“…From this, a total of 20 studies were identified for inclusion after the addition of 1 article identified by hand searching (Figure 1). Twelve studies recruited a heterogeneous group of patients, 8 , 20 , 21 , 23 , 26 – 33 3 studies recruited elderly patients, 34 – 36 2 studies each recruited post–liver transplant patients 37 , 38 and cardiac surgery patients, 7 , 22 and 1 recruited patients with acute kidney injury 39 . Fifteen articles 7 , 8 , 20 – 23 , 26 , 27 , 33 – 39 reported relevant clinical outcomes, of which 6 had low risk of bias 20 , 22 , 26 , 34 , 36 , 38 and were used to evaluate the association between malnutrition and ICU clinical outcomes (Table 1).…”