The ARDSnet Study

“In a world full of fluid trends, I want to remain classic.” – Iman



This multi-centre parallel group randomised control trial aimed to establish if a lung protective strategy using lower tidal volumes (6ml/predicted body weight in kg) in comparison to tradition tidal volumes (12ml/ kg) used at this time (US 1996) provided a mortality benefit, and increased the number of ventilator-free days in patients with established ARDS as per the usual criteria (pa02/Fi02 ratio <200, bilateral chest infiltrates suggestive of pulmonary oedema in the absence of left atrial hypertension).


This NIH/NHLBI funded study was undertaken in 10 university affiliated ARDS network hospitals from March 1996- March 1999, Study period was 0-28 days (29 days total).
A total of 861 patients were enrolled in the study; 423 patients were randomised to receive the novel (interventional) low volume ventilation strategy with a plateau pressure of less than 30cm H20, whilst 429 patients were randomised to receive the traditional  higher volume ventilation strategy with a plateau pressure of 50cm H20.
Study participants were matched for age, gender, ethnicity, APACHEII, mean tidal volume and all had a diagnosis of ARDS within 36 hours prior to enrolment.
There was a sensible exclusion criteria. Any obvious confounding factor impeding spontaneous ventilation was excluded such as longstanding COPD, neuro-muscular disease and increased intra-cranial pressure. Any factors that would grossly affect mortality outcomes were excluded, such as burns over 30% of body surface area, decompensated liver disease (Childs Pugh C), predicted mortality of greater than 50% at 6 months from another cause, super morbid obesity and bone marrow transplant/ immunosuppression.
Given the ventilator protocol to adjust airway
pressures and tidal volumes, it was not possible to ‘blind’ physicians or ITU nursing staff to the ventilation status of the patients. The patients however were intubated and sedated so were by default ‘blinded’ to the study.
The primary aim was to investigate mortality benefit in these patients, number of ventilator free days, and the number of patients spontaneously ventilating at day 28.
Organ dysfunction was evaluated as a secondary outcome. This included: the number of patients who had circulatory failure (defined by a systolic blood pressure of 90mmHg or were inotrope dependent), had biochemical evidence of clotting dysfunction (platelet count of less than 90), or had new onset AKI. The incidence of baro-trauma was also documented.
Follow up was: death within study time frame, outcome at 180 days following discharge from hospital, to see if the patients were self ventilating at home or at an alternative respiratory institution. (i.e. NIV/ rehab).




The overall relative decrease in mortality was 22% in the low ventilation strategy group. 
In the low ventilation group and traditional ventilation group mortality at 180 days was 31.0% versus 39.8% respectively, ventilator free days 12 vs 10 respectively, breathing without assistance by day 28 65.7% and 55.0% respectively, and the low ventilator group had fewer episodes of organ dysfunction.


The study was stopped after the 4th interim analysis as it demonstrated a statistically significant mortality benefit and it was therefore unethical to continue the study. The absolute risk reduction of 9% correlated with a number needed to treat of 11 patients in order to prevent one death. 




– This was hailed as the ‘landmark’ study.
– The scale of the study is remarkable, n =861 in ten different hospitals.
– Low ventilation strategy is now considered gold standard as per current guidelines in the US and in the UK (NICE, 2016 sepsis 6).
– Beyond of the scope of this summary: The study group utilised their recruitment resources well, and also took a measurement of plasma interleukin 6 at day 0 and day 3 to investigate if an endogenous SIRS process contributed to outcome. In addition, an anti-fungal (ketoconizole and lisofylline) was given to some of the patients. This was an attempt to conduct two trials simultaneously.



-This study assumed that patients have a common pathogenic pathway despite having different aetiology of their ARDS with both pulmonary and extra pulmonary causes.
-It is not possible to attribute outcomes solely on the basis of tidal volumes used. The patients in the low-ventilation group were given bicarbonate infusions (it is not stated how much) and their respiratory rate was amended to maintain strict normal ETCO2 parameters and a pH of 7.15. This was stated in the fine print of the paper.
-The statistically significant ARDSnet finding may have been related to the greater spread in tidal volumes used and plateau pressures between the control arm and the intervention arm. A large difference in independent variables lends itself to a positive finding and hence publishable results. 3 previous studies investigating low tidal volume ventilation had statistically insignificant results.
– Two trials were simultaneously conducted in the same patient group. Ketoconazole was compared with placebo in the first 234 patients and lisofyline was compared with placebo in the last 194 patients, no anti fungal agents were given to the middle 433 patients. This is a poor study design and may have contributed to study biases.
– Was the interventional arm using low tidal volume ventilation really ‘protective’ or was the ‘traditional’ ventilation strategy in comparison very damaging? ARDSNET used higher than normal tidal volumes in their control arm. 12 mls/ kg versus 10 mls/kg used as a standard in 1996. Was this a strategy to get statistically significant data?
– A very high PEEP protocol was used between 32-39 cm H20).
– The study used predicted body weight rather than actual body weight which as they study accepts could have lead to a 20% error margin affecting the tidal volume calculated.
– The paper provides no details about: power calculation, protocol on decision to re-intubate,  proportion of patients of each study arm on anti-fungals.
– The study has been criticised by the ethics committee OHRR stating that in their opinion, the informed consent educational materials failed to adequately describe foreseeable risk and discomforts to the patients. The investigators have therefore subject the patients’ to unnecessary harm.
Summary by Dr Shivani Amdekar

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s