[No conflicts of interest]

Transcription

1 [No conflicts of interest]

2 Patients and staff at:

3

4 Available evidence pre-calories Three meta-analyses: Gramlich L et al. Does enteral nutrition compared to parenteral nutrition result in better outcomes in critically ill adult patients? A systematic review of the literature. Nutrition 2004;20: Heyland DK et al. Canadian clinical practice guidelines for nutrition support in mechanically ventilated, critically ill adult patients. JPEN 2003;27: Simpson F, Doig GS. Parenteral vs. enteral nutrition in the critically ill patient: a meta-analysis of trials using the intention to treat principle. Intensive Care Med 2005;31: Conflicting evidence (and guidelines) regarding optimal route Within a context of concern about malnutrition in NHS hospitals

5 Commissioning brief (07/52)

6 Available evidence pre-calories Three meta-analyses: Gramlich L et al. Does enteral nutrition compared to parenteral nutrition result in better outcomes in critically ill adult patients? A systematic review of the literature. Nutrition 2004;20: Heyland DK et al. Canadian clinical practice guidelines for nutrition support in mechanically ventilated, critically ill adult patients. JPEN 2003;27: Simpson F, Doig GS. Parenteral vs. enteral nutrition in the critically ill patient: a meta-analysis of trials using the intention to treat principle. Intensive Care Med 2005;31: Conflicting evidence (and guidelines) regarding optimal route Concern about malnutrition in NHS hospitals Simpson & Doig Significant benefit favouring parenteral nutrition OR (mortality) 0.51 (0.27 to 0.97)

7 Hypothesis that, in critically ill adults, delivery of early nutritional support through the parenteral route is superior to delivery through the enteral route

8 Design A phase III, open, multicentre, pragmatic* randomised controlled trial with an integrated economic evaluation Nested in the national clinical audit database - Case Mix Programme reduce data collection burden enhance generalisability * does this intervention work under usual conditions?

9 Participating sites 33 adult general critical care units, England In a representative sample of NHS hospitals

10 Inclusion criteria sites Active participation in the national clinical audit Pre-existing, established protocols for nutritional support via both routes (and reflecting mainstream practice) Pre-existing implementation of bundles (as promoted by the NHS) to prevent development of: bloodstream infection ventilator-associated pneumonia thromboembolic events Glycaemic control in accordance with international guidelines (<180 mg/dl)

11 Inclusion criteria patients Adult (18+ years) Unplanned admission Expected to receive nutritional support for two or more days Not planned to be discharged within three days

12 Exclusion criteria - patients Pre-existing contraindication to either PN or EN Received PN/EN within the last 7 days PEG, PEJ, needle/surgical jejunostomy in situ Known to be pregnant Admitted for thermal injury (burns) Admitted for palliative care Been in a critical care unit for >36 hours Previously randomised into CALORIES Expected stay in UK less than 6 months

13 Timeline Admission to critical care Eligibility criteria met Informed patient consent, or Consultee agreement Up to 36 hours Randomise Start nutritional support

14 Delivery of nutritional support Randomisation - 24 hour telephone Parenteral route via CVC (dedicated lumen) Enteral route via nasogastric/ nasojejunal tube Followed exclusively for 5 days (120 hours) - target of 25 kcal/kg/day - goal of reaching target within 48 to 72 hours Exclusive oral feeding, unit discharge or death All other care during/after 5 days at discretion of treating clinician(s) If clinically indicated, then oral feeding could be started within the 5 days Enteral trickle feeding not permitted during the 5 days

15 Delivery of early nutritional support Local feeding protocols used All protocols reviewed by trial dietitians - to ensure common boundaries Standard feed within following ranges energy total kcal per bag nitrogen g per bag Additional micronutrients (if clinically indicated/prescribed) Adjust for obese patients, as required Adjust volume for fluid balance requirements

16 Sample size calculation Baseline 30-day mortality 32% Detectable difference RRR 20% (ARR 6.4%) Type I error rate (P value) P<0.05 Type II error rate (power) 90% power 1082 per arm Crossovers/exclusions 2% crossover each way 2% withdrawal (based on PAC-Man) 1200 per arm One planned interim analysis on first 1200 patients

17 Cumulative recruitment Cumulative recruitment N= Jul Jan Jul Jan Jul Jan 2014

18

19

20 Baseline characteristics Case mix Parenteral route Enteral route Age (years), mean (sd) 63.3 (15.1) 62.9 (15.4) Gender, males (%) 689 (57.9%) 725 (60.6%)

21 Baseline characteristics Severe comorbidities Severe comorbidities (APACHE II definitions), n (%) Parenteral route Enteral route Liver 29 (2.5%) 34 (2.8%) Renal 20 (1.7%) 15 (1.3%) Respiratory 34 (2.9%) 23 (1.9%) Cardiovascular 11 (0.9%) 14 (1.2%) Immunocompromised 78 (6.6%) 95 (8.0%)

22 Baseline characteristics Surgical status and mechanical ventilation Surgery within 24 hours prior to randomisation, n (%) Parenteral route Enteral route 162 (13.6%) 167 (14.0%) Mechanical ventilation during 24 hours following critical care admission, n (%) 979 (83.1%) 993 (83.8%)

23 Baseline characteristics Acute severity of illness APACHE II Acute Physiology Score, mean (sd) Parenteral route Enteral route 15.1 (6.2) 15.2 (6.2) APACHE II Score, mean (sd) 19.6 (6.9) 19.6 (7.0) APACHE II predicted risk of death, median (IQR) ICNARC Physiology Score, mean (sd) ICNARC model predicted risk of death, median (IQR) 0.34 ( ) 0.34 ( ) 25.6 (8.0) 25.8 (7.8) 0.42 ( ) 0.43 ( ) SOFA score, mean (sd) 9.5 (3.4) 9.6 (3.3)

24 Baseline characteristics Nutritional status Subjective assessment of severe malnutrition, n (%) Parenteral route Enteral route 151 (12.7%) 152 (12.7%) BMI, mean (sd) 27.7 (7.4) 28.2 (7.5) Degree of malnutrition, n (%) High (BMI<18.5 or weight loss>20%) Moderate (BMI<20 and weight loss>5%) 74 (6.4%) 81 (7.0%) 8 (0.7%) 10 (0.9%) No malnutrition 1070 (92.9%) 1070 (92.2%)

25 Non-adherence

26 Non-adherence Received nutritional support via assigned route and changed during first 120 hours 40 Parenteral route Enteral route Number of patients Time to crossover (hours)

27 Non-adherence Received no nutritional support for at least one full day during first 120 hours 20 Parenteral route Enteral route Number of Patients Days from Initiation of Nutritional Support

28 Clinical management Timing of nutritional support Time from original critical care admission to initiation of nutritional support (hours), median (IQR) Parenteral route Enteral route 24 (17-30) 22 (16-28)

29 Nutritional support

30 Clinical management Daily calories (in relation to the target of 25kcal/kg/day) D Percentage met Target (25kcal/kg/day) Parenteral route Parenteral route (adjusted) Enteral route Enteral route (adjusted) Days from Initation of Early Nutritional Support Adjusted for part days of feeding (e.g. death of a patient)

31 Clinical management Additives Patients receiving additives during the intervention period*, n (%) Parenteral route Enteral route Glutamine 50 (4.2%) 0 (0.0%) Selenium 268 (22.5%) 3 (0.3%) Fish oils 38 (3.2%) 1 (0.1%) * Only recorded for patients receiving nutritional support via the parenteral route

32 Clinical management Aspirates and put-back Total aspirates during intervention period* (ml), mean (sd) Parenteral route Enteral route 35 (265) 958 (1312) Total aspirates put-back during intervention period* (ml), mean (sd) 24 (170) 618 (863) * Only recorded for patients receiving nutritional support via the enteral route

33 Clinical management Prokinetics, vasoactives, insulin and blood glucose Patients receiving prokinetics during intervention period*, n (%) Parenteral route Enteral route 26 (2.2%) 426 (35.6%) Patients receiving vasoactives during intervention period, n (%) Patients receiving insulin during the intervention period, n (%) Total insulin received during intervention period (IU), mean (sd) 958 (80.9%) 1007 (84.6%) 694 (58.6%) 668 (56.1%) 154 (223) 122 (204) * Prokinetics - only recorded for patients receiving nutritional support via the enteral route

34 Clinical management Blood glucose Mean daily lowest blood glucose during intervention period (mg per dl), mean (sd) Parenteral route 127 (25) [7.0 (1.4)] Enteral route 118 (26) [6.5 (1.4)] Mean daily highest blood glucose during intervention period (mg per dl), mean (sd) 183 (43) [10.2 (2.4)] 181 (45) [10.0 (2.5)] [mmol per L]

35 Clinical management Diarrhoea and constipation Incidence of diarrhoea during intervention period, n (%) Parenteral route Enteral route 192 (16.2%) 250 (21.0%) Incidence of constipation during intervention period, n (%) 726 (61.3%) 643 (54.0%) Diarrhoea loose/unformed or liquid stools on 3 consecutive days Constipation - bowels not opened on 3 consecutive days

36 Primary outcome All-cause mortality at 30 days (from randomisation) Parenteral route, deaths (%) Enteral route, deaths (%) P-value 393 (33.1%) 409 (34.2%) 0.57 Effect estimate (95% CI) [P-value] Absolute risk reduction 1.15 (-2.65 to 4.94) Relative risk 0.97 (0.86 to 1.08) Adjusted* odds ratio 0.95 (0.79 to 1.13) [p=0.55] *Adjusted for age, ICNARC physiology score, surgical status, degree of malnutrition and a unit-level random effect

37 Secondary outcomes All-cause mortality at 90 days (from randomisation) Parenteral route, deaths (%) Enteral route, deaths (%) P-value 442 (37.3%) 464 (39.1%) 0.40 Effect estimate (95% CI) [P-value] Relative risk 0.96 (0.86 to 1.06) Adjusted* odds ratio 0.93 (0.77 to 1.11) [p=0.39] *Adjusted for age, ICNARC physiology score, surgical status, degree of malnutrition and a unit-level random effect

38 Secondary outcomes Duration of survival to 90 days Proportion of Patients Alive Parenteral route Enteral route Unadjusted (log rank test): P = 0.98 Adjusted* (Cox proportional-hazards model): Hazard ratio (95% CI) = 0.94 (0.82 to 1.07) P = Days since Randomization *Adjusted for age, ICNARC physiology score, surgical status, degree of malnutrition and a unit-level random effect

39 Secondary outcomes Duration of organ support in the unit Days alive and free (up to 30 days) of Parenteral route, mean (SD) Enteral route, mean (SD) Difference in means (95% CI) P-value Advanced respiratory 14.3 (12.1) 14.3 (12.2) 0.04 (-0.94 to 1.01) 0.94 Advanced cardiovascular 18.9 (13.5) 18.5 (13.6) 0.41 ( 0.63 to 1.53) 0.44 Renal 19.1 (13.9) 18.8 (14.0) 0.26 ( 0.85 to 1.47) 0.66 Neurological 19.2 (13.8) 18.9 (14.0) 0.34 ( 0.81 to 1.36) 0.57 Gastrointestinal 13.0 (11.7) 13.2 (11.8) 0.12 ( 1.05 to 0.80) 0.81

40 Secondary outcomes Number of new infectious complications per patient Parenteral route, mean (SD) Enteral route, mean (SD) Difference in means (95% CI) P-value Number of infectious complications per patient 0.22 (0.60) 0.21 (0.56) 0.01 (-0.04 to 0.06) 0.72 Either laboratory confirmed or strongly suspected and treated with antibacterials or antifungals (85% laboratory confirmed in both groups) Occurring from second day of nutritional support to discharge from critical care unit

41 Secondary outcomes Infectious complications Infectious complications Chest infection (pneumonia/ lower respiratory tract infection) Parenteral route, n (%) Enteral route, n (%) P-value 135 (11.3%) 143 (11.9%) 0.66 CVC infection 11 (0.9%) 9 (0.8%) 0.66 Other vascular catheter-related infection 4 (0.3%) 3 (0.3%) 0.73 Bloodstream infection 27 (2.3%) 21 (1.8%) 0.39 Infective colitis 8 (0.7%) 4 (0.3%) 0.26 Urinary tract infection 16 (1.3%) 15 (1.3%) 0.86 Surgical site infection 10 (0.8%) 12 (1.0%) 0.83 Other infectious complications 18 (1.5%) 24 (2.0%) 0.44

42 Secondary outcomes Non-infectious complications Non-infectious complications Parenteral route, n (%) Enteral route, n (%) P-value Episodes of hypoglycaemia 44 (3.7%) 74 (6.2%) Elevated liver function tests 212 (17.8%) 179 (15.0%) 0.07 Nausea requiring treatment 44 (3.7%) 53 (4.4%) 0.41 Abdominal distension 78 (6.5%) 99 (8.3%) 0.12 Vomiting 100 (8.4%) 194 (16.2%) <0.001 Pressure ulcers 181 (15.2%) 179 (15.0%) 0.91

43 Secondary outcomes Duration of stay (days) in Critical care -All - critical care survivors - critical care non-survivors Parenteral route, median (IQR) 8.1 ( ) 8.8 ( ) 5.2 ( ) Enteral route, median (IQR) 7.3 ( ) 8.2 ( ) 4.3 ( ) P-value Acute hospital -All - acute hospital survivors - acute hospital non-survivors 17 (8-34) 24 (14-43) 7 (3-16) 16 (8-33) 24 (14-43) 6 (2-14)

44 Serious adverse events Total number of patients with serious adverse events Parenteral route, n (%) Enteral route, n (%) P-value 58 (4.9%) 58 (4.8%) 1.00

45 Subgroup analyses Age Degree of malnutrition None Moderate/Severe APACHE II predicted risk ICNARC model predicted risk Mechanically ventilated No Yes Presence of cancer No Yes Time to start of feeding Less than 24 hours 24 hours or over P Value Adjusted for age, ICNARC physiology score, surgical status, degree of malnutrition and a unit-level random effect Favors parenteral route Favors enteral route Adjusted Odds Ratio

46 Adherence-adjusted analysis Adherence-adjusted analysis Relative risk (95% CI) [P-value] Original result 0.97 (0.86 to 1.08) Adherence-adjusted 0.96 (0.85 to 1.09) [p=0.55] Structural mean model with an instrumental variable of allocated group

47 Summary Primary outcome No significant difference in 30-day mortality by route of delivery of early nutritional support Secondary outcomes (significant) Increased episodes of hypoglycaemia with enteral route Increased episodes of vomiting with enteral route

48 Discussion Reported increase in infectious complications with parenteral route was not seen Improvements in management of vascular access? Improvements in prevention of VAP, etc.? Developments in feeding technology? Caloric targets were not met in either group Expected for enteral route? Lack of availability of nutritional product? Content (use of commercially available rather than individually titrated)? Delivery (delays and/or interruptions)? Clinical preference for lower target?

49 Conclusion Early nutritional support through the parenteral route as typically administered in ICUs in England - is neither more harmful nor more beneficial than through the enteral route One-year follow-up for the integrated economic evaluation ongoing

50 This project was funded by the National Institute for Health Research Health Technology Assessment Programme (project number 07/52/03) The views and opinions expressed are those of the authors and do not necessarily reflect those of the HTA Programme, NIHR, NHS or the Department of Health