Skip to main content
. 2009 May 12;338:b1732. doi: 10.1136/bmj.b1732

Table 2.

 Reporting of parameters required for a priori sample size calculation for the 215 articles

Parameter Reporting frequency (%)
α risk 191 (93)
 0.05 183 (96)
  Two tailed test 119 (65)
  One tailed test 7 (4)
  Unspecified 57 (31)
 0.025 for one tailed test 2 (1)
 Adapted for interim analyses 6 (3)
Power 200 (98)
 80% 107 (54)
 85% 9 (5)
 90% 66 (33)
 95% 4 (2)
 Other values 14 (7)
Assumptions for control group 165 (81)
 Justification of assumptions 81 (49)
  Results from previous trial 54 (67)
  Preliminary study 15 (19)
  Observational data 6 (7)
  Results of systematic review 2 (3)
  Others 4 (5)
Assumptions for the treatment effect 186 (91)
 Justification of the assumptions 50 (27)
  Analogy to another trial or treatment 41 (82)
  Clinical relevance 7 (14)
  Observational data 1 (2)
  Results of a meta-analysis 1 (2)
All parameters required for sample size calculation 113 (53)