Machine Learning-Based Prediction of Malnutrition in Surgical In-Patients: A Validation Pilot Study.

BACKGROUND: Malnutrition in hospitalised patients can lead to serious complications, worse patient outcomes and longer hospital stays. State-of-the-art screening methods rely on scores, which need additional manual assessments causing higher workload. OBJECTIVES: The aim of this prospective study was to validate a machine learning (ML)-based approach for an automated prediction of malnutrition in hospitalised patients. METHODS: For 159 surgical in-patients, an assessment of malnutrition by dieticians was compared to the ML-based prediction conducted in the evening of admission. RESULTS: The model achieved an accuracy of 83.0% and an AUROC of 0.833 in the prospective validation cohort. CONCLUSION: The results of this pilot study indicate that an automated malnutrition screening could replace manual screening tools in hospitals.

Pharmacokinetics Versus In Vitro Antiproliferative Potency to Design a Novel Hyperglycosylated hIFN-α2 Biobetter.

PURPOSE: IFN4N is a glycoengineered version of recombinant human interferon alpha 2 (rhIFN-α2) that was modified to exhibit four N-glycosylation sites. It shows reduced in vitro specific biological activity (SBA) mainly due to R23 mutation by N23. However, it has improved pharmacokinetics and led to a high in vivo antitumor activity in mice. In order to prepare a new IFN-based biobetter, this work compares the influence of glycosylation (affecting pharmacokinetics) with the in vitro antiproliferative SBA on the in vivo efficacy. METHODS: Based on IFN4N, three groups of muteins were designed, produced, and characterized. Group A: variants with the same glycosylation degree (4N) but higher in vitro antiproliferative SBA (R23 restored); group B: muteins with higher glycosylation degree (5N) but similar in vitro antiproliferative activity; and group C: variants with improved glycosylation (5N and 6N) and in vitro antiproliferative bioactivity. RESULTS: Glycoengineering was successful for improving pharmacokinetics, and R23 restoration considerably increased in vitro antiproliferative activity of new muteins compared to IFN4N. Hyperglycosylation was able to improve the in vivo efficacy similarly to or even better than R23 restoration. Additionally, the highest glycosylated mutein exhibited the lowest immunogenicity. CONCLUSIONS: Hyperglycosylation constitutes a successful strategy to prepare a novel IFN biobetter.

Depression, neuroticism and 2D:4D ratio: evidence from a large, representative sample.

A body of literature reports higher rates of depression and neuroticism in female samples compared to male samples. Numerous studies have investigated the role of prenatal sex hormone exposure in this sex difference, using the ratio between the second and fourth digit of the hand ("2D:4D") as a putative marker. However, the sample sizes of those studies were mostly small and results remained inconclusive. The aim of the present study is to test the suggested associations between depression, neuroticism and the 2D:4D ratio in a large, representative sample of over 3,000 German individuals. It was hypothesized that a higher 2D:4D (supposedly representing a more "feminine" prenatal hormone exposure) would positively predict (1) one's history of depression as well as (2) neuroticism rates and (3) acute depressive symptom scores. Controlling for biological sex, we only found suggestive evidence for linear associations with neuroticism in the case of left hand 2D:4D ratios and the mean 2D:4D of both hands. However, additional analyses indicated that these results may have been spurious due to confounding. Our findings suggest that the 2D:4D ratio is not a relevant predictor of depression, while there was mixed evidence in the case of neuroticism.