Likelihood-Ratio-Test Methods for Drug Safety Signal Detection from Multiple Clinical Datasets

Table 2

Summary of basic information of the PPIs study by trials (rr is for PLandPPI vs. PL).

Study	PL			PL and PPI			rr (95% CI)
	N (subject)	Event	Exposure (person-day)	N (subject)	Event	Exposure (person-day)
1st occurrence of AEOST analyses
s = 1	309	14	168947	25	1	10225	1.18 (0.16, 8.97)
s = 2	1647	77	1534190	245	20	181753	2.19 (1.34, 3.58)
s = 3	210	11	133813	20	2	9822	2.48 (0.55, 11.17)
s = 4	1058	167	470833	165	30	56845	1.49 (1.01, 2.19)
s = 5	950	144	713856	32	8	19575	2.03 (0.99, 4.13)
s = 6	150	17	87481	9	3	2884	5.35 (1.57, 18.26)
All s	4324	430	3109120	496	64	281104	1.65 (1.27, 2.14)
Repeated occurrences of AEOST analyses
s = 1	309	30	173796	25	4	10383	2.23 (0.78, 6.33)
s = 2	1647	179	1575058	245	67	196698	3.00 (2.26, 3.97)
s = 3	210	13	136128	20	3	10381	3.03 (0.86, 10.62)
s = 4	1058	228	513109	165	48	64764	1.67 (1.22, 2.28)
s = 5	950	169	789274	32	14	24188	2.70 (1.57, 4.66)
s = 6	150	21	95959	9	5	4266	5.36 (2.02, 14.20)
All s	4324	640	3283324	496	141	310680	2.33 (1.94, 2.79)