CODA: an open-source platform for federated analysis and machine learning on distributed healthcare data.

OBJECTIVES: Distributed computations facilitate multi-institutional data analysis while avoiding the costs and complexity of data pooling. Existing approaches lack crucial features, such as built-in medical standards and terminologies, no-code data visualizations, explicit disclosure control mechanisms, and support for basic statistical computations, in addition to gradient-based optimization capabilities. MATERIALS AND METHODS: We describe the development of the Collaborative Data Analysis (CODA) platform, and the design choices undertaken to address the key needs identified during our survey of stakeholders. We use a public dataset (MIMIC-IV) to demonstrate end-to-end multi-modal FL using CODA. We assessed the technical feasibility of deploying the CODA platform at 9 hospitals in Canada, describe implementation challenges, and evaluate its scalability on large patient populations. RESULTS: The CODA platform was designed, developed, and deployed between January 2020 and January 2023. Software code, documentation, and technical documents were released under an open-source license. Multi-modal federated averaging is illustrated using the MIMIC-IV and MIMIC-CXR datasets. To date, 8 out of the 9 participating sites have successfully deployed the platform, with a total enrolment of >1M patients. Mapping data from legacy systems to FHIR was the biggest barrier to implementation. DISCUSSION AND CONCLUSION: The CODA platform was developed and successfully deployed in a public healthcare setting in Canada, with heterogeneous information technology systems and capabilities. Ongoing efforts will use the platform to develop and prospectively validate models for risk assessment, proactive monitoring, and resource usage. Further work will also make tools available to facilitate migration from legacy formats to FHIR and DICOM.

Intravenous Vitamin C in Adults with Sepsis in the Intensive Care Unit.

BACKGROUND: Studies that have evaluated the use of intravenous vitamin C in adults with sepsis who were receiving vasopressor therapy in the intensive care unit (ICU) have shown mixed results with respect to the risk of death and organ dysfunction. METHODS: In this randomized, placebo-controlled trial, we assigned adults who had been in the ICU for no longer than 24 hours, who had proven or suspected infection as the main diagnosis, and who were receiving a vasopressor to receive an infusion of either vitamin C (at a dose of 50 mg per kilogram of body weight) or matched placebo administered every 6 hours for up to 96 hours. The primary outcome was a composite of death or persistent organ dysfunction (defined by the use of vasopressors, invasive mechanical ventilation, or new renal-replacement therapy) on day 28. RESULTS: A total of 872 patients underwent randomization (435 to the vitamin C group and 437 to the control group). The primary outcome occurred in 191 of 429 patients (44.5%) in the vitamin C group and in 167 of 434 patients (38.5%) in the control group (risk ratio, 1.21; 95% confidence interval [CI], 1.04 to 1.40; P = 0.01). At 28 days, death had occurred in 152 of 429 patients (35.4%) in the vitamin C group and in 137 of 434 patients (31.6%) in the placebo group (risk ratio, 1.17; 95% CI, 0.98 to 1.40) and persistent organ dysfunction in 39 of 429 patients (9.1%) and 30 of 434 patients (6.9%), respectively (risk ratio, 1.30; 95% CI, 0.83 to 2.05). Findings were similar in the two groups regarding organ-dysfunction scores, biomarkers, 6-month survival, health-related quality of life, stage 3 acute kidney injury, and hypoglycemic episodes. In the vitamin C group, one patient had a severe hypoglycemic episode and another had a serious anaphylaxis event. CONCLUSIONS: In adults with sepsis receiving vasopressor therapy in the ICU, those who received intravenous vitamin C had a higher risk of death or persistent organ dysfunction at 28 days than those who received placebo. (Funded by the Lotte and John Hecht Memorial Foundation; LOVIT ClinicalTrials.gov number, NCT03680274.).

The evolution of mean arterial pressure in critically ill patients on vasopressors before and during a trial comparing a specific mean arterial pressure target to usual care.

BACKGROUND: In randomized clinical controlled trials, the choice of usual care as the comparator may be associated with better clinician uptake of the study protocol and lead to more generalizable results. However, if care processes evolve to resemble the intervention during the course of a trial, differences between the intervention group and usual care control group may narrow. We evaluated the effect on mean arterial pressure of an unblinded trial comparing a lower mean arterial pressure target to reduce vasopressor exposure, vs. a clinician-selected mean arterial pressure target, in critically ill patients at least 65 years old. METHODS: For this multicenter observational study using data collected both prospectively and retrospectively, patients were recruited from five of the seven trial sites. We compared the mean arterial pressure of patients receiving vasopressors, who met or would have met trial eligibility criteria, from two periods: [1] at least 1 month before the trial started, and [2] during the trial period and randomized to usual care, or not enrolled in the trial. RESULTS: We included 200 patients treated before and 229 after trial initiation. There were no differences in age (mean 74.5 vs. 75.2 years; p = 0.28), baseline Acute Physiology and Chronic Health Evaluation II score (median 26 vs. 26; p = 0.47) or history of chronic hypertension (n = 126 [63.0%] vs. n = 153 [66.8%]; p = 0.41). Mean of the mean arterial pressure was similar between the two periods (72.5 vs. 72.4 mmHg; p = 0.76). CONCLUSIONS: The initiation of a trial of a prescribed lower mean arterial pressure target, compared to a usual clinician-selected target, was not associated with a change in mean arterial pressure, reflecting stability in the net effect of usual clinician practices over time. Comparing prior and concurrent control groups may alleviate concerns regarding drift in usual practices over the course of a trial or permit quantification of any change.

Optimal VAsopressor TitraTION in patients 65 years and older (OVATION-65): protocol and statistical analysis plan for a randomised clinical trial.

INTRODUCTION: Vasodilatory hypotension is common among intensive care unit (ICU) patients; vasopressors are considered standard of care. However, optimal mean arterial pressure (MAP) targets for vasopressor titration are unknown. The objective of the Optimal VAsopressor TitraTION in patients 65 years and older (OVATION-65) trial is to ascertain the effect of permissive hypotension (vasopressor titration to achieve MAP 60-65 mm Hg) versus usual care on biomarkers of organ injury in hypotensive patients aged ≥65 years. METHODS AND ANALYSIS: OVATION-65 is an allocation-concealed randomised trial in 7 Canadian hospitals. Eligible patients are ≥65 years of age, in an ICU with vasodilatory hypotension, receiving vasopressors for ≤12 hours to maintain MAP ≥65 mm Hg during or after adequate fluid resuscitation, and expected to receive vasopressors for ≥6 additional hours. Patients are excluded for any of the following: active treatment for spinal cord or acute brain injury; vasopressors given solely for bleeding, ventricular failure or postcardiopulmonary bypass vasoplegia; withdrawal of life-sustaining treatments expected within 48 hours; death perceived as imminent; previous enrolment in OVATION-65; organ transplant within the last year; receiving extracorporeal life support or lack of physician equipoise. Patients are randomised to permissive hypotension versus usual care for up to 28 days. The primary outcome is high-sensitivity troponin T, a biomarker of cardiac injury, on day 3. Secondary outcomes include biomarkers of injury to other organs (brain, liver, intestine, skeletal muscle); lactate (a biomarker of global tissue dysoxia); resource utilisation; adverse events; mortality (90 days and 6 months) and cognitive function (6 months). Assessors of biomarkers, mortality and cognitive function are blinded to allocation. ETHICS AND DISSEMINATION: This protocol has been approved at all sites. Consent is obtained from the eligible patient, the substitute decision-maker if the patient is incapable, or in a deferred fashion where permitted. End-of-grant dissemination plans include presentations, publications and social media platforms and discussion forums. TRIAL REGISTRATION NUMBER: NCT03431181.

Lessening Organ dysfunction with VITamin C (LOVIT): protocol for a randomized controlled trial.

BACKGROUND: Sepsis is a health problem of global importance; treatments focus on controlling infection and supporting failing organs. Recent clinical research suggests that intravenous vitamin C may decrease mortality in sepsis. We have designed a randomized controlled trial (RCT) to ascertain the effect of vitamin C on the composite endpoint of death or persistent organ dysfunction at 28 days in patients with sepsis. METHODS: LOVIT (Lessening Organ dysfunction with VITamin C) is a multicenter, parallel-group, blinded (participants, clinicians, study personnel, Steering Committee members, data analysts), superiority RCT (minimum n = 800). Eligible patients have sepsis as the diagnosis for admission to the intensive care unit (ICU) and are receiving vasopressors. Those admitted to the ICU for more than 24 h are excluded. Eligible patients are randomized to high-dose intravenous vitamin C (50 mg/kg every 6 h for 96 h) or placebo. The primary outcome is a composite of death or persistent organ dysfunction (need for vasopressors, invasive mechanical ventilation, or new and persisting renal replacement therapy) at day 28. Secondary outcomes include persistent organ dysfunction-free days to day 28, mortality and health-related quality of life at 6 months, biomarkers of dysoxia, inflammation, infection, endothelial function, and adverse effects (hemolysis, acute kidney injury, and hypoglycemia). Six subgroup analyses are planned. DISCUSSION: This RCT will provide evidence of the effect of high-dose intravenous vitamin C on patient-important outcomes in patients with sepsis. TRIAL REGISTRATION: clinicaltrials.gov, NCT03680274, first posted 21 September 2018.