Search

Abstract Background Early post-discharge assessments for newborns are recommended. Virtual care has become more prevalent during the pandemic, providing an opportunity to better understand its impact on the quality of post-discharge newborn care. The objective of this study was to understand whether primary care visit modality (in-person vs. virtual) is associated with early newborn hospital readmissions and emergency department (ED) visits. Methods We conducted a population-based, case-control study using linked health administrative databases between September 1, 2020 and March 31, 2022 in Ontario, Canada. We compared the modality of primary care visits among cases (hospital readmission within 14 days of life) and controls (newborns without a readmission), matched on infant sex, gestational age, and maternal parity. We included an alternative definition of cases as a composite of either a newborn hospital readmission or emergency department (ED) visit or in-hospital death within the first 14 days of life. Conditional logistic regression models were used to model odds ratios (ORs), comparing those exposed to a virtual visit versus in-person visit, adjusting for infant birth weight, birth hospitalization length of stay, neighbourhood level material deprivation, rurality and presence of active maternal comorbidities. Results Among 73,324 eligible newborns, 2,220 experienced a hospital readmission within 14 days of life and were matched to 8,880 controls. Jaundice was the primary reason for readmission (75% of readmissions). Compared to newborns who were seen in-person post-discharge, newborns who were seen virtually had higher odds of hospital readmission (adjusted odds ratio [aOR] 1.41 (95% CI 1.09, 1.83); the magnitude of effect was not different using the composite outcome (aOR 1.35, 95% CI 1.05, 1.75). Conclusions Newborns who receive a virtual post-discharge visit are more likely than those who receive an in-person visit to require hospital readmission.

Stress tests, e.g., the cardiac stress test, are standard clinical screening tools aimed to unmask clinical pathology. As such stress tests indirectly measure physiological reserves. The term reserve has been developed to account for the dis-junction, often observed, between pathology and clinical manifestation. It describes a physiological capacity that is utilized in demanding situations. However, developing a new and reliable stress test based screening tool is complex, prolonged, and relies extensively on domain knowledge. We propose a novel distributional-free machine-learning framework, the Stress Test Performance Scoring (STEPS) framework, to model expected performance in a stress test. A performance scoring function is trained with measures taken during the performance in a given task while exploiting information regarding the stress test set-up and subjects' medical state. Multiple ways of aggregating performance scores at different stress levels are suggested and are examined with an extensive simulation study. When applied to a real-world data example, an AUC of 84.35[95%CI: 70.68 - 95.13] was obtained for the STEPS framework to distinguish subjects with neurodegeneration from controls. In summary, STEPS improved screening by exploiting existing domain knowledge and state-of-the-art clinical measures. The STEPS framework can ease and speed up the production of new stress tests.