The Incidence and Nature of Adverse Events During Inpatient Rehabilitation: A Retrospective Case Series.

OBJECTIVE: To describe and compare adverse event (AE) incidence, type, severity, and preventability in the Canadian inpatient rehabilitation setting. DESIGN: In this retrospective case series, AEs were identified through chart reviews from two Canadian academic tertiary post-acute care hospitals. AEs were characterized through descriptive statistics and compared using the Mantel-Haenszel and Fisher's exact tests. RESULTS: During the study period, one site (n = 120) had 28 AEs and an incidence of 9.7 (95% CI 6.1-13.3) per 1000 patient days, and the other (n = 48) had 15 AEs and an incidence of 13.9 (95% CI 6.9-21) per 1000 patient days (p = 0.82). The two sites differed significantly in AE type (p = 0.033) and preventability (p = 0.002) but not severity. The most common AE type was medication/intravenous fluids-related (16/28, 57%) at one site and patient incidents (e.g., falls, pressure ulcers) at the other. Four percent (1/28) of AEs were preventable at one site, and 53% (8/15) at another. Most AEs at both sites were mild in severity. CONCLUSIONS: AEs significantly differed in type and preventability between the two sites. These results suggest the importance of context and the need for an organization-specific and tailored approach when addressing patient safety in inpatient rehabilitation settings.

Deep learning segmentation of orbital fat to calibrate conventional MRI for longitudinal studies.

In conventional non-quantitative magnetic resonance imaging, image contrast is consistent within images, but absolute intensity can vary arbitrarily between scans. For quantitative analysis of intensity data, images are typically normalized to a consistent reference. The most convenient reference is a tissue that is always present in the image, and is unlikely to be affected by pathological processes. In multiple sclerosis neuroimaging, both the white and gray matter are affected, so normalization techniques that depend on brain tissue may introduce bias or remove biological changes of interest. We introduce a complementary procedure, image "calibration," the goal of which is to remove technical intensity artifacts while preserving biological differences. We demonstrate a deep learning approach to segmenting fat from within the orbit of the eyes on T1-weighted images at 1.5 and 3 â€‹T to use as a reference tissue, and use it to calibrate 1018 scans from 256 participants in a study of pediatric-onset multiple sclerosis. The machine segmentations agreed with the adjudicating expert (DF) segmentations better than did those of other expert humans, and calibration resulted in better agreement with semi-quantitative magnetization transfer ratio imaging than did normalization with the WhiteStripe1 algorithm. We suggest that our method addresses two key priorities in the field: (1) it provides a robust option for serial calibration of conventional scans, allowing comparison of disease change in persons imaged at multiple time points in their disease; and (ii) the technique is fast, as the deep learning segmentation takes only 0.5 â€‹s/scan, which is feasible for both large and small datasets.