Single-cell type annotation with deep learning in 265 cell types for humans.

Motivation: Annotating cell types is a challenging yet essential task in analyzing single-cell RNA sequencing data. However, due to the lack of a gold standard, it is difficult to evaluate the algorithms fairly and an overfitting algorithm may be favored in benchmarks. To address this challenge, we developed a deep learning-based single-cell type prediction tool that assigns the cell type to 265 different cell types for humans, based on data from approximately five million cells. Results: We achieved a median area under the ROC curve (AUC) of 0.93 when evaluated across datasets. We found that inconsistent labeling in the existing database generated by different labs contributed to the mistakes of the model. Therefore, we used cell ontology to correct the annotations and retrained the model, which resulted in 0.971 median AUC. Our study reveals a limiting factor of the accuracy one may achieve with the current database annotation and points to the solutions towards an algorithm-based correction of the gold standard for future automated cell annotation approaches. Availability and implementation: The code is available at: https://github.com/SherrySDong/Hierarchical-Correction-Improves-Automated-Single-cell-Type-Annotation. Data used in this study are listed in Supplementary Table S1 and are retrievable at the CZI database.

Perceived barriers to early goal-directed mobility in the intensive care unit: Results of a quality improvement evaluation.

BACKGROUND: A multicentre randomised trial demonstrated improved outcomes for intensive care unit (ICU) patients using early, goal-directed mobility implemented by nurses. OBJECTIVES: The aim of the study was to evaluate barriers to nursing mobility, using a validated survey, during an ongoing quality improvement (QI) project (2019) in a medical ICU and determine changes from the pre-QI (2017) baseline. METHODS: Nurses, nurse practitioners, physician assistants, and clinical technicians completed the 26-item Patient Mobilization Attitudes and Beliefs Survey for the ICU (PMABS-ICU). An overall score and three subscale scores (knowledge, attitudes, behaviour), each ranging from 0 to 100, were calculated; higher scores indicated greater barriers. RESULTS: Seventy-five (93% response rate) nurses, eight (100%) nurse practitioners and physician assistants, and 11 (100%) clinical technicians completed the PMABS-ICU. For all respondents (N = 94), the mean (standard deviation) overall PMABS-ICU score was 32 (8) and the knowledge, attitudes and behaviour subscale scores were 22 (11), 33 (11), and 34 (8), respectively. Among all respondents completing the survey in both 2017 and 2019 (N = 46), there was improvement in the mean (95% confidence interval) overall score [-3.1 (-5.8, -0.5); p = .022] and in the knowledge [-5.1 (-8.9, -1.3); p = .010] and attitudes [-3.9 (-7.3, -0.6); p = .023] subscale scores. Among all respondents (N = 48) taking the PMABS-ICU for the first time in 2019 compared with those taking the survey before the QI project in 2017 (N = 99), there was improvement in the mean (95% confidence interval) overall score [-3.8 (-6.5, -1.1); p = .007] and in the knowledge [-6.9 (-11.0, -2.7); p = .001] and attitude [-4.3 (-8.1, -0.5); p = .027] subscale scores. CONCLUSIONS: Using a validated survey administered to ICU nurses and other staff, before and during a structured QI project, there was a decrease in perceived barriers to mobility. Reduced barriers among those taking the survey for the first time during the QI project compared with those taking the survey before the QI project suggests a positive culture change supporting early, goal-directed mobility implemented by nurses.

Evaluation of the Environmental Scoring System in Multiple Child Asthma Intervention Programs in Boston, Massachusetts.

OBJECTIVES: To test the applicability of the Environmental Scoring System, a quick and simple approach for quantitatively measuring environmental triggers collected during home visits, and to evaluate its contribution to improving asthma outcomes among various child asthma programs. METHODS: We pooled and analyzed data from multiple child asthma programs in the Greater Boston Area, Massachusetts, collected in 2011 to 2016, to examine the association of environmental scores (ES) with measures of asthma outcomes and compare the results across programs. RESULTS: Our analysis showed that demographics were important contributors to variability in asthma outcomes and total ES, and largely explained the differences among programs at baseline. Among all programs in general, we found that asthma outcomes were significantly improved and total ES significantly reduced over visits, with the total Asthma Control Test score negatively associated with total ES. CONCLUSIONS: Our study demonstrated that the Environmental Scoring System is a useful tool for measuring home asthma triggers and can be applied regardless of program and survey designs, and that demographics of the target population may influence the improvement in asthma outcomes.

"We make the path by walking it": building an academic community partnership with Boston Chinatown.

BACKGROUND: The potential for academic community partnerships are challenged in places where there is a history of conflict and mistrust. Addressing Disparities in Asian Populations through Translational Research (ADAPT) represents an academic community partnership between researchers and clinicians from Tufts Medical Center and Tufts University and community partners from Boston Chinatown. Based in principles of community-based participatory research and partnership research, this partnership is seeking to build a trusting relationship between Tufts and Boston Chinatown. OBJECTIVES: This case study aims to provides a narrative story of the development and formation of ADAPT as well as discuss challenges to its future viability. METHODS: Using case study research tools, this study draws upon a variety of data sources including interviews, program evaluation data and documents. RESULTS: Several contextual factors laid the foundation for ADAPT. Weaving these factors together helped to create synergy and led to ADAPT's formation. In its first year, ADAPT has conducted formative research, piloted an educational program for community partners and held stakeholder forums to build a broad base of support. CONCLUSIONS: ADAPT recognizes that long term sustainability requires bringing multiple stakeholders to the table even before a funding opportunity is released and attempting to build a diversified funding base.