The Collaborative Clinical Trials in Intensive Care Medicine Conference started yesterday (Monday 15 June 2015) at the Monash University Prato Centre (#pratoICM15). This multi-disciplinary conference, hosted by the Australian and New Zealand Intensive Care Research Centre (ANZIC-RC) aims to foster discussion among key stakeholders concerning the direction of ongoing clinical trials in the critically ill.
Here are are some highlights and discussion points from the first day of the conference.
ICU Recovery – It’s not just about mortality
The first session of the day explored the importance of quality of life (QOL) as an evolving outcome for critical care research. Dr. Carol Hodgson (Senior Research Fellow – ANZIC-RC; @carol_user) provided a comprehensive review of this topic, highlighting the variable tools used for this purpose (SF-36, EQ-5D, Barthel Index, GOSE), and the limitations and advantages of each. QOL appears to be an important health outcome for patients, but is often highly subjective; being heavily influenced by an individuals expectations, and background. It therefore remains uncertain whether we are capturing the correct data, and how these can be used to inform trial design. Importantly, a key driver of Post-ICU QOL appears to be pre-existing comorbidity, rather the aspects care provided in the ICU.
Early mobilization and physical therapy are evolving interventions that may improve clinical outcomes in the critically ill. This program of research is being explored by the ANZIC-RC, with data from a multi-center cohort study suggesting that ICU-acquired weakness is common, and associated with greater 90-day mortality. However, a recent trial exploring the efficacy and safety of very early mobilization within 24-hour of stroke, has demonstrated that this intervention was associated with a reduction in favorable outcomes at 3-months. It therefore remains uncertain whether we should simply adopt this intervention, which inherently seems a good idea, or pursue a large randomized controlled trial?
In this sense, a greater understanding of the biology of this process may be needed, in that our knowledge of the cause-effect relationship is very limited, e.g. are the weak dying, or are the dying weak??
ARDS – Reducing noise
The definition of ARDS has evolved in recent years, yet remains primarily based in on the classical features described by researchers in the 1960’s (Lancet 1967; 2:319–323). However as highlighted in this session, this often lacks specificity and therefore introduces ‘background noise’ when attempting to study interventions in this population. In this fashion the ACCP definition of ARDS was able to identify a cohort of critically ill patients that benefited from lower tidal volume ventilation, but multiple other interventions have been far less successful in the same group. While this may simply reflect a lack of efficacy, it could also be that we need to more precisely define the phenotype where such interventions are actually likely to work. An example would be determining a patients response to higher PEEP, as a surrogate endpoint that guides trial enrolment (where by if the patient’s oxygenation improves he or she is then randomized into a trial of varying PEEP strategies, focusing on patient-centered outcomes). This strategy allows you to reduce your sample size, and focus your intervention in a group of patients where it is likely to be of benefit. This reduces noise.
RBC transfusion – More like good wine
The results of the ABLE Study (N Engl J Med 2015; 372:1410-1418) were presented by one of the principle investigators (Paul Hébert). This large multicenter randomized controlled trial compared fresh RBC transfusion (storage duration < 8-days) versus standard RBC transfusion in critically ill patients across Canada. Groups were adequately matched at baseline, with excellent protocol compliance. Fresh RBCs were stored for a mean of 6.11 days versus 21.97 days in the control arm. There was no difference in the primary outcome of 90-day mortality; 37.0% with fresh RBC compared with 35.2% with standard RBC transfusion (absolute risk difference, 1.7 percentage points; 95% confidence interval [CI], –2.1 to 5.5). There was no statistically significant difference in any secondary outcomes.
These results appear at odds with the wealth of pre-existing observational data suggesting potential harm associated with older RBC transfusion. Importantly, by design, this study did not compare fresh RBC transfusion with truly old RBC, but rather standard practice. As such there was only a small number of very old units within control arm. As such whether there are any truly toxic effects from prolonged storage times remains uncertain. In addition, the observation that essentially all of the tested outcomes favored standard issue RBC (albeit non-significantly) raises the issue of whether we need to re-evaluate the proposed benefits of transfusing the freshest available blood. Like good wine, perhaps there may be some benefit in cellaring red blood cells, at least for a while.