Kristi Fender is a senior content specialist with Stephens & Associates, a Kansas City agency that works with animal health companies. Before joining S&A, she spent nearly 20 years in veterinary journalism with several animal health publications.Read Articles Written by Kristi Fender
Not too long ago, when we envisioned how artificial intelligence (AI) and the Internet of Things (IoT) would affect veterinary practices, our mental picture may have ranged from something out of The Jetsons to a dystopian sci-fi flick. Well, that day is here, and what it actually looks like is animal hospitals serving their clients, patients and teams exceptionally well.
Two veterinary professionals at the forefront of technology adoption are finding that solutions they’ve recently implemented solve a number of practice challenges — some they predicted and some they didn’t. Take a closer look at these issues and how they’ve been alleviated by technology.
1. Inefficient Communication
Veterinary Dental Services in Boxborough, Massachusetts, recently implemented an IoT-based workflow system that connects all of the practice’s smart devices (lab units, monitoring machines, etc.) to a patient-status whiteboard and its management software. The result, says practice manager Genel Les, MBA, CVPM, CCFP, is a huge gain in communication efficiency.
“It helps us see the day as an overview and what patients need to have what tasks done to them at what time,” Les says. “That cuts back on the communication where we say, ‘OK, what’s left to do to Fluffy? Did somebody already run blood work on Fluffy? What time are we premedicating Fluffy? Has Fluffy’s owner been called?’”
Now all of that information can be found in one place: a flat-screen TV mounted in the treatment room — the whiteboard — so the team can see what tasks have been done for each patient and what’s still ahead. The information is also accessible from any computer or tablet device in the hospital.
“Our head technician, who’s in charge of everything on the floor — she used to get bombarded,” Les says. “Sometimes the same question was asked of her three different times by three different people. It was no fault of theirs; they just wanted the information. It’s really helped to lighten the load on her.”
Not only does this enhance workflow within the practice, but it also helps improve communication with clients, Les says. For example, say Fluffy’s owner calls and asks for their pet’s status. Instead of the client care specialist putting the client on hold while she tracks down the information, she can glance at the whiteboard screen and give an immediate answer. “It helps with communications not only internally but also externally,” Les says.
2. Missed Charges
Another problem the workflow system solved was capturing all charges, Les says. “Say we give a patient an NSAID injection,” she explains. “On the anesthesia monitoring sheet within the system we’ll put our initials in next to the line saying we gave the injection. It automatically flows over into our practice management software, and not only does it put it into the medical record, it bills the client for it
One step — digitally initialing a line in the workflow software — has replaced writing down the procedure on a physical anesthesia monitoring sheet and remembering later to enter it in the practice software. “That saves us a lot of steps and a lot of time, and capturing those missed fees has really helped our bottom line,” Les says.
3. High Radiology Costs
Financial considerations piqued the interest of Karen Bradley, DVM, co-owner of Onion River Animal Hospital in Middlesex, Vermont, when she heard about AI-based veterinary radiology. Ever since her hospital switched to digital radiography 13 years earlier, they had built a radiologist review into their fee. While this step offered the doctors reassurance and confidence regarding their own interpretations, it was expensive.
Along came the opportunity for the hospital to try out a machine learning–based system designed to give real-time results on radiographic studies based on millions of X-ray data points. After a trial run, they were convinced to take the plunge. Notably, the cost per image was more than five times lower than it had been with a human radiologist review.
“Our goal was to reduce the number of cases and the cost associated with sending something to a radiologist,” Dr. Bradley says. “Here we are over a year later, and we have accomplished that.”
4. Human Error
Dr. Bradley had a personal interest in the technology as well. “I’m a breast cancer survivor. And on the human medicine side, we’re seeing that artificial intelligence is getting more accurate,” she says, referring to a study in which an AI program produced 5.7% fewer false positives and 9.4% fewer false negatives than a human radiologist.1 “That matters.”
From a clinical perspective, the biggest benefit Dr. Bradley sees in her own practice is that the team gets the assurance they aren’t missing anything so they can more confidently narrow their rule-out list. One case in particular stands out. A dog had come in for vomiting and stomach upset, so the team took radiographs as part of the workup. “It wasn’t the most perfect X-ray technique, because we had part of the lung field in the image,” she says.
“With the dog’s history of getting into the garbage, we were razor-focused on the abdomen,” Dr. Bradley continues. But the AI system offered an unexpected finding: a pulmonary mass — with the highest possible confidence interval. When the team looked at that part of the radiograph more closely, sure enough — there it was.
“That lung mass went completely undetected by multiple veterinarians who had looked at these X-rays,” Dr. Bradley says. “It forced us to look at that area and say, ‘Wow, now we need to take X-rays of the chest because there’s something going on there that we would have missed.’”
Of course, AI will never replace human radiologists, Dr. Bradley says. A best-practice approach relies on both humans and AI working together synergistically.2 She still utilizes her veterinary radiology consultants to look at a patient’s individual history and bring their years of experience to bear on a case. This is especially helpful in decisions about whether to go to surgery.
“They’re looking at more subtle things than the AI might pick up, and they’re recalling their experience,” she says. “If they suspect foreign material in the intestine and they’ve been right in their 20 previous similar cases that surgery was called for, then they can say, ‘Yes, this one needs surgery.’”
5. Need for Mentorship and Growth
Dr. Bradley has found over time that many veterinarians early in their career are looking for ways to grow in confidence. This is especially true in radiology. “They want the reassurance that they’re not missing anything on their X-ray read,” she says.
At the same time, while practice owners know that mentoring their young associates is important, finding the most effective ways to do that can be a struggle. Enter AI. “To me, this is one of the tools that can help be the little radiologist on their shoulder at the time that they need it,” Dr. Bradley says.
It’s also great for veterinary nurses/technicians who are interested in radiology. “They’ll look at the platform a lot and get information from it, or they like to learn themselves,” Dr. Bradley says. “It’s great for them to be able to say, ‘I thought that heart looked big, and look — it agreed with me.’ They can also be the ones who alert you to say, ‘You saw that the spleen was really large, right?’ And I say, ‘Oh, I missed that.’ Getting that communication where we are all sharing the same set of information is great.”
While the practice of the future may not (yet) involve humanoid robots or wearable jet packs, the advancements that are transforming team efficiency, client communication and patient care are no less inspiring. And with innovative companies working alongside leadership teams to transform veterinary care even more dramatically, who knows where technology will take us next?
Cognitive computing search tool delivers credible medical information in seconds
When Tyrrel de Langley, DVM, MRCVS, CIM, PMgr, CMgr, became area medical director for VCA Canada in London, Ontario, he knew things had changed since he’d last been in practice. He’d spent the most recent phase of his career in university lab animal medicine, and now he needed to get up to speed so he could support the veterinarians at the clinics in his region.
“I wanted to know what the current information was and how I could easily get it,” Dr. de Langley says. “I also wanted to know I could trust it. Many veterinary Facebook groups and message boards are full of anecdotal information and commentary. It’s not something I’m necessarily willing to bank on.”
Dr. de Langley heard about a veterinary medical search tool based on IBM Watson cognitive-computing technology that sounded promising, and it turned out to be the solution he was looking for. He could ask it a clinical question and receive relevant answers from highly regarded textbooks and peer-reviewed medical journals in seconds.
For example, recently he fielded questions from VCA veterinarians about Addison’s disease: What are the current recommended therapies? Do you use mineral corticoids or just a cortisone? How do you handle an Addisonian crisis? “I have a recollection of what I used to do, but what I used to do is not necessarily current practice,” Dr. de Langley says.
In these situations he turns to his search tool and quickly retrieves protocols from textbooks like Blackwell’s Five Minute Veterinary Consult and current journal article findings. He then passes the information along to his doctors who can immediately use it in practice. “They’re busy seeing appointment, appointment, appointment, and I can provide them the information they need in half an hour,” Dr. de Langley says. “It makes me look smart.”
AI has made Dr. de Langley’s job far easier and allowed him to be more efficient, he says. He and his colleagues also use it to educate clients, passing along printouts for pet owners to use as they encounter other information sources online.
“We just hand it to them and say, ‘Here’s our medical veterinary reference. We want you to know as much as we do. So when you’re Googling, keep this by your computer. If it’s hard to understand, give us a call and we’ll explain,’” Dr. de Langley says.
Besides increasing clients’ knowledge, the doctors have found that this strategy increases compliance. “When a client can see an acknowledged third party from a textbook saying, ‘Here’s how you approach a case; here’s how frequently you should do follow-up blood testing or follow-up radiographs,’ that’s powerful,” Dr. de Langley says.
And finally, having such ready access to reliable information facilitates better patient outcomes, Dr. de Langley says. “It makes our ability to reach that target easier. It’s another tool in our toolkit to compel the client to make the best decision for the care of their pet.”
- Killock D. AI outperforms radiologists in mammographic screening. Nat Rev Clin Oncol. 2020;17(3):134. doi:10.1038/s41571-020-0329-7
- Schaffter T, Buist DSM, Lee CI, et al. Evaluation of combined artificial intelligence and radiologist assessment to interpret screening mammograms. JAMA Netw Open. 2020;3(3):e200265. doi:10.1001/jamanetworkopen.2020.0265