Editors' Note: Neuro R 2 Score: Predicting High-Risk Neurologic Readmissions Within 30 Days

Abstract

In “Neuro R2 Score: Predicting High-Risk Neurologic Readmissions Within 30 Days,” Peacock et al. retrospectively reviewed the records of 4,876 neurologic admissions over a 5-year period and found that 314 (6.4%) patients were readmitted within 30 days of discharge. Predictors for readmission included the Charlson disease count (a score based on comorbid conditions), urgent or emergent admission, discharge to rehabilitation or skilled nursing facility, and diagnosis of cancer, brain tumor, or cerebrovascular disease. These were used to design and validate the Neuro R2 score to predict 30 days' readmissions. Majersik noted that this score can help neurologists predict unplanned readmissions, a quality metric monitored by the Centers for Medicare and Medicaid Services, and recommended focusing attention on reducing readmissions for patients with brain tumors and those discharged to skilled nursing facilities. Hunter was more skeptical of the benefits of this score and commented that the utility of risk stratification algorithms is limited because of implementation barriers, inaccurate predictions, and ineffective risk reduction strategies. For Peacock et al., Freeman responded that the Neuro R2 score is one of the largest multivariate models for 30-day unplanned neurologic readmissions but that it certainly would not predict (or prevent) all 30 days' unplanned neurologic readmissions. He further noted that external validation of this model, and the creation of additional models focused on neurologic patients, is needed. In “Neuro R2 Score: Predicting High-Risk Neurologic Readmissions Within 30 Days,” Peacock et al. retrospectively reviewed the records of 4,876 neurologic admissions over a 5-year period and found that 314 (6.4%) patients were readmitted within 30 days of discharge. Predictors for readmission included the Charlson disease count (a score based on comorbid conditions), urgent or emergent admission, discharge to rehabilitation or skilled nursing facility, and diagnosis of cancer, brain tumor, or cerebrovascular disease. These were used to design and validate the Neuro R2 score to predict 30 days' readmissions. Majersik noted that this score can help neurologists predict unplanned readmissions, a quality metric monitored by the Centers for Medicare and Medicaid Services, and recommended focusing attention on reducing readmissions for patients with brain tumors and those discharged to skilled nursing facilities. Hunter was more skeptical of the benefits of this score and commented that the utility of risk stratification algorithms is limited because of implementation barriers, inaccurate predictions, and ineffective risk reduction strategies. For Peacock et al., Freeman responded that the Neuro R2 score is one of the largest multivariate models for 30-day unplanned neurologic readmissions but that it certainly would not predict (or prevent) all 30 days' unplanned neurologic readmissions. He further noted that external validation of this model, and the creation of additional models focused on neurologic patients, is needed.