Skip to main content

Rise of the [Diagnostic] AI Machines!

Radiologist reviewing diagnostic imaging

Are we ready to welcome our AI overlords into healthcare yet? Will we welcome a fully autonomous Dr. Crusher? We’re not likely ready to completely abandon our human healers, but there have been a variety of algorithms and machine learning applications that have been creeping their way into our clinical specialties for many years now. The first AI algorithm was approved by the FDA in 1995, in the beginning of 2023 there are 520. The specialty with the largest number? Radiology! Check out this article from HealthExec, it provides some more detail on the approval and advantages of AI in imaging.

Radiology, and similar specialties that utilize imaging, have long embraced computer systems to improve quality, productivity, and sharing. The complexity behind the scenes of these systems can be quite surprising. Once an image is captured it may still go through multiple systems just to be viewed, PAC Systems, VNA, Dicom viewers, the route is anything but direct. If we want to add another step in the process to have the image analyzed, we are faced with yet another connection.

In a previous blog post we pointed out the challenges that staff face when one of these systems aren’t communicating, these certainly impact our physicians as well. As we add more diagnostic solutions to aid our physicians in treating patients, we want to ensure that their time is actually spent doing what they are trained for. The average radiologist makes roughly $300,000 per year, if we assume that radiologist works 40 hours per week, that’s about $144 per hour.

What happens when the machines stop talking? Likely the radiologist calls the help desk… 5 minutes on hold ($12), they will take 3 minutes to explain the problem to the help desk person ($7.20), who will then transfer them to a specialist where they take another 5 minutes talking over the problem ($12). Helpdesk will then forward the issue to multiple analysts and integration team. At-least 2 different analysts getting pulled into diagnosing the issue. Typically, integration and system analysts make 90k-130k per year or on average $50 per hour. These 2 analysts will work on this issue for about 30min ($100) to diagnose the issue. $131.20 may not seem like a significant amount, but now that radiologist may not be reviewing images, may not be discussing treatments with patients, they may not be doing what they’ve been trained to do. If the radiologist is only 50% as productive without this solution, every hour of outage costs $132…  That’s just for one Radiologist. And typically system issues affect all users. So once you multiply the outage costs for all radiologists and other team members it really adds up fast. There can also be significant downstream impacts to other services waiting on imaging as well that become harder to calculate.

Moving beyond the frustration of the healthcare teams, there can be a detrimental impact to patient satisfaction as well. As results are delayed and the team gets backed up, every hour of delay adds to the patient’s anxiety and fear. Anxiety and fear are not feelings we want our patients to experience, we want them to have confidence in our teams and systems.

Automated applications and integration monitoring can help quickly resolve these problems. With early notifications to the right team in a timely manner, problems can be quickly resolved to minimize downtime impacts. For over 10 years Tido Inc. has been partnering with health systems to help maximize their IT systems and quickly resolve issues as they arise, often before the end user even notices. Contact us today and so we can talk about how we can help you keep your systems working for your clinical teams.

Hospitals are losing billions in 2022, how can IT improve ROI with existing infrastructure?

healthcare financials

Hospital margins have been challenged by increases in labor expenses and shortages coupled with declining admissions and procedures. As financial pressures increase so too does the pressure on all departments to provide more value, more savings, and a greater ROI. Costs are rising faster than hospitals can raise revenues and prices.

There are a lot of solutions out there that offer promise of greater efficiencies for clinicians, new cloud based software that can provide greater insights into care practices and increased billing. Beyond the internal benefits, there are additional advances in hospital at home programs, and remote monitoring for chronic conditions. There is a learning curve to these systems, and for most it can often take months or years to realize the promised return.

How do you provide greater cost savings and efficiencies in care through existing IT networks and infrastructure? What can IT departments do now that doesn’t require education and training campaigns, massive investments in infrastructure or new systems?

In healthcare we know that early identification of problems is key to the most effective treatment. This is no different when it comes to IT systems ensuring safe and effective patient care.

How about monitoring the reliability of their current networks? All of these interconnected solutions require connected networks to function optimally and provide the necessary advances in patient safety and clinician efficiency. Buying the latest cloud-based AI solution to improve diagnosis, treatments, safety, and insurance denials, is only useful when the network is functioning.

Back in 2018 network reliability was identified as one of the risks to patient safety, what have systems done since then to ensure reliability? The pandemic likely radically changed or accelerated certain IT investments, moving up some upgrades or canceling others.

Calculating a basic ROI for pro-active network monitoring can be accomplished, just looking back at previous blog posts we can often see the time used by staff on this activity. If end-users are monitoring the network and checking for transmission of data and reports, their time is easily measured and value assigned. Are the IT departments measuring this time? Unlikely, but the unit managers can tell you how many hours each person spends on this.

  • When we considered the Cath Lab in a previous post, an RT or RN spends 3-5 hours every week checking feeds. Multiply this by every imaging area in the hospital, the numbers add up very quickly, at the most basic there may be 4 different imaging areas connected.
  • Pharmacy systems were also previously reviewed, how much time is spent reporting issues? It can be up to 20 minutes of a nurses time on the phone troubleshooting, likely with at least one additional call back.
  • How about remote monitoring programs? How much time will a nurse or physician spend on the phone with a patient trying to troubleshoot transmissions only to find that a connection was interrupted within the hospital’s own system?

Taking a look at the overall IT infrastructure the numbers can quickly add up throughout a hospital or health system. How often are feeds interrupted? Likely not that often, but even a single interruption after a monthly update can have significant ramifications to productivity when spread across an organization.

Those are the simple calculations to measure ROI, more difficult to measure are the ramifications to patient safety, staff satisfaction, and perhaps even future IT investment. If “nothing ever works around here” then there will be difficulty getting clinician buy-in to adopt the newest and best technologies.

Pro-active monitoring can enable early detection and warning. A simple message from IT can alert staff that IT is already aware of a problem and working on a resolution. Is IT optimizing the existing infrastructure to provide the best ROI?

Tido provides automated end to end monitoring solutions that will automatically alert your teams there is a disruption.