|
Context:
Labs are advised to streamline their processes before automating so they don't automate bad processes. Elmhurst's reference lab analyzed ordering and sample collection processes to reduce errors that occur before samples reach the lab.
Errors in information physicians receive from labs and delays in reporting can critically harm patients. Many lab errors result from manual steps performed by technologists. Automation decreases errors by reducing manual steps and stressful work conditions.
Objective:
Improve patient safety by reducing errors in test ordering, sample collection, lab processes.
Methods:
• Design: NA
• Setting: Elmhurst Memorial Healthcare system
(Elmhurst, Ill.)
• Participants: NA
• Intervention(s): Process re-engineering/laboratory
automation to eliminate manual steps, increase
standardization.
• Main outcome measure(s): Process measures.
Results:
Autoverification: 80-85% of more than 1 million annual chemistry/hematology results autoverified. TAT: Pre-automation, met goal of getting in-house test results to physicians for morning rounds 55% of time. Post-automation, goal is met 96% of time. Pre-automation, lab delivered timely troponin and myoglobin results to ED physicians 72% of time; post-automation, 90% of time. Pre-automation, met BMP TAT goal 58% of time. Post-automation, 94% of time.
Examples of improvements in ordering process (Feb. 2004 vs. Oct. 2004): On individual chart orders, duplicate orders decreased from 31% to 1%, orders signed by MDs increased from 86% to 100%, and orders requiring clarification decreased from 21% to 1%.
Conclusions:
Improved TAT and reporting results to clinicians reduces delays in treatment. Errors and error potential can be eliminated by analyzing processes, standardizing processes, minimizing variability, and eliminating manual steps. These same measures should be applied to the three phases of lab testing. As many manual steps as possible should be automated.
See more of Poster Session
See more of The 2005 Institute for Quality in Laboratory Medicine Conference