Do alternative methods for analysing count data produce similar estimates? Implications for meta-analyses

Table 3 Simulation results for the low mean (control 0.5, treatment 0.35)

	No overdispersion		Moderate overdispersion		High overdispersion
	RaR^a = 0.71		RaR^a = 0.73		RaR^a = 0.74
	Mean difference^b	SD	Mean difference	SD	Mean difference	SD
Dichotomise RR	0.05	0.08	0.05	0.08	0.06	0.08
Poisson RaR	<0.01	0.06	<0.01	0.01	<0.01	0.01
Simple RaR	<0.01	0.06	<0.01	0.01	<0.01	0.01
Time to the first event HR	0.01	0.08	0.01	0.09	0.02	0.09
Marginal model HR	0.01	0.08	0.01	0.08	0.01	0.08
Andersen-Gill HR	0.01	0.08	0.01	0.08	0.01	0.08
Ratio of means	<0.01	0.03	<0.01	0.03	<0.01	0.03
Ratio of medians	Only possible for 292, 717 and 2087 of the 10,000 simulations, respectively

ISSN: 2046-4053