As every minute passes, AI continues to alter the fabric of modern life as we know it. Its endless applications and vast potential to improve everything from mundane tasks to life-saving innovations mean it’s safe to say artificial intelligence is here to stay. Experts even claim AI will continue to change everyday life to the same degree as inventions like electricity and the internet.
Despite skepticism and fearmongering about AI glitches circulating in the news, many existing applications seamlessly rely on AI. Many of these products have been painstakingly developed and refined to virtual perfection.
Looking beyond the tech world, AI has played a significant role in improving modern medicine for over 20 years. Since 1995, the FDA has received over 300 submissions for drug and biological products with AI components and more than 700 submissions for AI-enabled devices. One of the most successful applications of artificial intelligence in human medicine is the interpretation of radiographs and other studies. This technology is available to veterinarians as well.
A Second Set of Eyes
The benefit of a second set of eyes cannot be underestimated. In the veterinary industry, it is standard procedure for practitioners to consult a remote radiologist when they are faced with a complex differential diagnosis. In an ideal world, a specialist would read every film to ensure the best possible outcome for every patient. However, outside consults are expensive, typically costing hundreds of dollars in fees that many clients are unable or unwilling to pay. These consults also take time, a commodity most busy practitioners do not have to spare.
In contrast, AI radiograph consults take a matter of minutes and, at a mere $10 to $20 per patient, are a fraction of the cost of a traditional radiology consult. Perhaps more compelling, AI radiology programs have been proven to outperform the diagnostic accuracy of even top radiologists. So, why isn’t every practitioner clamoring to get their hands on this technology?
Change can be challenging.
Despite blind studies affirming AI’s superiority in radiograph interpretation, some practitioners still refuse to believe in its potential. While, in truth, AI is only as good as the information it is based on, radiology is a field particularly well suited to AI. There are a near-limitless number of films and readings to input, enabling the technology to constantly learn and improve its accuracy.
The Malpractice Question
As it turns out, the primary reason these practitioners have not yet adopted AI radiology has less to do with their faith in its efficacy and more to do with misconceptions about protection from malpractice liability. These practitioners prefer to consult with “real” radiologists with the misunderstanding they will be better protected from malpractice liability in the event of a misdiagnosis.
However, in most jurisdictions, the outside consultant does not provide a protective cloak from liability. As all veterinarians know, the veterinarian-client-patient relationship obligates the practitioner to uphold the standard of care. Although malpractice cases can be litigated, unhappy clients are more likely to turn to their state’s board of veterinary medicine with a complaint, which can result in disciplinary actions, fines or worse (license revocation), a nightmare for any veterinarian. As a result, if a veterinarian consults an out-of-state professional who makes an incorrect diagnosis, the primary care practitioner licensed in the state where the patient was treated and who adopted the out-of-state expert’s reading is the one left standing before the board.
How is this the case? Most state statutes governing the practice of veterinary medicine provide an exception to licensure requirements for out-of-state consulting veterinarians. This is in recognition that a doctor who treats multiple species of living beings cannot be an expert in all of them. These states have determined it is in the best interest of their citizens to allow out-of-state veterinarians to assist with difficult cases, so long as the consultants have a valid license in their home state. However, regardless of whether the consultant is “in” or “out” of state, the duty of care remains with the treating veterinarian unless the consulting veterinarian examines the patient and creates a separate VCPR with the animal’s owner. Because the consulting veterinarian is frequently not licensed in the practitioner’s state, only the practitioner is subject to discipline by the board of veterinary medicine.
A Need to Indemnify
So, what makes artificial intelligence different? Certainly, it is no good to point a finger at a machine. The answer lies in reputable companies’ willingness to stand behind their programs’ accuracy. These companies have perfected their systems in many trials and over many years using massive amounts of data. Studies have repeatedly affirmed the superiority of AI readings to those made by human radiologists — even the best in the field.
With little fear regarding their programs’ accuracy, reputable companies should offer a warranty regarding their claim of superior accuracy. In the rare event a reading is incorrect, the company should indemnify and defend the veterinarian facing liability resulting from an incorrect AI reading. This contractual obligation is worth far more to the practitioner than what little support an individual consultant can provide.
Aside from cost savings, accuracy and protection from liability, another often overlooked benefit of using AI is its unmatched speed and efficiency. All busy practitioners could use extra time in their day. Even those who use outside radiologists can face a substantial turnaround time. In contrast,
AI results are available in a matter of minutes.
Even more impressive, if a practitioner questions the accuracy of an AI reading, a reputable company should offer a second reading. This time, the case is under closer scrutiny and is often vetted by a team of radiologists available for just that purpose. Even with a surcharge for the second reading, the practitioner gets the benefit of both AI and a specialist — at a fraction of the cost of a traditional consultation.
The Standard of Care
The final consideration is the standard of care. The legal analysis of the standard of care in veterinary malpractice is largely derived from the law of human medical malpractice. Before the internet, most states recognized that the standard of care for a physician could differ depending on the doctor’s practice location, available facilities, and sophistication. For example, the standard of care for an appendectomy in a large, urban, academic hospital was quite different from the standard that might be required of a general practitioner performing emergency surgery at a small country hospital. However, in recent years, recognition of this doctrine has waned, as the internet provides quick and accurate information to virtually everyone.
The logic that sophisticated information could not be accessed and utilized by the country doctor no longer applies. So, in some states, like Virginia, the “locality rule” has been abolished by statute, making every medical doctor, regardless of circumstances, subject to the same standard of care.
The Rule, Not the Exception
As artificial intelligence continues to gain footing in virtually every area of our lives, its use in veterinary practice will also expand.
When veterinarians’ options were limited to relying on their own readings or asking clients to pay a substantial upcharge for an expert consultation (which they would often decline), the standard of care did not require the consult. However, since AI is easy to use, faster, more efficient and costs a fraction of an expert consult, history suggests it will eventually become the rule and not the exception. Clients will come to expect the service from every practitioner. Or, at the very least, clinics offering AI services will be preferred by clients over those that don’t.
Eventually, veterinarians who fail to adopt AI will find that their practices no longer meet the basic requirements for the accepted standard of care. Considering the speed at which AI is being adopted worldwide, a new standard of care will be here before we know it.
LEARN MORE
The marketing company Full Slice compiled A Guide to AI Tools for Veterinary Medicine at bit.ly/4ely04e.