Machine learning to predict notes for chart review in the oncology setting: a proof of concept strategy for improving clinician note-writing

OBJECTIVE

Leverage electronic health record (EHR) audit logs to develop a machine learning (ML) model that predicts which notes a clinician wants to review when seeing oncology patients.

MATERIALS AND METHODS

We trained logistic regression models using note metadata and a Term Frequency Inverse Document Frequency (TF-IDF) text representation. We evaluated performance with precision, recall, F1, AUC, and a clinical qualitative assessment.

RESULTS

The metadata only model achieved an AUC 0.930 and the metadata and TF-IDF model an AUC 0.937. Qualitative assessment revealed a need for better text representation and to further customize predictions for the user.

DISCUSSION

Our model effectively surfaces the top 10 notes a clinician wants to review when seeing an oncology patient. Further studies can characterize different types of clinician users and better tailor the task for different care settings.

CONCLUSION

EHR audit logs can provide important relevance data for training ML models that assist with note-writing in the oncology setting.

Efficacy and Safety of Ensovibep for Adults Hospitalized With COVID-19 : A Randomized Controlled Trial

BACKGROUND

Ensovibep (MP0420) is a designed ankyrin repeat protein, a novel class of engineered proteins, under investigation as a treatment of SARS-CoV-2 infection.

OBJECTIVE

To investigate if ensovibep, in addition to remdesivir and other standard care, improves clinical outcomes among patients hospitalized with COVID-19 compared with standard care alone.

DESIGN

Double-blind, randomized, placebo-controlled, clinical trial. (ClinicalTrials.gov: NCT04501978).

SETTING

Multinational, multicenter trial.

PARTICIPANTS

Adults hospitalized with COVID-19.

INTERVENTION

Intravenous ensovibep, 600 mg, or placebo.

MEASUREMENTS

Ensovibep was assessed for early futility on the basis of pulmonary ordinal scores at day 5. The primary outcome was time to sustained recovery through day 90, defined as 14 consecutive days at home or place of usual residence after hospital discharge. A composite safety outcome that included death, serious adverse events, end-organ disease, and serious infections was assessed through day 90.

RESULTS

An independent data and safety monitoring board recommended that enrollment be halted for early futility after 485 patients were randomly assigned and received an infusion of ensovibep (n = 247) or placebo (n = 238). The odds ratio (OR) for a more favorable pulmonary outcome in the ensovibep (vs. placebo) group at day 5 was 0.93 (95% CI, 0.67 to 1.30; P = 0.68; OR > 1 would favor ensovibep). The 90-day cumulative incidence of sustained recovery was 82% for ensovibep and 80% for placebo (subhazard ratio [sHR], 1.06 [CI, 0.88 to 1.28]; sHR > 1 would favor ensovibep). The primary composite safety outcome at day 90 occurred in 78 ensovibep participants (32%) and 70 placebo participants (29%) (HR, 1.07 [CI, 0.77 to 1.47]; HR < 1 would favor ensovibep).

LIMITATION

The trial was prematurely stopped because of futility, limiting power for the primary outcome.

CONCLUSION

Compared with placebo, ensovibep did not improve clinical outcomes for hospitalized participants with COVID-19 receiving standard care, including remdesivir; no safety concerns were identified.

PRIMARY FUNDING SOURCE

National Institutes of Health.

Developing a pulse oximetry home monitoring protocol for patients suspected with COVID-19 after emergency department discharge

Objectives

Patients with COVID-19 can present to the emergency department (ED) without immediate indication for admission, but with concern for decompensation. Clinical experience has demonstrated that critical illness may present later in the disease course and hypoxia is often the first indication of disease progression. The objectives of this study are to (a) assess feasibility and describe a protocol for ED-based outpatient pulse-oximetry monitoring with structured follow-up and (b) determine rates of ED return, hospitalisation and hypoxia among participants.

Methods

Prospective observational study of patients presenting to a single academic ED in Boston with suspected COVID-19. Eligible patients were adults being discharged from the ED with presumed COVID-19. Exclusion criteria included resting oxygen saturation <92%, ambulatory oxygen saturation <90%, heart rate >110 beats per minute or inability to use the device. Study personnel made scripted phone calls on postdischarge days 1, 3 and 7 to review the pulse-oximetry readings and to evaluate for decompensation. Return visit and admission information were collected via medical record and 28-day follow-up calls.

Results

81 patients were enrolled of which 10 (12%) developed hypoxia after their initial discharge from the ED. Overall, 23 (28%) of the 81 patients returned to the ED at least once and 10 of those who returned (43%) were admitted. We successfully contacted 76/81 (94%) of subjects via phone at least once for follow-up assessment.

Discussion

Patients are eager and willing to participate in home monitoring systems and are comfortable with using technology, which will allow providers and health systems to extend our hospitals capabilities for tracking patient populations in times of crisis.

Conclusions

It is feasible to implement an outpatient pulse-oximetry monitoring protocol to monitor patients discharged from the ED with confirmed or suspected COVID-19.

Rare Complications of Acute Appendicitis: A Case Report

Introduction

Appendicitis is a common disease, and as we have improved in early diagnosis and management of this disease process, late stage complications have become extremely rare, but can have indolent presentations.

Case Report

A 37-year-old male with no past medical history presented to the emergency department (ED) with vague abdominal pain as well as 12 days of cyclical fever. He had no significant findings on laboratory workup with the exception of a mild aspartate transaminase and alanine transaminase and relative neutrophilia between outpatient, urgent care, and ultimate ED visit. His ED workup included cross-sectional imaging of his abdomen revealing multiple liver abscesses and septic thrombophlebitis secondary to ruptured appendicitis.

Conclusion

Liver abscesses and septic thrombophlebitis are an extremely rare complication of appendicitis that has only been documented twice previously.

Personalized cytomic assessment of vascular health: Evaluation of the vascular health profile in diabetes mellitus

BACKGROUND

An inexpensive and accurate blood test does not currently exist that can evaluate the cardiovascular health of a patient. This study evaluated a novel high dimensional flow cytometry approach in combination with cytometric fingerprinting (CF), to comprehensively enumerate differentially expressed subsets of pro-angiogenic circulating progenitor cells (CPCs), involved in the repair of vasculature, and microparticles (MPs), frequently involved in inflammation and thrombosis. CF enabled discovery of a unique pattern, involving both MPs and CPCs and generated a personalized signature of vascular health, the vascular health profile (VHP).

METHODS

Levels of CPCs and MPs were measured with a broad panel of cell surface markers in a population with atherosclerosis and type 2 diabetes mellitus (DM) and age-similar Healthy controls (HC) using an unbiased computational approach, termed CF.

RESULTS

Circulating hematopoietic stem and progenitor cell (CHSPCAng) levels were detected at significantly lower concentrations in DM (P < 0.001), whereas levels of seven phenotypically distinct MPs were present at significantly higher concentrations in DM patients and one MP subset was present at significantly lower concentration in DM patients. Collectively, the combination of CHSPC(Ang) and MP levels was more informative than any one measure alone.

CONCLUSIONS

This work provides the basis for a personalized cytomic vascular health profile that may be useful for a variety of applications including drug development, clinical risk assessment and companion diagnostics.

Plasma renin activity and aldosterone determination in patients with pseudotumor cerebri

A group of 6 women diagnosed as suffering from pseudotumor cerebri had measurement of their plasma renin activity and aldosterone levels at baseline, following plasma volume contraction by furosemide and after dexamethasone therapy. In each instance, the values recorded fell within the normal range and a normal renin-aldosterone axis is postulated.

Influence of vitamin D on bicarbonate reabsorption

The effects of acute administration of 25 hydroxycholecalciferol on bicarbonate reabsorption was studied in dogs. 25 hydroxycholecalciferol infused into the renal artery at a dose of 0.2 microgram/kg/hour led to a small but significant bilateral increase in bicarbonate reabsorption in intact dogs. 25 hydroxycholecalciferol had no effect on bicarbonate reabsorption in thyroparathyroidectomized dogs. The phosphaturia of bicarbonate administration was not altered by 25 hydroxycholecalciferol administration.

Effect of renin on extrarenal erythropoietin production

Extracts containing renal erythropoietic factor (REF) and others which contain renin reportedly enhance erythropoietin (Ep) titers in the plasma of hypoxic nephrectomized rats. Studies reported here were designed to elucidate the mechanism by which renin increases Ep production in the anephric rat. Injection of a renal extract containing renin significantly raised the blood pressure of anephric rats, and when it was injected just prior to exposure to hypoxia and 15 hour postnephrectomy, the resultant plasma Ep level exceeded that observed in rats exposed to the same hypoxic conditions immediately postnephrectomy. In contrast rats made hypoxic 15 hours after nephrectomy, but not given renin, had plasma Ep levels which were lower than those of anephric rats made hypoxic immediately postoperatively. When renin was injected immediately after nephrectomy or into normal rats, it failed to stimulate Ep production. When diazoxide was injected with the renin extract into hypoxic nephrectomized rats, the vasopressor effect of renin was abolished for 4 hours, and the plasma Ep levels were significantly lower than those of hypoxic nephrectomized animals injected only with renin, Injection of angiotensin II into anephric, hypoxic rats had an effect comparable to that of renin on extrarenal Ep roduction. REF was not detectable in the renin preparation. These results are best explained by the concept that renin and angiotensin increase extrarenal Ep production by causing vasoconstriction and consequently hypoxia in extrarenal sites of Ep production. Also of interest is the finding that plasma EP levels of rats, injected with renin and made hypoxic 15 hours postnephrectomy, are higher than those previously reported to occur in anephric rats.