A complete patient problem list is the cornerstone of the problem-oriented medical record. It serves as a valuable tool for providers assessing a patient's clinical status and succinctly communicates this information between providers. Accurate problem lists drive clinical decision support tools that improve quality, and an accurate problem list has been associated with higher quality care. Accurate problem lists are also critical for establishing accurate phenotypes for research and supporting quality improvement;however, problem lists in electronic health records are routinely incomplete. In this study, we propose to develop and validate problem inference algorithms to identify problems potentially missing from patient problem lists, and to conduct a randomized trial of these algorithms, studying their effects on quality of care. We call our approach IQ-MAPLE. Our project has three specific aims: 1) develop problem inference algorithms for heart, lung, and blood conditions, 2) implement problem inference alerts and optimize the workflow at four sites and 3) conduct a randomized controlled trial of the problem inference alerts, measuring the acceptance rate of alerts, the direct effect on problem list completeness and, critically, downstream impact of the alerts on key clinical quality measures, including both process and outcomes across a range of heart, lung and blood conditions. If successful, IQ-MAPLE will improve problem list accuracy, which has significant downstream implications: More accurate clinical decision support: Most clinical decision support is disease-oriented and, as such, depends on an accurate problem list. When problems are missing, opportunities to provide support to the clinician are missed, and quality suffers. Better quality measurement: Quality measurement today is often inaccurate, as patients are omitted from measures due to incomplete information on their clinical problems. A more accurate problem list would, in turn, lead to more accurate quality measurement. More accurate research: Clinical, translational and even basic science and genomic research increasingly use EHR data, particularly to identify patients with disease phenotypes, generally using problem lists. When problems are missing, the accuracy of research suffers. Better patient care: Secondary benefits aside, the problem list is, fundamentally, a tool for organizing patient care and communicating among providers. A more accurate, problem list supports these goals. The IQ-MAPLE rules, and our best practices for implementing them, will be freely available, ensuring broad dissemination of these benefits.
An accurate, complete clinical problem list the cornerstone of a problem-oriented medical record, and research shows that accurate problem lists improve healthcare quality;however, problem lists are often incomplete. In this study, we will develop and evaluate an intervention to identify gaps in problem lists for patients with heart, lung, and blood conditions. If successful, this system should improve problem list completeness and, ultimately, the quality of care delivered to patients.
|Song, Wenyu; Huang, Hailiang; Zhang, Cheng-Zhong et al. (2018) Using whole genome scores to compare three clinical phenotyping methods in complex diseases. Sci Rep 8:11360|
|Wright, Adam; McCoy, Allison B; Hickman, Thu-Trang T et al. (2015) Problem list completeness in electronic health records: A multi-site study and assessment of success factors. Int J Med Inform 84:784-90|