The need is urgent to provide older persons with individualized information about the benefits and harms of different diagnostic and treatment strategies. This need results from the growing recognition of the heterogeneity in outcomes in older persons with differing comorbidity profiles. The heterogeneity of benefits and harms resulting from treatment is not yet as well appreciated. Warfarin vs aspirin therapy for the reduction of stroke risk in nonvalvular atrial fibrillation provides an example of a treatment for which the benefit-to-harm ratio may actually reverse according to an older person's comorbidities, thus highlighting the importance of basing this treatment decision on individualized outcome data. Despite the wealth of studies in nonvalvular atrial fibrillation, many assumptions are necessary to calculate patient-specific outcomes, and these assumptions may lead to substantial overestimation or underestimation of benefits and harms. Improving care for patients with comorbidities will require substantive increases in the efforts and resources allocated to the collection and dissemination of outcome data for patients with varying comorbidities.