BACKGROUND: Octogenarian Israeli prime-minister Ariel Sharon recently sustained a mild, reversible stroke. A patent foramen ovale (PFO) was detected and anticoagulants were given pending PFO closure. A few days later, he sustained major intracerebral hemorrhage and has since remained in vegetative state. The events triggered serious criticism in the mass media, experts promoting one management option over others. Because knowledge of outcome and hindsight bias evaluation of appropriateness of care, we sought to systematically review the clinical case. METHODS: We performed a formal decision analysis to identify the preferred management between anticoagulation, antiplatelets, PFO closure, or no treatment. Using the best evidence available, we built a decision tree. MAIN OUTCOMES: recurrent stroke and treatment complications within 1 year. RESULTS: Optimal decision was found to be critically sensitive to assumptions about etiology, efficacy and safety of treatments, recurrence risk, and to small changes in utilities. In multiway sensitivity analysis, when the risk of recurrent stroke was <0.12 per year, no treatment was the best management. PFO closure is dominant only when the risk of recurrent stroke is >0.12 per year closure effectiveness is assumed to be <0.28. When closure effectiveness is >0.6, it is inferior to anticoagulation and antiplatelet management. CONCLUSIONS: Uncertainties precluded a clear-cut answer and choice was found to be a "toss-up," often associated with much controversy. Use of novel therapies, such as PFO closure, outside clinical trials will not reduce uncertainty about efficacy.
BACKGROUND: Octogenarian Israeli prime-minister Ariel Sharon recently sustained a mild, reversible stroke. A patent foramen ovale (PFO) was detected and anticoagulants were given pending PFO closure. A few days later, he sustained major intracerebral hemorrhage and has since remained in vegetative state. The events triggered serious criticism in the mass media, experts promoting one management option over others. Because knowledge of outcome and hindsight bias evaluation of appropriateness of care, we sought to systematically review the clinical case. METHODS: We performed a formal decision analysis to identify the preferred management between anticoagulation, antiplatelets, PFO closure, or no treatment. Using the best evidence available, we built a decision tree. MAIN OUTCOMES: recurrent stroke and treatment complications within 1 year. RESULTS: Optimal decision was found to be critically sensitive to assumptions about etiology, efficacy and safety of treatments, recurrence risk, and to small changes in utilities. In multiway sensitivity analysis, when the risk of recurrent stroke was <0.12 per year, no treatment was the best management. PFO closure is dominant only when the risk of recurrent stroke is >0.12 per year closure effectiveness is assumed to be <0.28. When closure effectiveness is >0.6, it is inferior to anticoagulation and antiplatelet management. CONCLUSIONS: Uncertainties precluded a clear-cut answer and choice was found to be a "toss-up," often associated with much controversy. Use of novel therapies, such as PFO closure, outside clinical trials will not reduce uncertainty about efficacy.