Table 10 Estimated RRs for use of ECIGs. Results of sensitivity analyses

-

Original estimate

30

1.88 (1.60–2.22)

1

At least four ratio estimates

30

1.88 (1.60–2.32)

At least six ratio estimates

15

2.18 (1.83–2.61)

2

Restricting attention to biomarkers with any slope

86

13.03 (11.91–14.25)

Restricting attention to biomarkers with slope p < 0.05

43

11.63 (8.79–15.40)

Restricting attention to biomarkers with slope p < 0.001

17

2.00 (1.67–2.39)

3

Restricting attention to studies with well-defined groups

22

1.98 (1.46–2.70)

(a) and to biomarkers with any slope

72

2.06 (1.74–2.23)

(b) and to biomarkers with slope p < 0.05

35

1.92 (1.54–2.39)

(c) and to biomarkers with slope p < 0.001

4

2.07 (0.80–5.36)

4

Varying the RR for snuff/snus from 0.80 to 1.00

27

1.92 (1.61–2.29)