CAP TODAY Pathology/Laboratory Medicine/Laboratory Management
Raymond D. Aller, MD, and Hal Weiner
How informatics tools can boost QA in anatomic pathology
Love and marriage, cookies and milk . . . quality assurance and informatics? In pathology, the latter pair are a natural fit, says Liron Pantanowitz, MD.
“If you look at the life cycle of a specimen in the laboratory, with all the data that gets generated and all the people who are touchpoints along that process, there are many opportunities to use informatics to improve the quality check measures and to do it on an ongoing basis,” says Dr. Pantanowitz, professor of pathology and biomedical informatics and director of pathology informatics at the University of Pittsburgh Medical Center. “If you have a quality issue,” he emphasizes, “you should look for an IT solution to improve it.”
Dr. Pantanowitz and UPMC colleague Douglas J. Hartman, MD, associate professor of anatomic pathology and associate director of pathology informatics, shared their knowledge of how to use informatics tools to solve quality related issues in pathology in a chapter of the new book from CAP Press titled Quality Management in Anatomic Pathology: Strategies for Assessment, Improvement, and Assurance (see story, page 5). The chapter targets not only pathologists but also laboratory administrators and supervisors who oversee informatics or quality assurance projects “but may not have put them together in the most effective ways,” Dr. Pantanowitz says.
The colleagues contributed the chapter, aptly titled “The intersection between quality assurance and informatics,” because at UPMC “we live and breathe this stuff and have published a lot of the work we’ve done using informatics tools to improve quality,” explains Dr. Pantanowitz. “Our intention was to let pathologists know that this synergy exists, to give them specific examples, and to show them how we do it at our institution.”
Case in point: pre-sign–out peer review. Having another pathologist review a case is a key component of quality assurance, but due to logistical hurdles, “it’s often done after the case has been signed out, when there’s less time pressure,” says Dr. Pantanowitz. The downside, adds Dr. Hartman, is that “after 20 to 30 days, there can be all kinds of problems in locating the slides. And often, when you do catch an error, the patient’s care plan has been initiated. We realized it was better to do it in real time, before the first pathologist signs out the case.”
Consequently, the IT team programmed into the LIS instructions to randomly flag cases for review before sign-out, without letting the pathologist know in advance that a case would be reviewed. “When the computer picks the case,” says Dr. Pantanowitz, “you hand the case over to another pathologist on service; they do a quick check and say whether they agree or disagree. Thereafter, you can go ahead and correctly sign it out—making sure no mistakes are made.” The percentage of cases selected for review can be customized for each pathologist; newer staff members might have as many as 20 to 30 percent of their cases reviewed, he adds.
Implementing quality assurance within a pathology department starts with collecting and analyzing data, and if the data are “hard to come by, hard to use, hard to manipulate or understand, that can impair the ability to perform good QA,” asserts Dr. Pantanowitz. Informatics tools such as LIS modules or middleware can simplify the collection and analysis of data and allow labs to apply the findings more broadly, he continues. “You can now start to tap data from many different systems and put them together. You can put the CAP benchmarks in place and have real-time assessments for quality improvement. . . . We’re a large pathology department spread over many different integrated hospitals. If we did not have [such] informatics tools, we would not be able to roll out our QA plan and program to the whole system.”
“I think it’s a challenge,” adds Dr. Hartman. “You really cannot effectively implement informatics without understanding the quality component. That’s where workflow becomes really important. These IT solutions, in order to improve safety and quality, have to be optimized for the workflow of the people in the lab.”
In addition to describing the UPMC peer-review solution, the book chapter, which is peppered with figures and tables, addresses numerous other suggestions for improving LIS functionality, such as employing biometrics, building decision support tools, and initiating tracking of operations, workload, and inventory. The authors also discuss regulatory compliance, coding, specimen labeling, and other factors that influence quality assurance efforts.
“Not all LISs do a good job of allowing us to collect, analyze, and manipulate data for QA purposes,” says Dr. Pantanowitz, adding that for some quality assurance tasks, data must be exported to an Excel spreadsheet or other tools. “Today, the issue is working with small amounts of data; in the future, it will be big data,” he notes. “Analyzing data from the LIS along with raw data from other systems allows you to do so much more, such as generate inferences and check on outcomes—you can determine whether the things you’re doing in the lab are having an overall positive or negative outcome. The LIS needs to be able to accommodate these needs.”—Jan Bowers
Carequality and CommonWell announce collaboration
Two names that have become synonymous with interoperability in health care are officially linked.
CommonWell Health Alliance, a nonprofit trade association of health information technology companies and other organizations operating a vendor-neutral network for data exchange nationwide, and Carequality, developer of a common interoperability framework that supports data exchange between and within various types of health data sharing networks, have joined forces to further their cause.
Under a three-pronged agreement announced last month: