Bioequivalence and Bioavailability Forum

Bravo? [General Statistics]

posted by d_labes – Berlin, Germany, 2011-09-02 11:28 (5403 d 09:57 ago) – Posting: # 7324
Views: 8,524

Dear ElMaestro, dear Helmut!

Cough ... :smoke:

The statistics you do here is at least very approximative.

Consider that you evaluate data from a crossover study that are not independent, i.e. the AE counts under Test or Reference may come from the same subject(s).

Keywords (evaluation of incidence of one sort of AEs, f.i. SOC General):
McNemar test if you neglect period effects,
Mainland-Gart or Prescott's test considering period effects.

See for this
Stephen Senn
"Cross-over Trials in Clinical Research"
Chapter 4
Wiley, Chichester 2002

See this online paper for a SAS solution and/or for a description how these tests work.

A test of the whole AE profile from a cross-over I'm not aware of, sorry.

@EM: Full ACK with Helmut's rating of the jargon. Brilliant as always from you :cool:

—
Regards,

Detlew

Complete thread:

Adverse events Dr_Dan 2011-09-01 11:27
- Significant differences in AEs Helmut 2011-09-01 12:18
  - Significant differences in AEs Dr_Dan 2011-09-01 15:44
- Significant differences in AEs ElMaestro 2011-09-01 13:40
- Adverse events ElMaestro 2011-09-01 18:12
  - Bravo! Helmut 2011-09-01 18:45
    - Bravo! Dr_Dan 2011-09-02 09:00
    - Bravo?d_labes 2011-09-02 09:28
      - Bravo? ElMaestro 2011-09-02 18:48
- Adverse events martin 2011-09-02 11:02

Bravo? [General Sta­tis­tics]

Complete thread:

Bravo? [General Statistics]