A yellow, white, purple, and green molecule

The immunosuppressant tacrolimus is a mainstay of post-transplant maintenance therapy, so precisely measuring blood tacrolimus levels remains critical for preventing organ rejection and optimizing transplant recipients’ care. Despite the integral role of tacrolimus in transplant outcomes, its unique biological profile poses clinical challenges. In particular, blood tacrolimus levels can be quite variable due to this medication’s narrow therapeutic index and susceptibility to drug-drug interactions.

As the largest transplant center in the state of California, the University of California, Los Angeles (UCLA) Health System serves patients from throughout Southern California. To meet the needs of this large population, we have implemented a unique system that expedites transplant patients’ follow-up visits. Our patients have routine laboratory tests and see their providers on the same day—just a few hours after the tests are performed. This makes it imperative that we optimize turnaround time (TAT) for whole blood tacrolimus testing. Favorable TAT not only contributes to patient satisfaction and the efficiency of post-transplant care but also serves as a quality metric for our laboratory’s operational performance. Given these factors we used Lean techniques to streamline our workflow and TAT for tacrolimus measurements.

UCLA uses Abbott Architect analyzer for our whole blood tacrolimus testing. Prior to being analyzed, tacrolimus must be extracted from red blood cells (RBCs). This requires exposing the RBCs to a lysing reagent and centrifuging them. We previously identified this extraction step as the bottleneck in our testing workflow and published about a batched extraction method that uses metal batch racks. This method, however, is not applicable for small and medium-sized laboratories that do not process a large number of specimens and rely on a manual extraction method using a standard centrifuge. This led us to our Lean analysis, which sought to determine the optimal batch size for the manual extraction method that minimizes TAT for whole blood tacrolimus testing.

A New Protocol

Our whole blood tacrolimus measurement occurs in three steps. First, we expose whole blood specimens to Architect tacrolimus whole blood precipitation reagent. We then manually vortex and centrifuge the treated samples. Finally, we place the samples in a standard rack and load them into the Architect. The average time per rack in the analyzer is 23 minutes.

Prior to this new protocol, each performing technician would determine at his or her discretion the batch size for the manual extraction phase. For example, if our laboratory received 36 patient specimens, a technician could either perform manual extraction on all 36 specimens at the same time or split them into smaller batches. We hypothesized that batch size variability was the bottleneck in the preanalytical phase. Based on this hypothesis, for 2 months and four different batch sizes—random (technician dependent), 15, 20, and 21 specimens—we recorded TAT, defined as the time between our lab’s receipt of specimens to when we verified results. We capped the sample size at 21 because the Xsystems centrifuge, commonly used for tacrolimus testing, has a maximum sample capacity of 21 tubes. Whenever we received more samples than the selected batch size, our laboratory technicians were instructed to perform the extraction step for the remaining specimens at hand once the first batch was loaded into the Architect for analysis.

During the 2-month period we collected 1,361 data points, excluding from our analysis two obvious outliers (3 and 994 minutes). Our mean TAT at baseline was 103 minutes versus 100, 83, and 77 minutes post-intervention for the 15-, 20-, and 21-sample groups, respectively. We compared the means by Student’s t-test with a p-value <0.05 to indicate statistical significance. The mean TAT for the baseline group (103 minutes) was significantly higher than that for the 20-, and 21-sample groups (83 minutes, p<0.0001; 77 minutes, p<0.0001). However, there was no statistical difference between the baseline and 15-sample groups (100 minutes, p=0.49).

Laboratory TAT is an important metric that measures the quality and performance of a clinical laboratory. Aside from serving as a quality indicator, achieving short TAT is critical to the success of UCLA’s current patient care model in which provider visits occur just hours from when we collect patients’ blood samples to determine their tacrolimus levels.

No More Waiting

We demonstrated that standardizing batch sizes during the manual extraction phase of our testing significantly reduced our overall TAT. One possible reason for this reduction comes from improved time management. Working on extraction while the first sample batch is being analyzed increases the instrument utilization rate and improves TAT. This method eliminates waiting, which is one of the seven wastes of Lean principle.

Based on this finding, we recommend that labs performing whole blood tacrolimus testing maximize the number of samples loaded into a single rack and perform extraction on the remaining samples while their instrument is analyzing the first rack. This study illustrates an excellent use case for Lean, which enables laboratories to identify and resolve obstacles in specimen processing by standardizing and devising novel strategies for process improvement.

The authors gratefully acknowledge the participation of Nathan Okawa, ASCP, and Vincent Buggs, MS, in designing, planning, and analyzing this quality improvement initiative and in reviewing this article.

Shohei Ikoma, MD, is a clinical informatics and surgical pathology fellow in the Department of Pathology and Laboratory Medicine at University of California, Los Angeles. +Email: SIkoma@mednet.ucla.edu

Kathleen A. Kelly, PhD, DABCC, FAACC, MT (ASCP), is interim co-chief of clinical chemistry, director of special chemistry and toxicology, and professor in the Department of Pathology and Laboratory Medicine at University of California, Los Angeles. +Email: KKelly@mednet.ucla.edu