The last of the trilogy of Early Goal Directed Therapy (EGDT) for sepsis management studies was published recently. The results, on the face of it, appeared surprising. There is no difference in outcome when either an algorithmic-based protocol or clinical gesalt is used in the management of patients with septic shock.
The reasons for this are not simply a failure of EGDT. (I recommend an excellent review of the ProMISE trial by Rick Body of St. Emyln’s.) The combined results of the three studies highlight, once again, that what is expected to be the case in clinical medicine often isn’t. Perhaps we should extend this caution to our assumptions about medical education.
I recently did a podcast with Prof Coats, one of the authors of a key EGDT paper, where he highlighted how the control group had provided a very good snap-shot of good clinical practice in sepsis management. Much of emergency and critical care research is pragmatic rather than explanatory in nature. The biases and confounders that surround patients who with complex diagnoses, multiple co-morbidities and a large number of potentially relevant outcomes mean one intervention rarely has just one outcome. Only by looking at what happens with an intervention in authentic practice can we understand its effect.
This is the challenge for medical education. Say you have an educational technique or practice that you wish to implement in a clinical setting. Defining the “control” can be problematic, while the delivery of the intervention in a ‘real-life’ setting is sometimes impossible. Education interventions are more complex with multiple influences when compared with a sepsis intervention that requires inserting (or not) a central venous line.
There is a paucity of large-scale, pragmatic, education interventions. Perhaps the solution is to incorporate education interventions into large, multicenter therapy trials. In the appendix of the study you discover that in both groups approximately 20% of patients had no consultant (attending physician) review patient care during the intervention period. A specific sub-group analysis has not been performed, but it seems unlikely these groups are significantly different. Too bad these groups were not targeted for a nested medical education study within the larger clinical trial. Could simulation, one-to-one mentoring or asynchronous learning be used with these trainees to improve outcomes for patients with septic shock? Large multicentre trials, like the EGDT trilogy, potentially have much to offer the medical education community, if we seek out opportunities to embed education questions within a clinical study.
Image by stockimages. From Freedigitalphotos.net