Improving Intensive Care Medication Safety through EHR-basedAlgorithms.

点击翻译按钮获取中文摘要

 DESCRIPTION (provided by applicant): In the field of patient safety, the paucity of systematic research is a critical barrier to progress. Notably missing are studies that meticulously investigate Electronic Health Records (EHR) and information technology in detecting intensive care-related errors. Our proposed study seeks to address an identified gap in the current knowledge of safety research by evaluating the usefulness of commercial IT systems and EHRs in reducing medical errors. In our study we seek to shift medication safety research from retrospective error identification towards a real-time automated and computerized approach to achieve a more comprehensive patient safety paradigm. The central hypothesis of our work is that by identifying discrepancies between medication order and administration data sources, we can detect and mitigate medication-related errors. In our study, we will 1) Use real time analysis to detect and intercept medication and parenteral nutrition (PN) administration errors identified by our recently developed Electronic Health Record (EHR) content-based algorithms (Aim 1); 2) Confirm performance of the algorithms in an external institution (Aim 2); and 3) Create new algorithms to identify smart pump infusion errors (Aim 3). By systematically detecting and intercepting medication and PN errors, we will shift medication safety from passive reporting of errors to proactive identification and mitigation of unsafe care. In Aim 1 we will reduce the time patients are at risk for harm through prospective identification o ameliorable medication and PN administration errors using CCHMC-developed medication administration error (MAE) detection algorithms. Using our EHR-based algorithms, we will detect administration errors in real-time and notify clinicians to decrease the time patients are a risk for harm. In Aim 2, we will evaluate the generalizability of the CCHMC-developed EHR-based medication administration error (MAE) detection algorithms by applying the algorithms to retrospective NICU and MICU data at an external institution. We will also develop a user-friendly demonstration package to facilitate usability of the algorithms and enhance the ability to observe their benefits. In Aim 3, we will develop novel algorithms to detect errors in smart pump use and evaluate system-level factors that contribute to pump errors. By detecting smart pump errors, the final step in medication and fluid administration, we will further reduce the rates of dangerous administration errors targeted in Aims 1 and 2. Our proposed work has the potential to accomplish a paradigm shift in the methods of patient safety research and clinical practice. The study is a fundamental step towards automating patient safety monitoring on a large scale and improving error identification and patient safety in the intensive care environment for millions of patients every year.