We address the problem of computing approximate answers to continuous sliding-window joins over data streams when the available memory may be insufficient to keep the entire join state. One approximation scenario is to provide a maximum subset of the result, with the objective of losing as few result tuples as possible. An alternative scenario is to provide a random sample of the join result, e.g., if the output of the join is being aggregated. We show formally that neither approximation can be addressed effectively for a slidingwindow join of arbitrary input streams. Previous work has addressed only the maximum-subset problem, and has implicitly used a frequencybased model of stream arrival. We address the sampling problem for this model. More importantly, we point out a broad class of applications for which an age-based model of stream arrival is more appropriate, and we address both approximation scenarios under this new model. Finally, for the case of multiple joins being executed with an overall memory constraint, we provide an algorithm for memory allocation across the joins that optimizes a combined measure of approximation in all scenarios considered. All of our algorithms are implemented and experimental results demonstrate their effectiveness.

Background: Electronic health records represent a large data source for outcomes research, but the majority of EHR data is unstructured (e.g. free text of clinical notes) and not conducive to computational methods. While there are currently approaches to handle unstructured data, such as manual abstraction, structured proxy variables, and model-assisted abstraction, these methods are time-consuming, not scalable, and require clinical domain expertise. This paper aims to determine whether selective prediction, which gives a model the option to abstain from generating a prediction, can improve the accuracy and efficiency of unstructured clinical data abstraction. Methods: We trained selective prediction models to identify the presence of four distinct clinical variables in free-text pathology reports: primary cancer diagnosis of glioblastoma (GBM, n = 659), resection of rectal adenocarcinoma (RRA, n = 601), and two procedures for resection of rectal adenocarcinoma: abdominoperineal resection (APR, n = 601) and low anterior resection (LAR, n = 601). Data were manually abstracted from pathology reports and used to train L1-regularized logistic regression models using term-frequency-inverse-document-frequency features. Data points that the model was unable to predict with high certainty were manually abstracted. Findings: All four selective prediction models achieved a test-set sensitivity, specificity, positive predictive value, and negative predictive value above 0.91. The use of selective prediction led to sizable gains in automation (anywhere from 57% to 95% reduction in manual abstraction of charts across the four outcomes). For our GBM classifier, the selective prediction model saw improvements to sensitivity (0.94 to 0.96), specificity (0.79 to 0.96), PPV (0.89 to 0.98), and NPV (0.88 to 0.91) when compared to a non-selective classifier. Interpretation: Selective prediction using utility-based probability thresholds can facilitate unstructured data extraction by giving "easy" charts to a model and "hard" charts to human abstractors, thus increasing efficiency while maintaining or improving accuracy.

Patient-centered organisationsHealthcare organisations now integrate patient feedback into value-based compensation formulas. This research considered Stanford Healthcare’s same-day feedback, a programme designed to evaluate the patient experience. Specifically, how did patients with cancer interviewed in the programme assess their physicians? Furthermore, how did assessments differ across emotional, physical, practical and informational needs when interviewed by volunteer patient and family partners (PAFPs) versus hospital staff?Patient–physician communication barriersIntegral to this research was Communication Accommodation Theory (CAT), which suggests individuals adjust interactions based on conversational roles, needs and understanding. Previous influential research was conducted by Frosch et al (2012) and Di Bartolo et al (2017), who revealed barriers to patient–physician communication, and Baker et al (2011) who associated CAT with these interactions. However, we still did not know if patients alter physician assessments between interviewers.Volunteers collect patient needsThis mixed methods study worked with 190 oncology unit patient interviews from 2009 to 2017. Open-ended interview responses underwent thematic analysis. When compared with hospital staff, PAFPs collected more practical and informational needs from patients. PAFPs also collected more verbose responses that resembled detailed narratives of the patients’ hospital experiences. This study contributed insightful patient perspectives of physician care in a novel hospital programme.

Medical education, research, and health care practice continue to grow with minimal coproduction guidance. We suggest the Commons Principle approach to medical education as modeled by Ostrom and Williamson, where we share how adapting these models to multiple settings can enhance empathy, increase psychological safety, and provide robust just-in-time learning tools for practice. We here describe patient and public coproduction in diverse areas within health care using the commons philosophy across populations, cultures, and generations with learning examples across age groups and cultures. We further explore descriptive, mixed methods participatory action in medical and research education. We adopt an “Everyone Included” perspective and sought to identify its use in continuing medical education, citizen science, marginalized groups, publishing, and student internships. Overall, we outline coproduction at the point of need, as we report on strategies that improved engagement. This work demonstrates coproduction with the public across multiple settings and cultures, showing that even with minimal resources and experience, this partnership can improve medical education and care.