Automatic Population of the Case Report Forms for an International Multifactorial Adaptive Platform Trial Amid the COVID-19 Pandemic.

OBJECTIVES: To automatically populate the case report forms (CRFs) for an international, pragmatic, multifactorial, response-adaptive, Bayesian COVID-19 platform trial. METHODS: The locations of focus included 27 hospitals and 2 large electronic health record (EHR) instances (1 Cerner Millennium and 1 Epic) that are part of the same health system in the United States. This paper describes our efforts to use EHR data to automatically populate four of the trial's forms: baseline, daily, discharge, and response-adaptive randomization. RESULTS: Between April 2020 and May 2022, 417 patients from the UPMC health system were enrolled in the trial. A MySQL-based extract, transform, and load pipeline automatically populated 499 of 526 CRF variables. The populated forms were statistically and manually reviewed and then reported to the trial's international data coordinating center. CONCLUSIONS: We accomplished automatic population of CRFs in a large platform trial and made recommendations for improving this process for future trials.

Lopinavir-ritonavir and hydroxychloroquine for critically ill patients with COVID-19: REMAP-CAP randomized controlled trial.

PURPOSE: To study the efficacy of lopinavir-ritonavir and hydroxychloroquine in critically ill patients with coronavirus disease 2019 (COVID-19). METHODS: Critically ill adults with COVID-19 were randomized to receive lopinavir-ritonavir, hydroxychloroquine, combination therapy of lopinavir-ritonavir and hydroxychloroquine or no antiviral therapy (control). The primary endpoint was an ordinal scale of organ support-free days. Analyses used a Bayesian cumulative logistic model and expressed treatment effects as an adjusted odds ratio (OR) where an OR > 1 is favorable. RESULTS: We randomized 694 patients to receive lopinavir-ritonavir (n = 255), hydroxychloroquine (n = 50), combination therapy (n = 27) or control (n = 362). The median organ support-free days among patients in lopinavir-ritonavir, hydroxychloroquine, and combination therapy groups was 4 (- 1 to 15), 0 (- 1 to 9) and-1 (- 1 to 7), respectively, compared to 6 (- 1 to 16) in the control group with in-hospital mortality of 88/249 (35%), 17/49 (35%), 13/26 (50%), respectively, compared to 106/353 (30%) in the control group. The three interventions decreased organ support-free days compared to control (OR [95% credible interval]: 0.73 [0.55, 0.99], 0.57 [0.35, 0.83] 0.41 [0.24, 0.72]), yielding posterior probabilities that reached the threshold futility (≥ 99.0%), and high probabilities of harm (98.0%, 99.9% and > 99.9%, respectively). The three interventions reduced hospital survival compared with control (OR [95% CrI]: 0.65 [0.45, 0.95], 0.56 [0.30, 0.89], and 0.36 [0.17, 0.73]), yielding high probabilities of harm (98.5% and 99.4% and 99.8%, respectively). CONCLUSION: Among critically ill patients with COVID-19, lopinavir-ritonavir, hydroxychloroquine, or combination therapy worsened outcomes compared to no antiviral therapy.

Hierarchical Adaptive Multi-task Learning Framework for Patient Diagnoses and Diagnostic Category Classification.

The problems a patient suffers from can be summarized in terms of a list of patient diagnoses. The diagnoses are typically organized in a hierarchy (or a lattice structure) in which many different low-level diagnoses are covered by one or more diagnostic categories. An interesting machine learning problem is related to learning of a wide range of diagnostic models (at different levels of abstraction) that can automatically assign a diagnosis or a diagnostic category to a specific patient. While one can always approach this problem by learning models for each diagnostic task independently, an interesting open question is how one can leverage the knowledge of a diagnostic hierarchy to improve the classification and outperform independent diagnostic models. In this work, we study this problem by designing a new hierarchical classification learning framework in which multiple diagnostic classification targets are explicitly related via diagnostic hierarchy relations. By conducting experiments on MIMIC-III data and ICD-9 diagnosis hierarchy, we demonstrate that our framework leads to improved classification performance on individual diagnostic tasks when compared to independently learned diagnostic models. This improvement is stronger for diagnoses with a low prior and smaller number of positive training examples.