Reducing Missed Laboratory Test Results Sureyya Tarkan, Catherine Plaisant, Ben Shneiderman UMD Department of Computer Science & HCIL sureyya@cs.umd.edu http://www.cs.umd.edu/hcil/sharp/
‘Missed Results’ Problem Mishandling of abnormal results & no follow-up (Wahls, 2007) Most common missing information: test results (Dovey, 2002) Delay in care, time/financial loss, pain/suffering, adverse clinical consequence (Hickner, 2008)
Test Process Steps Pre-analytic Analytic Post-analytic Ordering the test Implementing the test Analytic Performing the test Post-analytic Reporting results to the clinician Responding to the results Notifying the patients of the results Following-up to ensure the patient took the appropriate action based on test results Specimen lost Patient & Physician Specimen damaged during transport Lab technician Patient doesn’t go Specimen mislabeled in office Physician Results misplaced No documentation on record Nurse Provider on vacation
Motivation: Package tracking Clear expectation Constant feedback Explicit responsibility
Approach for Reducing Missed Laboratory Results Model lab tracking processes Assign temporal responsibility Derive user interface from processes From order time to until all follow-up actions are taken Support actor actions as part of tracking Provide retrospective analyses to identify common problems
Simple Model: Generic Test Process ACTOR: Patient TASK: Get Test Done START: 0 END: 6 UNIT: days ACTOR: Provider TASK: Analyze Result END: 1 ACTION: Access Report, Inform Patient, Schedule Visit, Repeat Test, Order New Test…
Extended Model: Laboratory Test Schedule Exam & Go to Appointment (4 days) Accept Patient (1 hour) Confirm Appointment Update Patient Info? Draw & Store Sample (1 hour) Prepare Equipment Send Sample (20 hours) Ship with Carrier or Transport by self Examine Specimen & Record Results (1 day) Access Patient Sample Conduct Test Write Results
Modeling Compound Tests Example: Strep Throat Rapid Test: Office Test (1 hour) Culture: Laboratory Test (1 week)
Multi-Step Task Alerting, Reminding, Tracking (MSTART) Prototype
Track November 14, 2018
Generating Actor Action Sheets: Result Review
Order November 14, 2018
Retrospective Analysis
Conclusion: Guidelines Make temporal responsibility visible Tracking should show what is pending (and how much time has passed since last action) Prioritize late and critical (abnormal) entries Integrate review of results with possible actor actions from within same screen Retrospective analysis improves accountability Acknowledgments Zach Hettinger, Daniel Murphy, Archana Laxmisan, Hardeep Singh, Dean Sittig Strategic Health-IT Advanced Research Projects (SHARP) Contact: sureyya@cs.umd.edu Website: http://www.cs.umd.edu/hcil/sharp/