Skip to main content

Table 1 Negative articles addressing power/sample size and confidence intervals

From: Prominent medical journals often provide insufficient information to assess the validity of studies with negative results

Journal

Power/Sample Size*

Confidence Intervals†

Power/Sample Size and Confidence Intervals*

 

n (%, 95% CI)

Annals

6/41 (15, 3–26)

33/41 (80, 68–93)

5/41 (12, 2–23)

BMJ

11/57 (19, 9–30)

46/57 (81, 70–91)

10/57 (18, 7–28)

JAMA

10/44 (23, 10–36)

24/44 (55, 39–70)

3/44 (7, 0–15)

Lancet

24/46 (52, 37–67)

34/46 (74, 61–87)

20/46 (43, 29–58)

NEJM

19/46 (41, 27–56)

34/46 (74, 61–87)

13/46 (28, 15–42)

Total

70/234 (30, 24–36)

171/234 (73, 67–79)

51/234 (22, 16–27)

  1. * P < 0.001 † P = 0.038