (not BE) │ Producer's Risk │
└──────────────────────────────────┴──────────────────┴─────────────────┘

α is a fixed value (generally 0.05, 0.025 for NTIDs in Brazil, or adjusted to ≤0.05 if a sequential design is used). Most people plan a study for 90% power (if all assumptions are fulfilled) and try to stay at 80% (assumptions violated: drop-outs, higher CV, T/R away from 1).