Table 4. Variables Associated With Publication of 127 Results From Logistic Regression .

Variable	Odds Ratio (95% CI)	P Value
Efficacy study (vs equivalence)	0.88 (0.26-3.01)	.83
Significant findings (vs not)	1.60 (0.58-4.38)	.36
Sample size (for every log-log unit increase)	1.92 (1.15-3.18)	.01
Parallel design (vs crossover)	1.04 (0.15-7.18)	.97
Full study (vs pilot)	1.13 (0.60-2.98)	.07
Registered (vs not)	13.54 (4.78-38.46)	<.001