Looking Beyond Single Randomized Controlled Trials

By critically appraising the entire body of literature on specific interventions and clinic outcomes, we have learned many lessons about what is most important in the delivery of critical care. However, we have learned that we must wait before we immediately embrace the results of a single randomized clinic trial (RCT) with very impressive results and instead base our clinic practices on more comprehensive, cautious, and critical appraisals of all of the available literature.

The last two decades of critical care research are filled with stories of impressive findings from single-center RCTs that could not be replicated in larger multicenter RCTs. Unfortunately, in many cases, the initial positive single-center results have been embraced by early adopters, only to have the results refuted by subsequent follow-up trials. The story of tight glycemic control in critical illness is illustrative. A single-center study of the management of hyperglycemia in a population consisting primarily of postcardiac surgical patients found that intensive glucose management with insulin infusion with a target blood glucose of 80 to 110 mg/dL dramatically reduced mortality when compared with a more lenient target blood glucose of 160 to 200 mg/dL. The results of this single-center study were embraced by many intensivists and rapidly generalized to a wide variety of critically ill patents. The factors behind this rapid adoption by the field are multiple, including ease of implementation and cost. Unfortunately, a subsequent similar study of medical patients showed no significant benefit of an intensive insulin therapy protocol in the critically ill medical patient. Ultimately, the most comprehensive multicenter trial of medical and surgical critically ill patients found significantly increased mortality in the group randomized to a tight glycemic control protocol, compared with targeting a blood glucose level of less than 180 mg/dL. This excess mortality was likely due to the much higher rates of severe hypoglycemia.

In 2001, the era of early goal-directed therapy (EGDT) was introduced through the publication of a single-center, randomized controlled trial. EGDT was widely adopted, and multiple subsequent published trials, all prospective cohort series, confirmed its benefit. More recently, two large RCTs failed to demonstrate a survival benefit when protocolized resuscitation was compared with “usual care.” It is possible that these results, at least in part, reflect the effect of the original EGDT trial; the widespread adoption of aggressive, early resuscitation; and the broad-based implementation of the Surviving Sepsis Campaign Guidelines and bundles. If this continues to define usual care, then perhaps it is no longer necessary to mandate specific protocols for resuscitation because it appears that standard sepsis management has evolved to be consistent with published protocols.

The evidence for the use of hydrocortisone in the treatment of septic shock is an example of a sepsis treatment in which the initial promising study was embraced quite early, only to be questioned by subsequent conflicting evidence. We are still awaiting the final answer about the utility of the administration of corticosteroids as an adjunctive therapy in septic shock.

Activated protein C is an example of how little we still currently know about the pathobiology of sepsis and the difficulty in developing targeted therapies. Activated protein C as an adjunct therapy for patients with sepsis initially was thought to be quite promising, but it was abandoned after subsequent randomized controlled trials failed to duplicate the original results.

Small Things Make a Big Difference

The evidence-based era has taught us that small, often neglected or overlooked details of everyday bedside care can play a large role in determining whether our patients survive their ICU stay. Pneumonia that develops after the initiation of mechanic ventilation (ventilator-associated pneumonia [VAP]) is associated with high morbidity and mortality and significantly increased costs for critically ill patients. Several simple targeted interventions to address this problem have significantly reduced VAP rates. Simply keeping our intubated patients’ heads elevated at least 30 degrees rather than leaving them supine (as was customary two decades ago) has resulted in major reductions in VAP. In addition, a focus on better oral hygiene of mechanically ventilated patients via the administration of oral chlorhexidine has even further reduced the VAP rates.

Another simple small intervention in the evidence-based era, the early mobilization of our critically ill patients, has also been found to significantly improve patient outcomes. We previously kept critically ill patients immobilized for weeks on end in the belief that this was necessary for their recovery. The result was very high rates of ICU-acquired weakness that required prolonged periods of rehabilitation in ICU survivors. More recent studies have shown dramatic improvements in functional status and significantly decreased ICU length of stay (LOS) when critically ill patients are mobilized as soon and as much as possible.

Accountability Is Important

Another important lesson learned during the evidence-based era is the importance of tracking clinic behavior through performance measures. Published reports have demonstrated a significant gap between intensivists’ perceptions of their ability to adhere to current evidence-based medicine and actual practice. This dichotomy has been noted in adherence to low tidal volume strategies in acute respiratory distress syndrome and other common “best ICU practices.” These findings have led to the development of checklists and performance metrics to foster clinician accountability that have provided tangible improvements in clinic care. Multifaceted interventions using checklists have dramatically reduced catheter-related blood stream infections as well as complications from surgical procedures.

In acute situations, checklists have also been shown to improve delivery of care. Continuous measurement of individual performance in the evidence-based medicine era has allowed ongoing, real-time feedback to individual clinicians and groups of providers. Application of this approach to sepsis care has resulted in significant improvement in adherence to evidence-based guidelines and in patient outcomes.

Do Less, Not More

The evidence-based era has also taught us that we often should do less, not more, to and for our critically ill patients. We have learned that interrupting sedation and awakening mechanically ventilated patients each day, and thus reducing the amount of medication administered, can significantly reduce ICU LOS. When coupled with a daily weaning trial, daily awaking of ICU patients significantly reduced mortality. We have also learned that decreasing the need for mechanic ventilation by first using noninvasive strategies in specific groups of patients with acute respiratory distress can improve outcome. In addition, use of smaller tidal volumes in mechanically ventilated patients has been shown to be lifesaving. We have also learned that reducing the amount of blood given to patients can significantly improve outcomes.