Table 1 Overview of studies investigating nutritional protocols in cardiac surgical patients

Rahman 2016 [48]

POS

Patients (n = 787) after cardiac surgery

No information

60.1% of patients received artificial nutrition support. Of these patients, 78% received enteral nutrition (EN) alone, 17% received a combination of EN and parenteral nutrition (PN), and 5% received PN alone. The remaining 314 patients (40%) received no nutrition

The mean (SD) time from ICU admission to EN initiation was 2.3 (1.8) days. The adequacy of calories was 32.4% ± 31.9% from EN and PN and 25.5% ± 27.9% for patients receiving only EN. In EN patients, 57% received promotility agents and 20% received small bowel feeding. There was no significant relationship between increased energy or protein provision and 60-day mortality. Postoperative cardiac surgery patients who stay in the ICU for 3 or more days are at high risk for inadequate nutrition therapy

Flordelis Lasierra 2015 [49]

POS

Cardiac surgery patients (n = 37) with hemodynamic failure requiring more than 24 h of mechanical ventilation

No information

Early EN protocol of institution

11 patients (29.7%) required mechanical circulatory support, and 25 (68.0%) met the criteria for early multiorgan dysfunction. Mortality was 13.5%. Mean EN duration was 12.3 days (95% confidence interval (CI) 9.6–15.0). The mean EN diet volume delivered/patient/day was 1199 mL (95% CI 1118.7–1278.8), and mean EN energy delivered/patient/day was 1228.4 kcal (95% CI 1145.8–1311). The set energy target was achieved in 15 patients (40.4%). The most common EN-related complication was constipation. No case of mesenteric ischemia was detected. Early EN is feasible in cardiac surgery patients and not associated with serious complications, but appropriate energy target by EN alone cannot be attained

Visser 2014 [50]

RCT

Patients (n = 33) undergoing off-pump coronary artery bypass grafting

Off-pump

Enteral, parenteral, or no nutrition (control) from 2 d before, during, and until 2 d after surgery

The myocardial arginine:ADMA ratio increased during surgery and was significantly higher in the enteral and parenteral groups than in the control group (median (IQR) 115.0 (98.0–142.2) (P = 0.012), 116.9 (100.3–135.3) (P = 0.004), and 93.3 (82.7–101.1)). The change in the preoperative to postoperative plasma arginine:ADMA ratio correlated with the change in myocardial glucose metabolism in positron emission tomography (r = 0.427, P = 0.033). Enteral or parenteral nutrition before, during, and after CABG may positively influence myocardial glucose metabolism by increasing the plasma and myocardial arginine:ADMA ratio

Umezawa Makikado 2013 [51]

POS

Cardiac surgery adult patients (n = 7) receiving VA ECMO for severe hemodynamic failure unresponsive to conventional therapies

No information

ICU EN Protocol (25 kcal/kg, to be reached over 4 days)

Two patients received EN within the first 24 h, and all patients were on EN by 48 h. More than 70% nutritional therapy (NT) was achieved within the first week in all cases. After 2 weeks, only 3 of 7 patients continued on ECMO, and a decrease in NT was noted. By day 7, the mean energy balance was –245.99 (range –75.82 to –555) kcal/day and was –373 (range –75.82 to –795) kcal/day at the end of the study. Mean cumulated deficit by day 7 was –603.54 (range –75.82 to –1721.92) kcal/day and was –3264.94 (range –75.82 to –7087.08) kcal/day at the end of the study. No serious adverse events were attributable to EN, such as aspiration of gastric contents, abdominal distension, ileus, vomiting, gastrointestinal bleeding, or bowel ischemia. Four patients met the criteria for constipation. Ventilator-associated pneumonia was diagnosed in one patient. One patient presented with a catheter-associated bloodstream infection Early EN is possible and safe in patients with severe hemodynamic failure receiving VA ECMO.

Tepaske 2007 [42]

RCT

Patients (n = 49) scheduled to undergo cardiac surgery with the use of extracorporeal circulation, aged 70 years or older, had a compromised left ventricular function, or were planned for mitral valve surgery

On-pump

Addition of glycine to a standard preoperative oral immune-enhancing nutrition supplement (OIENS)

Infectious morbidity was significantly lower in both treatment groups compared with the control group (P = 0.02). Less supportive therapy was necessary to stabilize circulation in both treatment groups compared with the control group. Median length of hospital stay was 7.0, 6.5, and 8.0 days in the OIENS + glyc, OIENS, and control groups, respectively. Inflammatory responses, as measured by systemic levels of pro-inflammatory cytokines and surface markers on polymorphonuclear cells were comparable for all groups. Preoperative OIENS reduces postoperative infectious morbidity and results in more stable circulation; the addition of glycine does not result in any beneficial effect over standard OIENS

Rapp-Kesek 2007 [52]

RCT

Patients (n = 16) undergoing cardiac surgery

No information

Enteral nutrition and dopexamine

Dopexamine and enteral nutrition caused no adverse effects on oxygen consumption or the oxygen extraction ratio. Enteral nutrition did not increase the splanchnic blood flow or cardiac index. Dopexamine increased the systemic blood flow with only a transient effect on the splanchnic blood flow. Dopexamine increased the lactate concentration, possibly indicating a more ischemic condition.

Berger 2005 [53]

POS

Patients (n = 70) after cardiac surgery with extracorporeal circulation, staying 5 days in the ICU, with acute cardiovascular failure

On-pump

Enteral nutrition

Forty patients required artificial nutrition. Energy delivery was very variable. There was no abdominal complication related to EN. As a mean, 1360+/-620 kcal/kg/day could be delivered enterally during the first 2 weeks, corresponding to 70+/-35% of energy target. Enteral nutrient delivery was negatively influenced by increasing dopamine and norepinephrine doses, but not by the use of IABP. EN is possible in the majority of patients with severe hemodynamic failure, but usually results in hypocaloric feeding. EN should be considered in patients with careful abdominal and energy monitoring

Sustić 2005 [54]

RCT

Patients (n = 40) after coronary artery bypass grafting

No information

Early EN and effects of metoclopramide

Enteral feeding with isoosmotic enteral formula was initiated by nasogastric tube 18 h after surgery. After 6 h, feeding was stopped, and paracetamol solution (1000 mg) and 10 mg of metoclopramide IV or 2 ml of saline IV were concurrently administered. The plasma paracetamol concentrations 15, 30, 60, and 120 minutes after the administration of paracetamol were significantly higher in metoclopramide group than in control group: (t(+15)) 5.4-/+2.7 vs 3.3-/+2.5 (Mann-Whitney U test; P = 0.017); (t(+30)) 6.7-/+2.4 vs 3.7-/+2.0 (P = 0.006); (t(+60)) 7.7-/+2.5 vs 5.1-/+3.2 (P = 0.008); (t(+120)) 8.5-/+2.2 vs 5.2-/+2.8 (P = 0.005). The AUC value was 34% larger in the metoclopramide group vs control group (574-/+296 vs 429-/+309; P = 0.027). There were no significant differences in gallbladder ejection fraction between groups (group metoclopramide vs control group: (t(0)-t(+15)) –2% vs –2%; (t(+15)-t(+30)) 1% vs 4%; (t(+30)-t(+60)) 0% vs –1%; (t(+60)-t(+120)) 1% vs 3%; P = NS). In CABG surgery patients with early enteral feeding, a single dose of intravenous metoclopramide effectively improves gastric emptying, but does not have any prokinetic effect on gallbladder motility

Hartwell 2003 [55]

POS

Patients (n = 15) after coronary artery bypass grafting

No information

Dietary advice

Dietary intake was assessed on three occasions (preoperatively, 2 months after surgery and 1 year after surgery) by use of a food amount frequency questionnaire that had been previously validated. Patients were also asked to provide information on any dietary advice they had received. The absolute mean intakes of total fat, saturated fat and dietary cholesterol significantly increased 1 year after CABG surgery by 21%, 36% and 51%, respectively, and the choice of food items reflected this change in nutrient intake. These undesirable changes occurred despite the provision of dietary advice.

Kesek 2002 [56]

POS

Patients (n = 73) scheduled for coronary artery bypass grafting and/or valvular disease, thoracic or thoraco-abdominal aortic aneurysms or other combined procedures

No information

Enteral nutrition

In 59/73 patients, EN was started within 3 days. EN was discontinued in half of the patients when they were able to feed themselves. Twelve patients vomited, one of them severely. Dislocation of the nasogastric tube occurred in 28 patients. The 15 patients with diarrhoea were treated with 2–6 broad-spectrum antibiotics during their ICU stay. Out of 73 patients, 40 did not show any gastric residual volume (GRV). GRV decreased during EN in 50% of the patients with fairly large or large residual volumes. The incidence of aspiration pneumonia was 10%. In the cardiothoracic ICU, individually adjusted early EN is feasible with few problems

Tepeske 2001 [57]

RCT

Patients (n = 50) scheduled for coronary artery bypass grafting

No information

Oral immune-enhancing nutritional supplementation containing L-arginine, omega3 polyunsaturated fatty acids, and yeast RNA (n = 25), or a control (n = 25) for a minimum of 5 days

Preoperative expression of HLA-DR epitopes on monocytes was significantly higher in patients given the study treatment (109% (95% CI 92–128)) than those given the control (69% (58–82)) compared with baseline (100%) (P = 0.02, repeated measures ANOVA). However, concentration of interleukin 6 was significantly lower in the treatment group (0.90 pg/L (0.69–1.18)) than in the control group (1.94 pg/L (1.45–2.59)) (P = 0.032, repeated measures ANOVA). Additionally, delayed-type hypersensitivity response to recall antigens improved preoperatively and remained better until hospital discharge

Revelly 2001 [58]

POS

Patients (n = 9) 1 day after cardiac surgery under cardiopulmonary bypass

On-pump

Isoenergetic EN via a post pyloric tube while catecholamine treatment remained constant. Baseline (fasted) condition was compared to continuous EN condition

Cardiac index (CI), mean arterial pressure (MAP), pulmonary and wedge pressures, indocyanine green (ICG) clearance, gastric tonometry, plasma glucose and insulin, and glucose turnover (6,62H2-glucose infusion) were determined repetitively every 60 minutes during 2 h of baseline fasting condition and 3 h of EN. During EN, CI increased (from 2.9 +/- 0.5 to 3.3 +/- 0.5 l min-1 m-2), MAP decreased transiently (from 78 +/- 7 to 70 +/- 11 mmHg), ICG clearance increased (from 527 +/- 396 to 690 +/- 548 ml/min), and gastric tonometry remained unchanged, while there were increases in glucose (158 +/- 23 to 216 +/- 62 mg/dl), insulin (29 +/- 23 to 181 +/- 200 mU/L), and glucose rate of appearance (2.4 +/- 0.2 to 3.3 +/- 0.2 mg min-1 kg-1). The introduction of EN in these postoperative patients increased CI and splanchnic blood flow, while the metabolic response indicated that nutrients were utilized. These preliminary results suggest that the hemodynamic response to early EN may be adequate after cardiac surgery even in patients requiring inotropes

Berger 2000 [28]

POS

Cardiac patients (n = 39)

No information

Early EN in patients with adequate hemodynamic status, and in patients with hemodynamic failure

Absorption was strongly reduced on day 1 in all patients after gastric administration (lower peak paracetamol and AUC), but normal after post pyloric delivery. Duration of anesthesia and of circulatory bypass did not affect paracetamol absorption. On day 3, AUC was close to normal in the case of hemodynamic failure. Peak absorption on day 1 was negatively correlated with opiate dose (r2 = 0.176, P = .008). Hypocaloric enteral nutrition was well-tolerated

Fiaccadori 1994 [59]

POS

Cardiac patients undergoing mitral valve replacement (n = 12)

On-pump

A lipid emulsion containing 10% medium-chain triglycerides (MCT) and 10% long-chain triglycerides (LCT) was infused at a rate of 1 ml/kg/h (3.3 mg/kg/min) for 2 h to 24 h after surgery

Fat emulsions containing both MCT and LCT, when given at 3.3 mg/kg/min for 120 minutes following valvular heart surgery, do not exert negative cardiopulmonary effects and could represent a source of rapidly metabolized substrates

Behrendt 1984 [60]

RCT

Cardiac patients undergoing CABG (n = 40)

No information

Patients were treated with 3 types of hypocaloric parenteral nutrition consisting of the same content of amino acids (75 g/day) but different amounts of carbohydrates: 120 g dextrose, 120 or 200 g dextrose/fructose/xylitol (1:2:1) per day; 18 patients received only 75 g/day of dextrose (controls)

No beneficial effects on the concentrations of visceral serum proteins (transferrin, pre-albumin, retinol binding protein, cholinesterases) were found

Weidler 1984 [61]

RCT

Cardiac patients (n = 20)

No information

Both groups received identical carbohydrate calories (2000 kcal/day) group 1 received only essential amino acids while in group 2 a combined pattern of essential and non-essential amino acids were infused

Routine parameters evaluating protein metabolism in both groups did not differ significantly. Nitrogen balance was positive only in group 2; the difference between the two groups concerning other parameters was minimal

Haider 1981 [62]

RCT

Patients (n = 22) after open heart surgery

On-pump

Control group received Ringer lactate as postoperative infusion (RL-group), the 2nd group was given 50% glucose (0.5 g/kg/h) and insulin (250 U/1000 cc) (GI-group)

In contrast with the RL-group in the GI-group there was a significant decrease in FFA-serum level and cAMP serum level, which developed during the infusion. Urine output and urinary glucose excretion was nearly equal in both groups. Urinary potassium excretion in the GI-group remained significantly one third lower than that of the RL-group, in spite of the potassium supply to the GI-group being nearly twice as much and serum potassium level approximately equal. Urinary sodium excretion in the GI-group on the other hand was approximately 15% higher than that in the RL-group. In relation to preoperative values, postoperative urinary N-excretion in the GI-group was unchanged, whereas in the RL-group the postoperative N-excretion was significantly 30% increased; in postoperative alpha-amino-N-excretion there was only a small difference between the groups, which indicates an insulin-modifiable increase in protein breakdown rather than decreased protein synthesis

Fisch 1981 [63]

RCT

Patients (n = 13) following cardiopulmonary bypass procedures

No information

Two intravenous fat emulsions (Liposyn 10%, Abbott Laboratories, North Chicago, IL, USA and Intralipid 10%, Cutter Laboratories, Berkeley, CA, USA)

Neither intravenous fat emulsion was observed to exert significant changes in left ventricular stroke work, left ventricular filling pressure, cardiac output, systemic vascular resistance, mean systemic arterial blood pressure, central venous pressure, or mean pulmonary artery pressure. This study confirmed that the administration of 10% fat emulsions available in the USA does not exert significant untoward hemodynamic changes, even in patients with severe cardiac impairment recovering from recent open-heart surgery

Abel 1976 [64]

RCT

Malnourished (n = 44) and non-malnourished adult patients (n = 20) after cardiac surgery

No information

Parenteral nutrition compared to routine postoperative infusions for 5 days or until oral intake was deemed adequate

The overall mortality in the malnourished group was 12.5%, in the malnourished patients receiving parenteral nutrition it was 20% and in the non-malnourished group it was 0%. Malnutrition itself appears to significantly increase morbidity and mortality following cardiac surgery. Although total parenteral nutrition can safely be administered in the early postoperative period following open heart surgery, an average of 5 days of prophylactic parenteral nutrition did not appear to be effective in lessening the untoward effects of the malnourished state