College of American Pathologists
Printable Version

  Full automation as solution—U of Iowa’s story


CAP Today




September 2011
Feature Story

Sue Zaleski, SCT(ASCP)HT
Jeff Kulhavy, MT(ASCP)

The laboratory at the University of Iowa Hospitals and Clinics—a 735-bed health system—was like all other labs: facing continuing budget constraints, staffing shortages, competition for employees, turnaround time pressure, and an ever-present need to reduce label and aliquot errors due to human mistakes. Manual processes that were mundane and repetitive contributed to these errors and led to a high turnover rate and low employee morale.

We started planning for automation in 2005, and by 2007 a preanalytical solution had been installed and integrated into a series of chemistry analyzers. In 2008 we implemented a digital cell morphology system with automated differential and middleware in three lab locations. We’re fully automated today, but the process continues, with old chemistry analyzers being replaced and an archival/retrieval unit in our immediate plans. The latter will help the laboratory manage add-on orders and be an asset to our biobank initiative.

The UIHC experience is a practical example of how technology can advance the practice of laboratory medicine by making the process of generating results more efficient with no loss of accuracy.

The decision to automate was not made quickly, and it wasn’t taken lightly. UIHC administrators had to be confident that the technology would make financial sense and create a better environment for employees with no negative effect on patient care. Lab management presented a strong case to administration that focused on these core elements:

  • Patient safety. Accurate pathology tests are critical in ensuring that a physician can accurately diagnose and appropriately treat a patient. In a manual process, errors can occur at any of several steps. An error that is caught may require subjecting the patient to additional blood draws; an error that is not caught can have far more serious consequences.

  • Employee safety and morale. Manual testing presents a biohazard risk for staff. Automation takes the samples out of employees’ hands and reduces the risk. At the same time, it frees employees from redundant bench protocols and allows them to take on other responsibilities such as validating new tests. The resulting improvement in morale, along with the state-of-the art automation, means an employer-of-choice environment for prospective employees and those already employed.

  • Customer/medical staff satisfaction. Greater speed, accuracy, and consistency mean customers who are satisfied.

  • Financial. A more streamlined process and even workflow means a laboratory is able to run more samples without needing more staff.

  • Once the lab was given the go-ahead, the next step was to select a small team to lead the charge. This team consisted of seven hand-picked laboratory staff members who were diverse in their education, skills, and expertise. The deliberate exclusion of managers and administrators ensured that the team would consist of people who knew and were responsible for the work and would be the most engaged in the process. These “change agents” remained aligned in their mission by establishing and operating under three goals: 1) increase productivity, 2) improve employee morale, and 3) add value to the physicians, nurses, and patients.

    Selecting a vendor was next. UIHC sought out a vendor that could offer customizable solutions related to project management, middleware, and service support. Most important, though, was to partner with a company that could offer valuable expertise and help the lab evolve. The key to automation is to not automate a bad process. UIHC found that its partner, Roche Diagnostics, was (and still is) critical in evaluating every step along the way. For example, to determine the efficiencies of the preanalytic and analytic systems, Roche conducted a full systems audit. Roche collected data for 25 days on such parameters as collection to receipt and receipt to load on the preanalytical system, including event frequencies and error rates for loading and aliquot functions. Roche also evaluated analyzer reruns and repeat frequencies.

    Once processes were agreed upon, automation began to be rolled out in a stepwise fashion. Since there are employees working all day and night, the training had to take place around the clock. A rigorous process was put in place to ensure that all employees were educated before the full transition.

    With automation in place, staff members were freed up to take on other activities that by their very nature needed to be conducted by people. An example of this is the call center UIHC added to its laboratory. Now, when customers and medical staff contact the laboratory with inquiries and requests, the calls go directly to a dedicated laboratory staff member. Since the people who are working the bench no longer answer incoming calls, they do not have to navigate competing priorities and can focus on testing.

    The lab also takes advantage of push technology. For instance, add-on requests are printed in the call center where the staff can triage the requests. Push technology also helps to build standard work practices for situations that rarely occur. This can be seen in the rare event where the automation process does not autoverify a sample. A printout will appear and direct the person working the bench in exactly what needs to be done to verify the case. The printouts are color-coded according to priority and all steps are included. With this, the person working the bench no longer needs to commit lesser-known protocols to memory, which can create a possibility for error. This clarity allows the staff member to work more effectively with greater confidence in the procedure being used.

    Laboratory staff members and UIHC management and administration have celebrated the full automation of the lab as a great success. Productivity, efficiency, and reliability are greatly improved. Today, about 99 percent of samples are autoverified. Once a sample goes onto the automation highway it is out of human hands; therefore, labeling and aliquot errors have been eliminated.

    Now, there is no distinction between a routine and a stat test because all tests are handled in the same way and processed as quickly as possible. Turnaround time is shorter and highly predictable within set goals. This means far fewer calls to the laboratory.

    From a business perspective, the up-front investment has proved to be financially beneficial. In fact, billable test volume is up 30 percent from 2005 while, during the same period, total hours worked are down by 11 percent. The result is a 47 percent increase in productivity.

    Finally, staff members are more efficient, and they have additional opportunity to innovate and to help move the lab in new directions. They are able to spend time on building test volume and validating new tests for potential future use rather than on manual activities such as sample handling and processing. With this new approach, morale is up. This has helped UIHC to attract and retain the best and the brightest talent at a time when staffing is a challenge elsewhere.

    So, what lessons did we learn? Automation is best done when you make incremental improvements and create a culture of continuous improvement. This means that the chosen employee team needs to look continually at the process and evaluate ways to use employees and technology more efficiently. Other lessons: involve the people who will do the work, keep in contact with the team and stay on task, focus on improving the workflow and work processes, communicate the vision to staff repeatedly, develop a thoughtful and uncomplicated system, and have adequate technical support.

    Automation creates a natural fear of job loss. Management had to convince administration that automation does not equal job reduction. In fact, middleware created new work assignments. Today, five medical laboratory scientists are trained to write middleware rules. As the lab gained efficiencies and needed fewer staff, positions were eliminated by attrition or moved to other laboratory services.

    To maintain success, ongoing measurement and reporting are critical. The lab runs real-time reports every two hours. The results are recorded on a whiteboard so everyone working in the lab can see how quality measures are tracking. If it appears there are problems getting the samples onto the automation highway in a timely fashion, a laboratory staff member investigates and looks for patterns. The two-hour reporting continues day and night, and the reports are shared with staff working different shifts. These numbers help keep people on track by inciting a bit of healthy competition and allowing for better communication between those who are working within the same job function but do not cross paths because of scheduling.

    Sue Zaleski is laboratory manager, Department of Pathology, University of Iowa Hospitals and Clinics, Iowa City. Jeff Kulhavy is assistant laboratory manager.

    Related Links Related Links