Tuesday, February 19, 2008

Complex trials for complex condition can be ..well complex

The history of ARDS ( acute respiratory distress syndrome) is a relatively short narrative in contrast to some other diseases and their treatments. I remember writing a review of a book about the topic as a fellow in pulmonary disease circa the early 1970s at that time very little was written about that topic, the condition having been recognized somewhere around 40 years ago.

We now talk about ALI (acute lung injury) as well as ARDS, the former being a sort of "ARDS- lite"with less severe oxgenation issues than full blown ARDS and it considered a precursor to ARDS. The "A" in ARDS first stood for "adult" and now is taken to mean " acute". Specific definitions of ALS and ARDS appeared in 1994 based on oxygenation levels defined by the ratio of arterial oxygen tension and the fraction of inspired oxygen.

Ventilatory support of these patients can be lifesaving but the optimal techniques or "settings of the ventilator" continue to be a source of controversy.Too high a tidal volume or a pressure and we risk various forms of barotrauma and too little of either seems to be associated with atelectasis ,blood shunting and another form of ventilator induced lung damage from forces related to opening and closing of gas exchange units.

Recent published clinical trials involving various techniques to ventilate these patients report in- hospital mortality rates for the treatment and the control arms in the range of 30 to 40% while a 2005 observational trial of 467 patients reported a 60% mortality rate which may well more accurately reflect real world outcomes,the old efficacy-effectiveness gap issue.

Two large clinical trials were published in the Feb. 13,2008 issue of JAMA as were as two editorials on the subject. I believe this brings to three the total of clinical trials that investigated the use of "high"er versus "low"er levels of PEEP and all found no statistically significant difference in terms of mortality although some less final measures tended to favor the higher levels. There were fewer days on ventilator support in one trial and fewer deaths from progressive hypoxemia in another.

Also favoring the higher levels are current theoretical considerations regarding what seems to be a major pathophysiological factor in ARDS. There appears to be considerable non-cardiac pulmonary edema in ARDS and the suggestion that these wet alveloi are capable of being ventilated (recruitable alveolar units) and since PEEP should enable lung unit expansion the subset of ARDS patients with more edema might significantly benefit from PEEP while other patients with less edema might not. With the now commonplace use of low tidal volumes one would expect more micro-atelectasis and it seems reasonable that higher PEEP might be needed to mitigate that tendency.

Accordingly, one editorialist suggested trials in the future might well use some measure of edema to stratify patients.The thought here is that those patients with more edema might benefit more from higher PEEP.Interestingly, in some patients PEEP seems to cause harmful overdistention without eliminating the atelectasis.

Clinical trials in complex patients pose a number of problems not seen in simpler randomized clinical trials such as those in which two medications are compared in more stable patients.

One of the editorials discusses some of the problems seen in trials such as those reported in ARDS including the necessary lack of blinding and the difficulty and traps involving in the crucial definition of exactly what the control group will be.This involves in part how much discretion are the treating docs given in the control group. Further,"inferences in causality can only be drawn about the entire intervention and not about any specific piece."

Interesting (and somewhat disconcerting to me) even though we now have three major clinical trials that all failed to demonstrate a mortality benefit to higher versus lower PEEP, all of the authors of these latest two article and one of the editorials agree that because of the compelling theoretical arguments favoring adding high PEEP to the current low tidal volume standard of care and the favorable morbidity measures that higher PEEP should be part of the de facto standard of care. If I were still twiddling the dials of the ventilators in ICU in patients with ARDS I would probably opt for higher PEEP as it seems to being about (at least in some patients) higher oxygen levels and lung docs always seem pleased with higher values and feel that we have accomplished something.


James said...

Maybe PEEP and insulin have similar theraeutic indices. Studies of lower glucose control with insulin showed no great advantage in terms of CV mortality, at least some of the studies failed to show our dogma was correct.
So oxygen and pulmonary pressures can be bad, too. Maybe both types of patients need to have their treatment carefully calibrated to their individual response. If pO2 fails to rise with a bit more PEEP, cut it back until the pO2 is optimized.

james gaulte said...

Careful individualized and calibrated treatment of ARDS patients or anyone with a complex,rapidly changing clinical problem is what typically takes place in ICU and it is quite a lot to expect planners to be smart enough to devise protocols to replace that.