Table 2.

Grading of Recommendations Assessment, Development and Evaluation (GRADE) evidence profile

Quality assessment						No. of patients		Effect		Quality
No. of studies	Risk of bias	Inconsistency	Indirectness	Imprecision	Other considerations	Omega-3 supplementation	Control	Relative (95% CI)	Absolute (95% CI)
Mortality
17	Not serious	Not serious	Seriousa	Not seriousb	None	149/629 (23.7%)	161/610 (26.4%)	RR 0.85 (0.71 to 1.03)	40 fewer per 1000 (from 8 more to 77 fewer)	Moderate
ICU length of stay
12	Seriousc	Seriousd	Seriousa	Not serious	None	469	456	–	MD 3.79 days fewer (5.49 fewer to 2.09 fewer)	Very low
Duration of mechanical ventilation
7	Seriousc	Seriouse	Seriousa	Not serious	None	254	241	–	MD 2.27 days fewer (4.27 fewer to 0.27 fewer)	Very low

CI confidence interval, RR relative risk, MD mean difference

^aWe rated down the quality of evidence by one level for multiple sources of indirectness. Population: mechanical ventilation and sepsis severity varied as inclusion criteria across studies. Intervention: content of enteral/parenteral formulations differed across studies (10 used fish oil alone while 7 used formulae with additional supplements such as mRNA, arginine, and selenium). Outcome: different mortality definitions (28-day, 60-day, in-hospital, ICU)

^bWe did not rate down the quality of evidence for imprecision. The CI included both significant benefit and small harm, but the number of events was not small

^cWe rated down the quality of evidence by one level for risk of bias. Several studies showed high risk of attrition bias and performance bias

^dWe rated down the quality of evidence by one level for significant unexplained heterogeneity (P < 0.00001, I ² = 82%)

^eWe rated down the quality of evidence by one level for significant unexplained heterogeneity (P = 0.02, I ² = 60%)