read time: 5 minutes
Technology & Innovation
In the past 20 years, automation has added significant value to the diagnostic laboratory. In the near future, advances in automation, and especially the arrival of artificial intelligence (AI), will bring even more important changes to the lab. That is the message from James Davis, chairman, chief executive officer and president of Quest Diagnostics, in a recent edition of the American College of Healthcare Executives podcast, “The Healthcare Executive.”
“When I first came to Quest Diagnostics, having come from other manufacturing industries, I was struck by the level of automation in those other industries versus the level of automation that I saw in the laboratory industry,” says Davis. “The lab industry was still very much at the early stages, and so these are the things that we started to focus on.”
Davis was trained as an aeronautical engineer, and began his career designing turbine blades for jet engines for GE. “I still get passionate about seeing airplanes in the sky,” he says. He moved into GE’s healthcare division in 2001, “and I fell in love with healthcare.” His main focus was radiologic diagnostics, specifically diagnostic magnetic resonance imaging. He joined Quest in 2013, becoming senior vice president of operations in 2014; he became CEO and president in 2022. “It's been a great 10-year run so far. I am just as passionate about lab-based diagnostics as I am about radiological-based diagnostics.”
Automation has been an enormous benefit to the diagnostic lab, Davis says, especially for routine tasks such as dividing a sample, for instance a tube of blood, into separate aliquots for all the different tests that must be run on it. “It's hard to attract and retain people to sit and pipette specimens all day long, it's just not a very fun or pleasant task.” Automating such tasks also means fewer mistakes, and greater safety for lab workers. For COVID-19 testing, for example, nasal swabs are automatically decapped and the fluid is automatically pipetted into the analyzer, avoiding human exposure to potentially infectious virus. “So, these were some of the quick wins,” he says.
The shortage of workers to staff a round-the-clock high-tech diagnostic lab is significant, and is a driver of automation. Samples are collected during the day, and must be rapidly processed, which means the lab doesn’t sleep. “We start in earnest at 9 or 10 o'clock at night and work through 6 or 7 o'clock in the morning to process most of the specimens. And it's tough finding people that want to work those hours. And so, there's definitely a role here, when you can't get enough people to do this work.”
Additionally, “There is an improvement in quality; the fewer times you actually physically touch a specimen, the better. If you're not putting human hands on, introducing heat or other potential contaminants into the specimens, it's much better from a quality standpoint.” And, he adds, freeing up staff to do higher-value tasks is a benefit for both the worker and the lab.
Quest has begun to move far beyond relatively basic automated operations, he emphasizes. “We have a couple of very advanced facilities where we've completely automated the laboratory. The minute the specimen comes in, [the label] is scanned, so that we know what tests we need to run on that specimen. And then we put it in a puck,” or tube holder, “and the puck carries it to the various analyzers in the lab that need to do testing on that specimen. Not all of our labs are at that level of automation, but we're quickly moving into that direction.”
Artificial intelligence (AI) is the most important new development in diagnostic lab automation, and its effects are only just beginning to be felt. “When it comes to the use of artificial intelligence, we're always going to start with the precept that it has to improve quality,” Davis says. AI is finding its highest initial value in the pathology lab, where digital scans of tissue samples can be screened for anomalies that are passed along to human pathologists for diagnosis. And AI systems don’t tire after looking at hundreds of images, a real benefit in controlling quality in a busy lab.
AI may also help with the aging of the trained labor force. “This country still does probably between 40 and 50 million pap smears a year, and if I look at the average age of the people in our laboratories that read these pap smears, pathologists and cytotechnologists, it is probably well north of 55 years of age,” and fewer young people are training to go into these fields.
The goal of AI in the diagnostic lab is to aid human intelligence, not replace it, “and to improve the quality of the work that the laboratory business does,” Davis says. “The goal is to reduce the number of false positives and false negatives, and to help the lab scientist make decisions much more quickly.”
The entire podcast is available here: https://soundcloud.com/healthcare-executive