Search

OBJECTIVES: Electroencephalogram features predict neurologic recovery following cardiac arrest. Recent work has shown that prognostic implications of some key electroencephalogram features change over time. We explore whether time dependence exists for an expanded selection of quantitative electroencephalogram features and whether accounting for this time dependence enables better prognostic predictions. DESIGN: Retrospective. SETTING: ICUs at four academic medical centers in the United States. PATIENTS: Comatose patients with acute hypoxic-ischemic encephalopathy.None. MEASUREMENTS AND MAIN RESULTS: We analyzed 12,397 hours of electroencephalogram from 438 subjects. From the electroencephalogram, we extracted 52 features that quantify signal complexity, category, and connectivity. We modeled associations between dichotomized neurologic outcome (good vs poor) and quantitative electroencephalogram features in 12-hour intervals using sequential logistic regression with Elastic Net regularization. We compared a predictive model using time-varying features to a model using time-invariant features and to models based on two prior published approaches. Models were evaluated for their ability to predict binary outcomes using area under the receiver operator curve, model calibration (how closely the predicted probability of good outcomes matches the observed proportion of good outcomes), and sensitivity at several common specificity thresholds of interest. A model using time-dependent features outperformed (area under the receiver operator curve, 0.83 ± 0.08) one trained with time-invariant features (0.79 ± 0.07; p < 0.05) and a random forest approach (0.74 ± 0.13; p < 0.05). The time-sensitive model was also the best-calibrated. CONCLUSIONS: The statistical association between quantitative electroencephalogram features and neurologic outcome changed over time, and accounting for these changes improved prognostication performance.

OBJECTIVES: Electroencephalogram features predict neurologic recovery following cardiac arrest. Recent work has shown that prognostic implications of some key electroencephalogram features change over time. We explore whether time dependence exists for an expanded selection of quantitative electroencephalogram features and whether accounting for this time dependence enables better prognostic predictions. DESIGN: Retrospective. SETTING: ICUs at four academic medical centers in the United States. PATIENTS: Comatose patients with acute hypoxic-ischemic encephalopathy.None. MEASUREMENTS AND MAIN RESULTS: We analyzed 12,397 hours of electroencephalogram from 438 subjects. From the electroencephalogram, we extracted 52 features that quantify signal complexity, category, and connectivity. We modeled associations between dichotomized neurologic outcome (good vs poor) and quantitative electroencephalogram features in 12-hour intervals using sequential logistic regression with Elastic Net regularization. We compared a predictive model using time-varying features to a model using time-invariant features and to models based on two prior published approaches. Models were evaluated for their ability to predict binary outcomes using area under the receiver operator curve, model calibration (how closely the predicted probability of good outcomes matches the observed proportion of good outcomes), and sensitivity at several common specificity thresholds of interest. A model using time-dependent features outperformed (area under the receiver operator curve, 0.83 ± 0.08) one trained with time-invariant features (0.79 ± 0.07; p < 0.05) and a random forest approach (0.74 ± 0.13; p < 0.05). The time-sensitive model was also the best-calibrated. CONCLUSIONS: The statistical association between quantitative electroencephalogram features and neurologic outcome changed over time, and accounting for these changes improved prognostication performance., SCOPUS: ar.j, info:eu-repo/semantics/published

Lidia MVR Moura,1 Eli L Schwamm,1 Valdery Moura Junior,1 Michael P Seitz,1 Daniel B Hoch,1 John Hsu,2,3 Lee H Schwamm1 1Department of Neurology, 2Mongan Institute for Health Policy, Massachusetts General Hospital, 3Department of Medicine and Department of Health Care Policy, Harvard Medical School, Boston, MA, USA Objective: Many effective medical therapies are available for treating neurological diseases, but these therapies tend to be expensive and adherence is critical to their effectiveness. We used patient-reported data to examine the frequency and determinants of financial barriers to medication adherence among individuals treated for neurological disorders. Patients and methods: Patients completed cross-sectional surveys on iPads as part of routine outpatient care in a neurology clinic. Survey responses from a 3-month period were collected and merged with administrative sources of demographic and clinical information (eg, insurance type). We explored the association between patient characteristics and patient-reported failure to refill prescription medication due to cost in the previous 12 months, termed here as “nonadherence”. Results: The population studied comprised 6075 adults who were presented between July and September 2015 for outpatient neurology appointments. The mean age of participants was 56 (standard deviation: 18) years, and 1613 (54%) were females. The patients who participated in the surveys (2992, 49%) were comparable to nonparticipants with respect to gender and ethnicity but more often identified English as their preferred language (94% vs 6%, p<0.01). Among respondents, 9.8% (n=265) reported nonadherence that varied by condition. These patients were more frequently Hispanic (16.7% vs 9.8% white, p=0.01), living alone (13.9% vs 8.9% cohabitating, p<0.01), and preferred a language other than English (15.3% vs 9.4%, p=0.02). Conclusion: Overall, the magnitude of financial barriers to medica