Table 3: Change in trial quality over time and odds of a positive trial relative to trial quality.

Quality itemQuality item reported in trial
N (%)
Association between reporting of quality item and publication year
OR per year post 1979 (95% CI), p
Association between reporting of quality item and likelihood of a positive trial
OR (95% CI), p

Risk of bias
 Allocation concealment68 (35.1)1.06 (1.00, 1.14), 0.092.04 (0.67, 6.16), 0.21
 Sequence generation58 (29.9) †1.0 (0.85, 1.17), 0.993.27 (0.27, 40.2), 0.35
 Detection bias addressed90 (46.4) ‡1.02 (0.96, 1.08), 0.550.83 (0.36, 1.91), 0.67
Intervention dose
 Dose schedule report148 (76.3)1.06 (1.01, 1.10), 0.011.52 (0.76, 3.09), 0.24
 Dose intensity report69 (35.6)1.06 (1.01, 1.12), 0.020.49 (0.26, 0.92), 0.03
 Dose justified36 (18.6)1.01 (0.95, 1.06), 0.890.59 (0.28, 1.26), 0.17
Trial design
 Hypothesis stated71 (36.6)1.09 (1.03, 1.15), 0.0021.33 (0.73, 2.46), 0.36
 Primary outcome stated49 (25.3)1.16 (1.08, 1.26), <0.0011.22 (0.61, 2.45), 0.56
 Trial phase stated15 (7.7)1.09 (0.98, 1.20), 0.120.70 (0.40, 1.24), 0.22
 Registration of trial10 (5.2)3.32 (1.58, 6.95), 0.0010.61 (0.37, 5.49), 0.61
 Use of committees1 (0.5)1.33 (0.68, 2.63), 0.41N/A §
 Study sites specified124 (63.9)1.01 (0.96, 1.05), 0.870.99 (0.53, 1.84), 0.97
 Biological rationale of intervention stated60 (30.9)1.11 (1.04, 1.17), 0.0012.18 (1.14, 4.19), 0.02
 Compliance with intervention reported69 (35.6)1.05 (1.01 1.10), 0.0481.33 (0.71, 2.49), 0.37
 Actual intervention received reported22 (11.3)1.14 (1.03, 1.27), 0.021.48 (0.57, 3.84), 0.42
 Adverse events reported22 (11.3)1.09 (1.01, 1.19), 0.071.42 (0.56, 3.62), 0.46

OR= odds ratio.
Logistic regression adjusted for number of outcomes and sample size.
†Out of 113 trials that were scored in the corresponding Cochrane review.
‡Out of 179 trials that were scored in the corresponding Cochrane review.
§As "design committee= 1" predicted positive trial perfectly, odds ratio cannot be calculated.