An ethical hierarchy for decision making during medical emergencies

Abstract

Evidence from well-designed clinical trials may guide clinicians, reduce regional variation, and lead to improved outcomes. Many physicians choose to ignore evidence-based practice guidelines. Using unproven therapies outside of a randomized trial slows recruitment in clinical trials that could yield information on clinical and economic efficacy. Using acute stroke therapy as an illustration, we present an ethical hierarchy for therapeutic decision making during medical emergencies. First, physicians should offer standard care. If no standard care option exists, the physician should consider enrollment in a randomized clinical trial. If no trial is appropriate, the physician should consider a nonrandomized registry, or consensus-based guidelines. Finally, only after considering the first 3 options, the physician should use best judgment based on previous personal experience and any published case series or anecdotes. Given the paucity of quality randomized clinical trial data for most medical decisions, the "best judgment" option will be used most frequently. Nevertheless, such a hierarchy is needed because of the limited time during medical emergencies for consideration of general principles of clinical decision making. There should be general agreement in advance as to the hierarchy to follow in selecting treatment for critically ill patients. Were more clinicians to follow this hierarchy, and choose to participate in clinical trials, the generation of new knowledge would accelerate, yielding rigorous data supporting or refuting the efficacy and safety of new interventions more quickly, thus benefiting far more patients over time.

Figure 1. MRI and MRA Scans

Images were made 3.5 hours after symptom onset, and 1 hour after the completion of intravenous rt-PA therapy. In Panel A, the left internal carotid and middle cerebral arteries do not appear, consistent with occlusion. In Panel B, a diffusion-weighted image from a mid-parietal section, a large area of diffusion restriction is seen in the territory of the right middle cerebral artery. In Panel C, an Apparent Diffusion Coefficient image at the same level as in Panel C confirms that the diffusion restriction seen in Panel B is consistent with a very acute ischemic event.

Figure 2. Intervention

Following intravenous rt-PA, the patient was studied with digital subtraction angiography. Panel A confirms that the right middle cerebral artery remains occluded at 7 hours following onset of symptoms. In Panel B,, successful recanalization of the middle cerebral artery is demonstrated, although reperfusion is not complete in the microcirculation.

Figure 3. Hypothetical Scheme for Applying Clinical Trial Data to Individual Patients

The curve is a speculative representation of the relationship between study criteria and the individual patient. Conceptually, the further the patient is from the sample population contained in the clinical trial, the lower the likelihood of benefit.

Figure 4. Example of Diffusion-Perfusion Mismatch

In Panel A, there are multiple areas of restricted water diffusion, interpreted as small regions of irreversible injury. In Panel B, a perfusion image shows an area of diminished perfusion much larger than the area of diffusion restriction. The hypothesis to be tested is that the brain with reduced perfusion showing minimal diffusion restriction remains salvageable. To date, the mismatch hypothesis has not been proven conclusively.