We commend Prabhakaran and colleagues1 on their trial of tPA use in warfarin-receiving patients with acute ischemic stroke. Guidelines allow tPA if the baseline international normalized ratio is less than 1.7; however, guideline-driven care led, in this study, to worse outcomes with a significant excess of symptomatic intracerebral hemorrhages.
There is still uncertainty about how, when, to whom, and in which settings guidelines should be applied and also about the net effect of standardized guideline-driven management in terms of success vs failure, not a negligible point for physicians or patients. Several reasons may underlie this perceived skepticism toward guidelines. First, recommendations could be based on limited and low-quality data that support misleading conclusions, and this could be the case for TPA use in warfarin-receiving patients. Furthermore, randomized trials often exclude large subsets of patients, ie, those older than 80 years who carry the greatest risk for ischemic stroke, and guideline-driven care seems to focus on the disease rather than the patient, which implies a linear cause-and-effect relationship with only 1 or a few relevant target end points that are clinically monitored. Nonetheless, all comorbidities, natural history and disease progression, effect of therapies, and many other confounding factors add considerable complexity and nonlinearity that is far behind the linear and reductionist view of guideline-driven care.2 Patients should be viewed rather as complex systems in which predictably linear relationships of cause and effect do not result, by themselves, in a predictable behavior of the entire system, ie, the patient.3,4 We need to consider the reality of the complexity of both patients themselves and disease course in terms of comorbidities, drug interactions, multilevel interactions, heterogeneity, unpredictability, and exceptions to the general rules and their clinical significance when making clinical decisions among different options.