Table 6.

Reporting outcomes and precision metrics on harms

	Primary harm 1 (n = 54)	Primary harm 2 (n = 9)	Secondary harm 1 (n = 70)	Secondary harm 2 (n = 33)	Secondary harm 3 (n = 15)
Effect measure
Missing	1 (2)	–	–	2 (6)	–
Mean	19 (35)	4 (44)	12 (17)	8 (24)	2 (13)
Median	7 (13)	–	7 (10)	3 (9)	2 (13)
Percentage	8 (15)	1 (11)	7 (10)	5 (15)	2 (13)
Absolute number	3 (6)	–	12 (17)	7 (21)	3 (20)
Absolute number + percentage	15 (28)	3 (33)	30 (43)	8 (24)	5 (33)
Cumulative incidence	1 (2)	1 (11)	–	–	–
Absolute number + mean	–	–	1 (1)	–	–
Ratio	–	–	1 (1)	–	–
Rate/100 patient‐years	–	–	–	–	1 (7)
Precision measure of effect		–
Missing	23 (43)	4 (44)	46 (66)	21 (64)	10 (67)
P value	1 (2)	–	1 (1)	1 (3)	1 (7)
95 per cent c.i.	6 (11)	1 (11)	6 (9)	1 (3)	‐
s.d.	13 (24)	4 (44)	9 (13)	6 (18)	1 (13)
i.q.r.	1 (2)	–	2 (3)	–	1 (7)
Range	5 (9)	–	4 (6)	4 (12)	1 (7)
P value and range	1 (2)	–	1 (1)	–	–
s.e.m.	3 (6)	–	1 (1)	–	–
Difference measure
Missing	4 (7)	1 (11)	19 (27)	10 (30)	4 (27)
Risk ratio	4 (7)	1 (11)	1 (1)	–	–
Hazard ratio	7 (13)	–	2 (3)	1 (3)	–
Odds ratio	4 (7)	1 (11)	2 (3)	1 (3)	–
Difference in mean	18 (33)	4 (44)	13 (19)	8 (24)	2 (13)
Difference in percentage	7 (13)	–	26 (37)	9 (27)	6 (40)
Difference in median	7 (13)	–	5 (7)	3 (9)	2 (13)
Risk difference	3 (6)	1 (11)	1 (1)	1 (3)	–
Difference in cumulative incidence	–	1 (11)	–	–	–
Effect size	–	–	1 (1)	–	–
Difference in rate/100 patient‐years	–	–	–	–	1 (7)
Precision measure of difference
Missing	3 (6)	1 (11)	11 (16)	5 (15)	2 (13)
P value	29 (54)	4 (44)	51 (73)	25 (76)	13 (87)
95 per cent c.i.	6 (12)	–	1 (1)	–	–
P value and 95 per cent c.i.	14 (26)	4 (44)	6 (9)	3 (9)	–
90 and 95 per cent c.i.	1 (2)	–	1 (1)	–	–
P value, 95 per cent c.i. and number needed to treat	1 (2)	–	–	–	–

Values in parentheses are percentages.