Wessex ICS Blog

Summary of ESICM Hot Topics Session

Author of summary: Dr Emma Fitzgerald
Follow @emma_zcharm6
Follow @WessexICS

ESICM Hot Topics 2014

The effects of decontamination of the oropharynx and intestinal tract on antibiotic resistance in ICUs

Eradication of and prevention of potential pathogens
Tobramycin, colistin and amphotericin B
SDD: 3rd generation cephalosporin added
Used for patients with expected LOS of >48hours
Topical components until end of ICU discharge

Previous studies have shown that SOD not different to SDD (Dutch multi-centre study)

This study:
2 Armed cluster randomisation. No control as not deemed ethical in Netherlands.
An ICU was randomised to one for 12 months, 1 month cross over and then a 2nd 12 month period. 16 ICUs, 31264 eligible patients.

Included: 5881 SOD patients, 6116 SDD patients

Of note, there is a very low rate of drug resistance in the Netherlands as compared with rest of the world

Sample form very patient in the ICU once a month to look at the ecology of the ICU
Intestinal colonisation – highly resistant micro organisms –more present in SOD group
Aminoglycoside resistance was higher in SOD group.
Respiratory carriage – no difference
Mortality – no difference, even with adjusted analysis for 28 day, ICU or hospital mortality.

Subgroup analysis
- Surgical patients (theory that they may benefit more from SDD): No differences seen.
- Bacteraemia – higher number of positive cultures in SOD patients, esp enterobacter.
- Aminoglycoside resistance is lower in SDD group

- Low levels of antibiotic resistant
- Intestinal tract carriage – SDD lower than SOD
- Gradual increase in Gram negative aminoglycoside resistance
- Respiratory tract – no difference
- No LOS or mortality differences in SOD or SDD groups

- JAMA editor in chief: Feeling that it is relevant and provides useful information, despite the ecology and resistance profile being quite different in the Netherlands

[email protected]: Effect of High-Dose Vitamin D3 on Hospital Length of Stay in Critically Ill Patients With Vitamin D DeficiencyThe VITdAL-ICU Randomized Clinical Trial
- Vitamin D is a steroid hormone, regulates about 200 genes
- It has many other functions that are not well recognised.
- High prevalence of low vitamin D levels in the ICU

Is vitamin D a marker (only) or a contributor?

Pilot study performed in a single centre – biochemical primary endpoint to establish the correct dose and establish safety.

This Study
Hospital LOS primary endpoint
Given Loading plus maintenance dose monthly for 6 months
Used a pre-defined threshold for deficiency

Median LOS in ICU and hospital was the same.
No differences in mortality
Even in the treatment group, they still remained vitamin D deficient (approx 50% of patients)
No toxicity
Some hypercalcaemia occurred but no treatment needed

Subgroup – severe deficient
- Big reduction in mortality in this group
- NNT 7

Post hoc power analysis gives a power of 0.7 for the secondary outcome.

The question is now, should we measure vitamin D levels of our patients to stratify them?
Need more trials!!

EPO-ACR 02 Trial - Early high dose erythropoietin therapy after OOHCA
We need more neuroprotective treatments
There may be potential neuroprotective effects of EPO, pleiotropic effects.

Pilot study conducted.
French multi-centre trial
Hypothesis - early high dose EPO could improve neurological outcomes post cardiac arrest.

Inclusion criteria:
ROSC had to be within 60 minutes
Needed GCS<7

Double blinded RCT
Immediate administration of EPO and 5 subsequent doses
Primary end point – Level 1 on CPC scale at D60
Secondary endpoints – other CPC outcomes, mortality
501 included, 25 excluded - 476 overall
Baseline characteristics were similar, including the time to ROSC

32% had CPC level 1, but no significant difference between the two groups.
Higher rate of adverse events – mainly thrombotic events
Mortality similar

Similar to other studies in other groups of patients – no improvement in outcome seen

- Long ROSC time. Question asked about whether some had too long a down time to see a good effect – subgroup analysis didn't show any benefit in those that had EPO early
- Subsequent question was labour whether it should be given before ROSC? Thought to be too hard to do a study or subsequent y to translate into clinical practice
- Are we done with EPO?!? It doesn't seem to work! Answer – it maybe useful in neonates but doesn't seem to help anywhere else!!

The ARISE Trial – Goal-Directed Resuscitation for Patients with Early Septic Shock

Evaluating role of early goal directed therapy.
3 SSC guidelines since Rivers study.
ProCESS trial – patients having EGDT or Protocolised care had no benefit over usual care

EGT is an intervention and it depends on the context in which they are delivered. Are the results applicable in all health care settings and also can it actually be delivered across all healthcare settings if they are applicable?

Prospective, unblinded, randomised, parallel group trial
Conducted in 51 sites
Diversity of healthcare settings involved (tertiary and non-tertiary)

Prior to starting – prospective trial was performed to assess what “usual” care was in the centres. Of note, usual care did not include SVO2 measurement

This study
6 hour intervention period
Randomisation had to occur within 2 hours
Strict inclusion and exclusion criteria
In the usual care group – SVO2 was not permitted in the 6 hour period
The algorithm was the same as Rivers and the ProCESS trials.
Primary outcome – all cause mortality at 90 days
Numerous secondary outcomes.
Baseline mortality was estimated from large databases and previous trials at 28%
Predefined sub groups were analysed

3559 met the inclusion criteria, 1600 were enrolled. 792 in EGT group, 796 in usual care group after exclusion etc
Both groups had similar fluids pre-study, BP and lactate also similar.
Time to ED presentation to randomisation was <3 hours in both groups
Similar baseline characteristics. Of note, most came from home
Blood cultures positive in 38% of both groups.

- 90% of EGT had CVC In time allowed
- 61% had a cvc in usual group, Similar time to insertion
- More fluids given in EGT group
- Hourly compliance – high compliance in all components
- No difference in the resuscitation end points between groups
- The only difference was a difference in the MAP, this was not clinically significant though
- Mortality was 18.6 vs 18.8%. Not significant.
- EGT had shorter LOS in ED
- More EGT Patients went to ITU, but some needed to go there to deliver the bundle
- No other mortality differences
- No other differences in LOS or organ support

- Largest RCT on EGT and the only one in a variety of healthcare settings in different countries
- EGT used more fluid, RBC and vasopressor agents but this didn't seem to matter

- Is it generalisable given that 50% were not included?? Paul Hebert responds for them! – the internal validity is outstanding. External validity is too, fits in with all of the other trials that are done and actually matters less seeing how good the internal validity is!
- Hard to define “usual” treatment to your juniors. The SVO2 target didn't seem to matter. Usual care is good and does not seem to need the SVO2 measurement.
- If the SVO2 was >70% at cvc insertion, hard to see how targeting this would make a difference.
- Are we looking at patients who aren't that sick?? Low rates of intubation. Response – similar to the PROcESS trial. This study used less fluid and more vasopressors. Maybe different interpretation of the algorithm.

TRISS Trial: Lower versus Higher Hemoglobin Threshold for Transfusion in Septic Shock

Scandinavian Trials Group
Aim – To assess the effects and safety of a lower vs higher Hb in patients with septic shock in the ICU
Transfusion triggers of 7g/dl or 9g/dl
Needed to have septic shock AND an HB of 9 or lower to be included
Primary outcome – mortality at 90 days
Variety of secondary outcomes
Exclusion of those with acute coronary syndromes, or transfusion already.
1005 were randomised
With loss of some… Final numbers 998 (977 for secondary outcomes) 502 (lower) vs 496(higher)

Lower threshold transfused approx half that of the higher threshold group.
They achieved their Hb thresholds in both groups

Mortality – no differences.
No difference in life support use at day 5
Only 1 adverse event
No difference in ischaemic events

No difference in outcomes (even adjusted etc) in low threshold group. Suggests a trigger of 7 is safe

- Extraordinary study that will change practice. Should NOT be called a negative study!!
- Was there a relationship with lactate and other parameters? No data to answer this. But the baseline characteristics were very similar
- Did you look at the amount of fluid being given? The cumulative amounts were similar. No other data collected on this
- The study seems to support that a trigger of 7 is the way to go in most patients.
- This blood was Leuco-depleted – different to TRICC
- The rates of ARDS and pulmonary oedema were much higher in the TRICC study. Here they found none. Here, it may be that some of these patients and reflects the fact that these patients are sicker and you may not see small changes in oxygenation etc here as you might in the TRICC study patients.

CALORIES: Trial of the Route of Early Nutritional Support in Critically Ill Adults

When the trial was started, there were 3 meta-analysis that all had conflicting results. This study reflects the commissioning brief set out in the UK
A phase III multi centre RCT
Inclusion criteria aimed to standardise other aspects of care
Excluded if in ITU for more than36 hours or if they had been fed in that last 7days…plus others
Randomised to parenteral or enteral route.
Target of 25kcl/day to be met by 48 hours.
Dietician input into trial design.
Interim analysis halfway – trials continued.
1200 in each group. A few lost --> 1188 parenteral and 1195 enteral
Baseline characteristics were balanced.
92% deemed not to be malnourished.
Non adherence – related to new line insertion
The time targets were met- feed was started early
The calorie targets were not met. No real differences in protein intake.
Approx 80% of both groups had vasoactive medication.

All cause mortality at 30 days – no difference and close to baseline assumptions.
No difference in mortality at 90days either
No difference in organ support used
No difference in new infectious complications overall, or with site of infection
Higher hypoglycaemia and vomiting in enteral route.
LOS – no difference between groups and they did stay the 3 days predicted by the clinician enrolling them

- No difference in mortality with route of delivery of nutritional support
- No increase in infectious compilations ?due to better cvc are and VAP prevention now?
- Caloric targets were not met in either group. Expected in enteral group? Why in TPN group though?

- Lots of issues around PRIOR nutrition, but there did seem to be a lack of malnutrition
- Maybe reaching the calorific targets may have shown a difference between the two groups? It is a problem with pragmatic trials.
- Use the NG route and establish how much to give