64 Assessing the value of a trainee-driven manual review process for reference testing

Laboratory testing that is sent out to reference laboratories can be costly, requires effort to process, and may not provide a result fast enough to have the intended impact on medical decision-making. Many laboratories have established oversight over this area to prevent low-value testing while helping clinicians navigate an increasingly complex set of available reference tests. Previous studies have found that, although time-consuming, a manual review process can decrease expenditures and improve communication between clinicians and the laboratory. This study aims to evaluate the cost-benefit, efficiency, educational value, and trainee opinion of a pathology trainee-driven manual review process for send-out approval at a large academic institution. Methods At our institution, 316 chemistry and 83 microbiology reference tests are on approval; these requests are reviewed by pathology lab medicine residents (LMRs) or fellows prior to send out. This process involves review of patient history, specific testing requirements, and discussion of clinical utility with ordering providers. A retrospective review of send-out testing was performed from July 2023 through June 2024, using the internal-use-only LMR call log as the data source, which included cancellation rates and costs per test. Costs used were the direct expense charged by referral laboratories per test. To identify tests with significantly higher or lower cancellation rates relative to others, a hypergeometric distribution was used to model random variation in cancellation rates assuming a similar underlying test cancellation rate. The Benjimini-Hochberg method was used for false discovery rate correction, and significance was set at corrected p <.05. An anonymous survey was sent to LMRs, consisting of multiple choice, Likert-scale response, and free-text questions about this process. Results Over the study period, LMRs reviewed 1,597 requests for 269 unique chemistry or microbiology send-out tests, with an overall cancellation rate of 35%. This conferred a $108,812, or 23.6%, savings. Outliers analysis identified Autoimmune Encephalopathy, CSF and Serum; and Broad Range 16S and 18S Sequencing as having significantly lower cancellation rates relative to all tests. T3, Reverse; Arbovirus Panel, IgG; West Nile Virus, IgG; Alkaline Phosphatase, Bone Specific; and Hepatitis B e Antibody had significantly higher cancellation rates. Most LMRs (89%) reported spending 10-30 minutes per low-complexity request; for high-complexity requests, most (67%) responded 1-2+ hours. 78% agreed this process was valuable to their education. Only 33% thought clinicians were receptive to it. Through free-text response, LMRs shared that time spent in this process is not always proportional to the benefits, and could be optimized through more automation. Conclusion A trainee-driven manual review of send-out testing is an effective yet time-consuming process to reduce unnecessary testing and expenditures. However, further work is needed to evaluate more efficient lab stewardship methods for high-volume reference tests with unexpectedly high or low cancellation rates.