Restrictive fluid management for ARDS saves lungs, but does it also fry brains? (AJRCCM)

Restricting IV fluids for ARDS patients helps them get off the ventilator faster, in general. But a small follow-up study to the FACTT trial suggests that for some ARDS survivors, that approach may come at a cost of increased cognitive dysfunction for more than a year after hospital discharge.

Since publication of the FACTT trial in NEJM in 2006, a conservative approach to providing IV fluids to patients with ARDS has been considered standard care (by pulmonologists, at least). In case you missed it, FACTT showed that among 1,000 patients with ARDS, being very conservative with IV fluids, and giving lots of diuretics to maintain net fluid balance, resulted in ARDS patients getting ~7 liters less fluid over 7 days, and escaping the ventilator and the ICU 2 days sooner. All without an increased risk for shock, and a (non-significant) absolute 3% lower 28-day mortality rate. What’s not to like?

Given the shortage of positive clinical trials in critical care, you’d think we could all stop and rest for a decade-long congratulatory back-pat after that. But somewhat uncomfortably, Mark Mikkelsen, Jason Christie, Derek Angus et al are reporting a markedly higher incidence of long-term cognitive dysfunction in patients from the FACTT trial who were treated with conservative fluid management. The article is in the June 15 2012 AJRCCM.

What They Did

There were only 406 eligible patients at the time this study started (they were using a subset of surviving FACTT patients for whom quality-of-life data was available). Authors were able to actually talk on the phone extensively with 102. Only 75 patients (~37 from each arm) could complete a battery of validated neuropsychological tests over the telephone. The interviews took place about one year after hospital discharge.

Patients were similar at baseline, and in their clinical outcomes, except the liberal group had had more severe illness (higher APACHE scores, which a priori, should have biased against the observed finding).

What They Found

Half (55%) of patients had cognitive impairment on neuropsychological testing; this is broadly consistent with other studies on cognitive dysfunction after ARDS. But cognitive impairment was significantly more common in the fluid-conservative arm (p=0.005), and was pronounced in executive dysfunction domains.

Lower oxygen saturation (SaO2) during ARDS, irrespective of fluid group assignment, was also associated with impaired cognition (an average PaO₂ of 86 non-impaired vs 71 mm Hg impaired). Two-thirds of patients had psychiatric impairment; this was more likely in those with lower SaO2 during ARDS.

After conducting a conservative sensitivity analysis–i.e., assuming all the patients they couldn’t test completely, who were lost to follow-up, etc. were not cognitively impaired–the association between conservative fluid management for ARDS and cognitive dysfunction persisted, with an odds ratio of about 2.5 (online supplement).

What It Means

This is too small a sample size (75 out of the original 1,000) to conclude that conservative fluid management for ARDS causes cognitive dysfunction. There are built-in biases due to attrition (only survivors were eligible) and variability of who could be reached on the phone. Neuropsychological testing is difficult to perform, especially over the telephone–authors note that patients completed testing “to varying degrees” due to “frustration and fatigue.” Nevertheless, the finding is concerning.

Our medical care system, our natural intensivist mind-set, and the logistical and budget limitations of clinical trials together push us toward measurement of short-term outcomes, with minimal follow-up. This study is nettlesome because answering this question more clearly would require a large, expensive, and ethically complex randomized trial–one that is unlikely to occur (much less conclude anything) in the near future. In the meantime many of us will feel obligated to use a therapy (conservative fluids for ARDS) we know is helpful in achieving short term outcomes, but at a possible price to some patients’ long-term brain health.

Mikkelsen ME et al. The Adult Respiratory Distress Syndrome Cognitive Outcomes Study: Long-Term Neuropsychological Function in Survivors of Acute Lung Injury. AJRCCM 2012;185:1307-1315.