“Intelligent automation”

In an effort that began in February of 2008, the hematology leadership team at the University of Iowa Hospital and Clinics, directed by Nancy Rosenthal, MD, transitioned its manually-driven laboratory into a highly efficient “intelligent automation” environment that has resulted in a single-tube touch point on every blood sample, auto-verification of 80% to 85% of samples, and a paper-free environment. These changes are significant, since the hematology lab’s patient population has more than 50% abnormal samples.

The University of Iowa Hospital and Clinics is an academic teaching facility that includes a medical school. It has a 690-bed inpatient facility, a 69-bed NICU, which boasts a 94% nationally-recognized survival rate, and a 20-bed PICU—all of which send samples to the core laboratory. Samples are also received from a liver transplant unit, and the facility recently made headlines with its first total artificial heart transplant. Also contributing to this significant and sophisticated workload is a new, self-contained cancer center which, unlike the previous cancer center, does not house a dedicated hematology laboratory specifically for these patients. Through a pneumatic tube system, the hematology department receives 100 to 150 cancer patient specimens from the new center daily.

Defining the problem

Like other laboratories across the country, this hematology department was facing increased workloads and more complex samples in the face of a declining medical technology staff. “It was obvious that we could no longer handle single specimens three to four times during processing. We needed to implement the highest level of automation,” says Mary Capper, MT(ASCP)SH, hematology supervisor and 25-year veteran of the lab. In 2007, the hematology lab’s workload included 5,000 CBCs, 12,000 CBCs with automated diffs, and 2,000 hemoglobin and hematocrits per month. This was a challenge for its existing system, whose instruments were 10 years old and had significant reliability problems. “I never really knew if we were going to have an instrument to run a test on,” says Capper. The hematology lab was also responsible for urinalysis studies and a complete battery of hemostasis testing that included thrombophilia workups, all factor assays, and routine prothrombin times and APTTs. Two technologists per day were assigned on a rotating basis to assist on bone marrow procedures, which appreciably increased from 80 to 130 per month.

The lab’s journey toward what Capper saw as its “ideal automated environment” began with the necessary physical alterations. Once the space was ready, the lab was able to build its high-speed automation line, which consisted of two high throughput hematology analyzers and a slidemaker/stainer. The laboratory also invested in a middleware software product, which would direct and guide the processing of all samples on the high throughput line according to rules that the laboratory developed to enhance safety and efficiency, considering its patient population and sample type mix.

“Without increasing our number of technologists, we have been able to withstand a significant increase in workload on the same complicated specimens we have always had,” says Capper.

Meeting the challenge

It was a challenge for technologists to go from a totally hands-on approach to a paperless, one-touch, walk-away approach. More than 40 technologists required training on the new system. Because the laboratorians were accustomed to touching the tube two to four times per sample, their learning curve was steep. “Imagine the process of having a sample come into the lab to have a CBC run and a smear made,” says Capper. “With the automated line, the sample is touched once as it is introduced onto the line. The line processes both initial and rerun/reflex sampling as well as smear preparation based on specific laboratory criteria.” The automation has been extended to smear review, where the operator is able to take the stained smears from the line and load them onto the digital image system for cell identification. Capper is very happy with a manual differential review rate of 25% to 27%, since the abnormal patient sample population is well over 50%. “Our aim was not to look at less but to look at those needing review,” she says.

Senior hematology technologists Roxanne Bohlke, MLS, CLA(ASCP), Dean Aman, MT(ASCP), and Pam Goldsmith, MT (ASCP), were instrumental in setting up the decision rules by which the middleware helped standardize the way samples and data were processed. This was no small endeavor, since the project began in 2008 when the implementation of hematology rules was in its infancy. “Previously, all we had was auto-verification for normal or close to normal values, not a big help with all those abnormal samples. We couldn’t build a rule that said, “Make a slide for specific doctors,” or “Reflex to reruns or smear making.” The ability to define such complex decisions in the software enabled us to achieve 80%-to-85% auto-verification, from less than 40%, and in a paper-free environment,” says Bohlke.

As the primary rules builder and tester of their middleware software, Bohlke needed to have a great deal of interaction with the middleware vendor’s support team. “We were so green,” she recalls. “We required a great deal of help and guidance with initial implementation, and continued receiving ongoing support after we went live. We began with about 160 rules that were refined as we started to gain confidence in the system and workflow. We now have about 100 rules in our software, providing streamlined sample and data processing. Considering we had 35 people to train on the new workflow, I think we did an excellent job. In the meantime, we got smarter and so did the manufacturer, and now we are able to efficiently update rules ourselves so even that aspect of the automation has improved considerably.”

The lab also encountered another challenge: the move of its Oncology Center and the integration of the oncology lab testing into its core lab. With this move came the relocation of existing equipment, a bench-top automation platform. The staff again met this challenge without hesitation and came up with another innovative approach. With a 69-bed NICU, microtainer samples are a way of life for this laboratory. While the initial intention was to repurpose the testing equipment for cancer center samples only, this resourceful team came up with another innovative idea. They decided to test the feasibility of dedicating this “alpha” bench-top system for just microtainer samples and body fluids, which disrupted workflow on their core laboratory equipment. This new approach has led to another efficient work area setup.

The team was not done yet. They decided to further streamline testing by looking at their approach to handling STAT samples. Now the hematology lab considers STAT samples only as those coming from the OR (for a liver transplant, for example) or on a critical patient. These samples are identified when hand delivered with verbal communication from specimen control personnel. On a routine basis, 97% of all CBCs + automated diffs are turned around in 15 minutes.

A model for your lab?

If a hematology laboratory is contemplating a leap to a fully automated setup to help it meet common challenges and discover the best way to approach its situation, its staff might consider the experiences the University of Iowa Hospital and Clinics hematology laboratory has shared. The right combination of automation and intelligent software, innovation, leadership, and teamwork has helped the Iowa laboratory achieve a paperless, ergonomic, and less stressful work environment for its staff.

Mary Capper sums it up best: “We are an academic center, so we believe in education. We do research and we do patient care. So we are pleased to share with other laboratories how we are doing things that can help increase the quality of care.”