Predictors of clinical trial data sharing: exploratory analysis of a cross-sectional survey

Table 4 Respondents’ reasons for sharing clinical trial data*

	Reasons for sharing clinical trial data*
	Respondents	Administrative requirements (%)	P value	Promote open science (%)	P value	Academic benefits or recognition (%)	P value	Other reasons (%)
Overall	317	55 (17)		248 (78)		133 (42)		56 (18)
Trialist academic productivity (number of research articles published in past three years)			0.64		0.26		0.94
1 to 10 articles	71	10 (14)		58 (82)		31 (44)		8 (11)
11 to 25 articles	117	20 (17)		95 (81)		49 (42)		21 (18)
>25 articles	129	25 (19)		95 (74)		53 (41)		27 (21)
Trialist geographic location			0.02		0.46		<0.0001
United States or Canada	167	32 (19)		135 (81)		52 (31)		35 (21)
Western Europe	113	12 (11)		86 (76)		65 (58)		18 (16)
Other	37	11 (30)		27 (73)		16 (43)		3 (8)
Trial funding source			0.62		0.26		0.28
Government	120	24 (20)		99 (83)		45 (38)		17 (14)
Industry or mixed funding	152	24 (16)		113 (74)		65 (43)		28 (18)
Other	45	7 (16)		36 (80)		23 (51)		11 (24)
Trial size			0.36		0.82		0.48
≤239 subjects	80	18 (23)		64 (80)		33 (41)		14 (18)
240 to 2,016 subjects	158	24 (15)		124 (78)		71 (45)		22 (14)
≥2,017 subjects	19	13 (16)		60 (76)		29 (37)		20 (25)
Journal in which the trial was published			0.27**		0.55		0.46
NEJM	113	17 (15)		87 (77)		46 (41)		22 (19)
Lancet	70	15 (21)		51 (73)		33 (47)		14 (20)
JAMA	43	6 (14)		33 (77)		14 (33)		8 (19)
Annals of Internal Medicine	24	8 (33)		20 (83)		13 (54)		2 (8)
The BMJ	53	8 (15)		46 (87)		20 (38)		7 (13)
PLoS Medicine	14	1 (7)		11 (79)		7 (50)		3 (21)

*If respondent had not been asked to share data they were asked to share their hypothetical reasons for sharing. **Fisher exact test. JAMA, Journal of the American Medical Association; NEJM, New England Journal of Medicine.

ISSN: 1745-6215