Table 2Results of subgroup and sensitivity analyses in randomized trials

All-cause mortalityCardiac mortality
Studies (Patients)RR (95% CI)Meta-regression (P value)Heterogeneity P value (I2%)Studies (Patients)RR (95% CI)Meta-regression (P value)Heterogeneity P value (I2%)
All studies17 (51264)0.95 (0.89, 1.01)0.39 (5)14 (48500)0.89 (0.83, 0.96)*0.63 (0)
Subgroup analyses
Randomization mode clear
    Yes17 (51264)0.95 (0.89, 1.01)0.39 (5)14 (48500)0.89 (0.83, 0.96)*0.63 (0)
    No
Double-blinding
    Yes12 (20636)0.95 (0.89, 1.02)0.740.59 (0)10 (18032)0.91 (0.84, 0.99)*0.240.53 (0)
    No5 (30628)0.93 (0.77, 1.13)0.12 (45)4 (30468)0.82 (0.70, 0.95)*0.75 (0)
Drop-out rates <20%
    Yes11 (38839)0.97 (0.91, 1.03)0.090.45 (0)10 (36418)0.92 (0.84, 0.99)*0.150.65 (0)
    No6 (12425)0.85 (0.76, 0.95)*0.69 (0)4 (12082)0.79 (0.68, 0.93)*0.65 (0)
Population group1
    Group 111 (30788)0.93 (0.88, 0.99)*0.410.60 (0)9 (28144)0.89 (0.83, 0.96)*0.620.49 (0)
    Group 23 (19328)0.90 (0.59, 1.39)0.18 (41)2 (19208)0.86 (0.55, 1.34)0.47 (0)
    Group 33 (1148)0.69 (0.39, 1.23)0.29 (20)3 (1148)0.64 (0.35, 1.18)0.41 (0)
Industry funding
    Yes11 (42744)0.94 (0.84, 1.04)0.570.14 (33)10 (42521)0.89 (0.83, 0.96)*0.920.69 (0)
    No6 (8520)1.01 (0.85, 1.19)0.93 (0)4 (5979)0.80 (0.54, 1.19)0.23 (30)
Performed in the United States
    Yes3 (661)0.72 (0.34, 1.53)0.50.29 (19)3 (661)0.76 (0.36, 1.64)0.740.52 (0)
    No14 (50603)0.95 (0.89, 1.01)0.37 (7)11 (47839)0.89 (0.83, 0.96)*0.50 (0)
Performed in Countries With Background Intake
    Yes4 (20008)0.91 (0.64, 1.29)0.40.23 (31)4 (20008)0.85 (0.57, 1.27)0.920.73 (0)
    No13 (31256)0.93 (0.88, 0.99)*0.52 (0)10 (28492)0.88 (0.80, 0.96)*0.39 (5)
Study size
    Exclude 3 largest studies214 (14321)0.99 (0.86, 1.13)0.750.57 (0)11 (11557)0.83 (0.67, 1.02)0.560.62 (0)
    Only 3 largest studies3 (36943)0.94 (0.85, 1.04)0.09 (59)3 (36943)0.89 (0.82, 0.98)*0.33 (9)
Internal inconsistencies
    Yes1 (240)0.52 (0.28, 0.95)*0.11
    No13 (48260)0.90 (0.83, 0.97)*0.81 (0)
*

p<0.05

1

Group 1: patients with coronary artery diseases, undergoing coronary artery bypass surgery or with recent myocardial infarction; Group 2: patients with intermittent claudication or hypercholesterolemia; Group 3: patients with implantable cardioverter defibrillators.

2

GISSI-Prevenzione, (12-14) GISSI-Heart Failure(15;51;52) and JELIS.(20;53)

From: Results

Cover of Effects of Eicosapentanoic Acid and Docosahexanoic Acid on Mortality Across Diverse Settings: Systematic Review and Meta-Analysis of Randomized Trials and Prospective Cohorts
Effects of Eicosapentanoic Acid and Docosahexanoic Acid on Mortality Across Diverse Settings: Systematic Review and Meta-Analysis of Randomized Trials and Prospective Cohorts: Nutritional Research Series, Vol. 4.
Technical Reviews, No. 17.4.
Trikalinos TA, Lee J, Moorthy D, et al.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.