Search

Secondary use of medical data and use of observational data for causal inference has been growing. Yet these data bring many challenges such as confounding due to unobserved variables and variation in medical processes across settings. Further, while methods exist to handle some of these problems, researchers lack ground truth to evaluate these methods. When a finding is not replicated across multiple sites, it is unknown whether this is a failure of an algorithm, a genuine difference between populations, or an artifact of structural differences between the sites. We show how agent-based simulation of medical interventions can be used to explore how bias, error, and variation across settings affect inference. Our approach enables users to model not only interventions and outcomes, but also the complex interaction between patients with different risks of mortality and providers with different observed and latent treatment effects. Ultimately we propose that such simulations can be used to better evaluate the behavior of new methods with known ground truth and better calculate sample size for EHR-based studies.

Communication and information sharing are critical parts of teamwork in the hospital; however, achieving open and fluid communication can be challenging. Finding specific patient information within documentation can be difficult. Recent studies on handoff documentation tools show that resident handoff notes are increasingly used as an alternative information source by non-physician clinicians. Previous findings also show that residents have become aware of this unintended use. This study investigated the alignment of resident note updating patterns and team note viewing patterns based on usage log data of handoff notes. Qualitative interviews with clinicians were used to triangulate findings based on the log analysis. The study found that notes that were frequently updated were viewed significantly more frequently than notes updated less often (p < 2.2 × 10(-16)). Almost 44% of all notes had aligned frequency of views and updates. The considerable percentage (56%) of mismatched note utilization suggests an opportunity for improvement.

Handoffs are a critical component of coordinated patient care; however, poor handoffs have been associated with near misses and adverse events. To address this, national agencies have recommended standardizing handoffs, for example through the use of handoff documentation tools. Recent research suggests that handoff tools, typically designed for physicians, are often used by non-physician providers as information sources. In this study, we investigated patterns of edits of an electronic handoff tool in a large teaching hospital through examination of its usage log data. Qualitative interviews with clinicians were used to triangulate log data findings. The analysis showed that despite its primary focus on facilitating transitions of care, information in the handoff documentation tool was updated throughout the day. Interviews with residents confirmed that they purposefully updated information to make it available for other members of their patient care teams. This further reiterates the view of electronic handoff tools as facilitators of team communication and coordination. However, the study also showed considerable variability in the frequency of updates between different units and across different patients. Further research is required to understand what factors drive such diversity in the use of electronic handoff tool and whether this diversity can be used to make inferences about patients' conditions.

We have developed an extensible, open source framework for collecting data from bedside medical devices which support RS-232 serial output. The system uses inexpensive hardware to convert RS-232 signals to Universal Serial Bus (USB) format, enabling "plug-and-play" device communication at low cost. A prototype of the system has been created and tested with Puritan-Bennett 840 ventilators.