Electronic health record-based triggers to detect potential delays in cancer diagnosis
This is a summary of Murphy et al.'s 2014 article "Electronic health record-based triggers to detect potential delays in cancer diagnosis."
This article describes the development of electronic algorithms (called “triggers”) to find patients who may be lost to follow-up for colorectal or prostate cancer.
The purpose of this work is to eventually develop a more automated means (using queries) to keep track of patients who may be “falling through the cracks” and not receiving follow-up or treatment for serious but treat-able cancers.
A group of physicians developed the clinical criteria for each algorithm, and then using EHR data from the data warehouses of two difference healthcare systems, the clinical criteria were translated into computer logic, and the resulting queries (“triggers”) were run on two EHR data repositories over specified time periods. The patient charts that the queries “triggered” (as potentially lost to follow-up) were then reviewed by clinician chart reviewers to determine the true accuracy of the algorithms.
The trigger algorithms for both prostate and colorectal cancer were found to have positive predictive values (meaning the “true positive” triggered charts for both cancers) between 58% and 70%, making the use of the triggers more efficient than the typical random chart reviews performed for quality reasons.
This article both makes the case for and demonstrates the successful development and use of implementing a computer-driven method to surveil patient safety and quality, which is more efficient than current methodologies (chart reviews).
I chose this article because this area of research is the focus of the team with whom I work, and through my work, I have come to believe that trigger-like tools are the future of patient safety and quality measurement.
Trigger tools can be utilized in data mining as well. In Radiology, critical results communication must be tracked to ensure timely communication to requesting physicians. Radiologists need to be properly trained to make sure this tool is utilized to its maximum potential.
- Electronic health record-based triggers to detect potential delays in cancer diagnosis. Murphy, D. R., Laxmisan, A., Reis, B. A., Thomas, E. J., Esquivel, A., Forjuoh, S. N., … Singh, H. BMJ Qual Saf. 2014 Jan;23(1):8-16. doi: 10.1136/bmjqs-2013-001874. Epub 2013 Jul 19. http://www.ncbi.nlm.nih.gov/pubmed/23873756