TRIMEDX Chief Information Officer Brad Jobe recently contributed thoughts in the industry on the impact of artificial intelligence on healthcare technology management. A full-length article appears below, and links to published stories include:
- Healthcare Business Today: published Dec. 2, 2023
- AAMI News: published Dec. 5, 2023
- Quality Digest: published Dec. 13, 2023
Artificial intelligence (AI) has the potential to reshape the healthcare industry. There is a massive amount of healthcare data available for artificial intelligence to process. Nearly a third of the world’s data volume is generated by the healthcare industry, and the volume of big data is projected to increase faster in health care than in any other field. As health systems continue to deal with razor-thin margins and consumers demand high-quality, cost-effective patient care, leveraging large datasets with AI can drive cost efficiencies across the industry, including within the healthcare technology management (HTM) field.
While AI does not replace human oversight and management, it allows BMETs and other HTM professionals to work more efficiently and focus on their core skillset and responsibilities. The HTM field should embrace developing AI opportunities, while being mindful of the potential risks and concerns that come with the implementation of these emerging technologies.
Increasing Day-To-Day Technician Efficiency
Artificial intelligence has the capability to significantly increase BMETs’ efficiency when completing day-to-day tasks. Authors of a 2023 Nature Biotechnology report examined two decades of U.S. and European Patent Office Data to find AI and medical machine learning (MML) patent trends. The report found 26% of the patents with MML claims were aimed at technology improvements, such as platform tools and medical devices.
The ability to predict device failures before they happen can decrease equipment downtime, save money on expensive device repairs, improve the patient experience, and make BMETs’ jobs less stressful. Predictive work systems, powered by AI, will be able to learn from device logs to spot warning signs that a device will soon have a larger problem. This gives BMETs the ability to order necessary parts and make quick repairs before the device reaches a point of failure. Health systems will have fewer unexpected equipment failures, which lead to significant downtime and rescheduled patient appointments. BMETs will also have fewer device emergencies that force them to postpone other responsibilities.
Machine learning can also assist BMETs when they need to repair a piece of equipment. By learning from the large amount of data available—including manuals and device history—the technology will be able to generate a work order for the technicians dictating the problem, steps to solve it, and what parts will need to be ordered. Once the repair is completed, AI will be able to write up the work order summary for the technician. While some healthcare providers have been wary of implementing AI directly in patient care, they have made use of it as a scribe or back office organizer. BMETs can also use the technologies to reduce paperwork and other repetitive tasks.
Using artificial intelligence as a component to test medical devices is another opportunity to increase efficiency and accuracy. Running tests through mobile apps which then deliver results to the system will reduce technician workload. Technicians don’t have to write down or document the results by hand, reducing human error. AI will be able to learn from that bank of data to validate that test results are within permissible limits.
Predictive work systems and automated testing equipment are already saving health systems time and money, and the benefits are sure to grow as the technologies advance. However, human oversight is still needed. A technician should still use their expertise to determine if a computer-generated solution is indeed appropriate, rather than trusting the proposed solution without critical thinking. It’s also important for health systems to engage with technicians when deploying AI tools—to gather feedback and ensure new procedures & technologies are enhancing their work experience rather than adding to their list of responsibilities.
Enhancing Productivity Across The Health System
Artificial intelligence has the potential to enhance productivity, save time, and decrease administrative tasks throughout the entirety of a health system. Machine learning can play a valuable role in supply chain management, capital planning decisions, legal contracts & review, and cybersecurity.
Health systems could use AI to generate recommendations for the best equipment to fit their needs. AI could use industry and hospital data to evaluate equipment by capability, durability, total cost of ownership, and expected utilization to help determine what type of equipment to select and whether they should purchase or rent it. It could also use vendor data and invoices to help health systems confidently select vendors or identify vendor issues, and then assist in ordering appropriate parts from the preferred vendor. At a time when supply chains and inflation are especially challenging for health systems, machine learning would prove to be a useful resource.
Large language models can examine contracts or other legal agreements to spot any outliers or risks. This will allow health systems to make sure they’re informed and give them the opportunity to alter language or terms that aren’t in their best interest.
In addition, AI can be a key tool in a health system’s cybersecurity toolkit. With AI-powered solutions, health systems are able to quickly detect and analyze threats. As part of a comprehensive cybersecurity program, AI can evaluate data and patterns to help predict and respond to cyber-risks.
Risks And Concerns Must Be Top-Of-Mind
The stakes of using artificial intelligence, and any other developing technology, are higher in health care than other industries because human lives are in the balance. While the opportunities for AI to enhance patient care and health system operations are vast, there are also concerns that need to be top-of-mind when using these technologies. While concerns are primarily focused on applying AI to make clinical decisions, every innovation in the industry must be thoroughly vetted to ensure patients receive reliable, quality care.
President Biden recently issued an executive order with several elements focused on AI in health care, including calling on the Department of Health and Human Services to create a safety program to receive reports of—and act to remedy—harms or unsafe healthcare practices involving AI.
AI technologies and machine learning are only as good as the data they’re built upon. A major challenge to functioning machine learning models is building a foundation of accurate, complete, and representative datasets. When dealing with the size and scope of data inside a health system, it is guaranteed to have gaps and inaccuracies. Applying machine learning engines on top of these imperfect data sets can create a “garbage in, garbage out scenario.” AI technology may not be able to weed out the noise within the data sets, resulting in inaccurate analysis and outputs. Theoretically, the impacts of this inaccuracy could be as far-reaching as misdiagnoses, medication dosage miscalculations, and improper patient instructions.
AI is not built to understand human moral values, which could lead to morally flawed healthcare practices. Machine learning is built on the data and processes currently in place, which have been built by humans with a variety of beliefs and biases. This can translate into biased AI databases, resulting in automated gender, ethnicity, or socioeconomic discrimination. President Biden’s executive order includes several directives aimed at preventing algorithmic discrimination.
These concerns illustrate why AI should not completely replace human oversight. An experienced professional should still monitor these technologies, investigating further if analysis or output appears suspect.
Finally, health-related information is some of the most sensitive and legally protected data categories. Ensuring the ongoing security and privacy of protected health information (PHI) is a major challenge in the development of large-scale AI models. PHI laws are numerous and complex and often vary on federal, state, and local levels. Health systems must be vigilant to avoid leveraging patient data inappropriately or exposing high-risk data in an unsecure way.
These are just a few of the myriad ways health systems can leverage AI to increase productivity, reduce costs, prevent human errors, and improve patient & staff experience. These are exciting and constantly evolving technologies, with regulations and industry standards in very early development stages. To ensure responsible and secure use, health systems should work with a trusted partner who is well-versed in AI and capable of closely monitoring and abiding by new regulatory & legislative policies.