Restrictive and liberal red cell transfusion strategies in adult patients: reconciling clinical data with best practice

Table 2 Strength of association between red cell transfusion and purported clinical adverse effect

	Highest-level clinical studies
Adverse effect	Showing correlating effect	References	Showing no correlation	References
Multisystem organ dysfunction	Observational studies	[11,12]	Level 1 RCT	[3,8]
Nosocomial infection	Observational and retrospective studies	[11,13]	Level 1 RCT (2), RCT (1), RCT meta-analysis	[1,4,5,8,9]
Allergic or immunomodulation, tumor promotion	Observational and retrospective studies	[14-16]	Level 1 RCT (2)	[8,9]
Pulmonary edema	Level 1 RCT	[3,9]	Level 1 RCT (3), RCT meta-analysis	[1,5,7-9]
Pulmonary (non-edema) including ARDS	Observational studies	[11,17]	Level 1 RCT (3), RCT (1)	[4,5,7-9]
Acute kidney injury	Observational and retrospective studies	[18,19]	Level 1 RCT (2), RCT (1)	[4,5,8,9]
Myocardial ischemia	Prospective cohort	[20,21]	Level 1 RCT (4), RCT (1)	[3-5,7-9]
Cerebral ischemia	Observational and retrospective studies	[10,20,22]	Level 1 RCT (3), RCT (1), RCT meta-analysis	[1,4,5,7-9]
Shock	Observational study	[11]	Level 1 RCT	[3,5,8]
Cardiac arrest	Prospective cohorts	[20,21]	Level 1 RCT	[3,5,7,8]
Bleeding/coagulopathy	Observational study	[11]	Level 1 RCT	[5,8]

Highest-level study reflects the rank order of scientific merit typically afforded to studies based on trial design. Highest to lowest: prospective RCTs, prospective subset analyses of randomized studies, nonrandomized controlled trials, observational case series (including prospective and retrospective cohort analysis), and consecutive and nonconsecutive case series. ARDS, acute respiratory distress syndrome; RCT, randomized clinical trial.

ISSN: 1364-8535