NICE Sugar

Screen Shot 2018-08-07 at 10.46.16.png

Intensive insulin therapy in the MICU/SICU

Included 6104 medical/surgical patients with an anticipated ICU length of stay > 3 days
Randomized patients to intensive control (BG 81-108 mg/dL) vs. conventional control (<180 mg/dL)
Primary outcome (90-day mortality) was more common with intensive control (27.5% vs. 24.9%, p=0.02, NNH 38) and was primarily driven by cardiovascular causes
Subgroup analysis of primary endpoint showed possible benefit with intensive control with trauma (OR 0.77, 95% CI 0.5-1.18, p=0.07) and with corticosteroids at baseline (OR, 0.88, 95% CI 0.66-1.19, p=0.06)
No difference in 28-day mortality (22.3% vs. 20.8%, p=0.17)
Mean morning BG values were 118 mg/dL (intensive) vs. 145 mg/dL (conventional)
More patients in the intensive group received corticosteroids (34.6% vs. 31.7%, p=0.02). The significance of this finding is very controversial
Severe hypoglycemia (BG < 40 mg/dL) occurred more with intensive control (6.8% vs. 0.5%, p<0.001, NNH 16)
There were no differences in morbidity outcomes, such as length of stay, duration of mechanical ventilation
There were no differences in morbidity outcomes, such as length of stay, duration of mechanical ventilation, need for renal replacement therapy, or infections
Two major differences in comparison to the Leuven trials (van den Berghe
et al) are that NICE-SUGAR was multi-center and used less parenteral nutrition — both of which increase external validity

TBL

Among critically ill patients, intensive glucose control increased 90-day mortality and the incidence of severe hypoglycemia compared to conventional therapy.

See the paper here

NINDS

Screen Shot 2018-08-07 at 10.54.19.png

Alteplase within 3 hours for acute ischemic stroke

Included 624 patients presenting within 3 hours of ischemic stroke not meeting a long list of exclusion criteria, many related to risk of bleeding. There was no exclusion on the basis of severity of stroke or maximum age.
Primary outcomes were split between two phases of the study. In phase I (n=291), the primary outcome was NIHSS improvement at 24 hours. In phase II (n=333), the primary outcome was a global assessment of four clinical outcome scales at 3 months
Baseline NIHSS score was median of 14-15
At 24 hours (phase I), there was no difference in NIHSS improvement
At 3 months (phase II), the odds ratio of a “favorable outcome” (global test) was 1.7 (95% CI 1.2-2.6), favoring TPA
The NIHSS subgroup of the global test was improved compared to placebo (31% vs. 20%, p=0.0033, NNT=9)
Symptomatic ICH occurred more frequently with TPA (6.4% vs. 0.60/0, p<0.001, NNH=17)
ICH was more common in patients with more severe stroke (NIHSS 20 vs. 14) or CT evidence of cerebral edema (9% vs. 4%)
Minor bleeding was more common with TPA (23% vs. 3%, NNH 5)
There was no difference in mortality

TBL

In patients presenting within 3 hours of ischemic: stroke, alteplase improved 3-
month neurological function (NNT=9) but did not impact 24-hour symptoms or mortality. In patients receiving alteplase, approximately one-quarter had minor bleeding, and 6.4% had symptomatic ICH (NNH=17).

See the paper here

NUTRIREA 1

Screen Shot 2018-08-07 at 11.45.43.png

Monitoring gastric residuals in mechanically ventilated patients

Included 449 intubated patients who were expected to required more than 48 hours of mechanical ventilation and who received enteral nutrition within 36 hours of intubation
Excluded recent abdominal surgery; those with any history of upper GI surgeries; and those with any history, of GI bleeding
Randomized patients to intervention (no gastric residual monitoring; intolerance determined based on vomiting) or control (residuals monitored Q6hr; intolerance defined as residual > 250 mL or vomiting)
Ventilator-associated pneumonia (VAP) suspicion required a positive chest x-ray and at least two of the following criteria: WBC > 10 or < 4; temp > 38.5 or < 35.5; Purulent tracheal aspirates.
VAP was confirmed in those with suspicion of VAP who had a confirmatory culture from BAL, protected brush, or tracheobronchial aspirate.
VAP diagnosis was adjudicated by a blinded committee.
Primary endpoint (patients with at least one VAP episode) was non-inferior between intervention and control (16.7% vs. 15.8%, 90% CI-4.8% to 6.7%). The study did rule-out a pre-specified 10% non- inferiority margin.
Other outcome measures related to VAP were also non-inferior (eg, cumulative VAP incidence, total number of VAP episodes)
Vomiting was more common in the intervention group (39.6% vs. 27%, OR 1.86 [90% CI 1.32-2.61], p=0.003)
Intervention patients had a lower mean caloric deficit within the first seven dayS (319 vs. 509 calories), although this difference is unlikely to be clinically relevant
Feeding intolerance (defined differently between the two groups) was more common with the control group (39.6% vs. 63.5%)
Vomiting is a concerning endpoint in that it may have led to adverse events that were not captured by the Primary endpoint. As an example, non-bacterial pneumonitis from aspiration would not have met the; primary endpoint criteria.
Although there were no difference in ICU length of stay, duration of mechanical ventilation, or mortality, it is not clear whether the study was powered to detect these endpoints specifically.
90-95% of patients in the study were medical (non-surgical). It is not clear whether these results are applicable to surgical patients who are more likely to experience feeding intolerances.
Guidelines differ on the appropriate volume to hold enteral nutrition.
Other guidelines recommend a holding parameter of 500 mL, which would have altered this study’s incidence of “feeding intolerance” in the control group.
The authors fairly argue that the standard practice of gastric residual monitoring is not evidence-based.
There are no data supporting the practice of checking gastric residuals. supporting specific cut-offs that should prompt intervention to improve feeding intolerance, or that checking gastric residuals decreases VAP incidence rates.