Artificial Intelligence is playing an increasingly important role in digital health. Optimizing surgery scheduling is an area where AI can have a big impact. By slotting AI and machine learning algorithms into EHR and scheduling systems, practices will be able to improve patient safety and increase revenue while taking into account the constraints of all surgical elements. The potential of AI in surgical scheduling is just being realized and can yield incredible results.
Optimized block time
Getting the most out of block time is crucial for both practices and hospitals. A recent study shows that the average value of an OR room is $62 per minute. With each minute so valuable, a practice needs to ensure optimal usage.
Currently, scheduling systems do not incorporate any machine learning. In fact, many of the systems used by hospitals and clinics were designed for in-office scheduling – and not surgical scheduling.
While schedulers may be able to fill the surgeon’s block time for the day, they may not have taken into account other factors – a surgery taking longer than estimated, the complexity of the case, and so on. This can lead to surgeons working longer than their block time, thereby increasing risk to patients.
Mayo Clinic case study
In 2012, the Mayo Institute reviewed their own spinal surgical scheduling in an effort to find a better way to structure their scheduling process. They found a great deal of inefficiency, and doctors working a significant overtime – both potentially harmful for patients and costing the clinic substantial amounts of money.
Using data mined from previous surgeries performed at the hospital, the Mayo began mapping out the time actually spent on each surgery. They took into account the patient’s profile, the doctor’s previous cases, and numerous other factors. The information was fed into an algorithm with several complex models that attempted to maximize surgeons block time while preventing overtime in the OR. A scheduler would then be shown, through a color-coded system, the optimum openings for a patient based on surgeries already scheduled that day.
After rigorous testing and tweaking, the Mayo Clinic successfully deployed this scheduling algorithm. During the pilot program – which ran from June 2012 to July 2013) – the Clinic increased utilization for spine surgeries by 19% and reduced overtime by 10%. These numbers are significant for clinics as large as Mayo, but the results can be just as significant for smaller size practices.
Slash down time
One would assume that “maximizing block time” simply means cramming a surgeon’s schedule with back-to-back surgeries. What is less known, however, is that there’s plenty of downtime for a surgeon between surgeries. Technically, a surgeon is not needed in the OR room when the patient is being prepped, or closed up.
Using AI and machine learning can enable the scheduling system itself work out the most efficient use of multiple OR rooms. It may find that based on the resources available, having back-to-back surgeries in the same OR room is the best schedule for a surgeon. Alternatively, it may also find that using a Knife to Skin based schedule will actually allow for over 100% use of the surgeon’s time.
Practices look at various factors, from patient’s health information (age, BMI, etc) to more technical aspects of a surgery such as surgery length and time it takes to clean up an OR room. Using this information will be key for a practice to create a successful AI system that yields efficient results and reduces overtime.
Planning for the best
One potential flaw in manually building a surgery schedules is ensuring the correct pre-surgery care plan and clearances needed. This is often determined by the health team, and not the surgeon. Certain factors may not be accounted for and critical tests missed. This can end up delaying the surgery or have grave consequences for the patient.
Using data fed from medical journals, EHR databases, and previous surgical procedures can help teach the system all the factors that must be considered. The AI system will analyze the patient’s health data to trigger alerts for the proper pre-surgery care plan and tests required. It would also alert the schedulers to any errors made by the staff.
The right tool for the job
Surgeons, like the rest of us, are creatures of habit, and will often use the same tools for the same surgery. While these may be educated decisions, they may be missing out on opportunities to use the best devices for their procedures. This can impact the efficiency and outcome of the surgery.
AI can step in and recommend the best equipment for the procedure in question. Similar to the planning stage, big data can be fed into the algorithm. It will be important that the system takes into account the patient’s age and medical record to be able to properly assess the impact different devices will have. It will also alert the scheduler to any equipment accidentally omitted.
AI and machine learning have the potential to dramatically change how practices schedule their surgeries. It is important for managers to understand the factors relevant to maximizing their resources. By creating the right algorithm and applying proper use of machine learning, practices will be able to increase efficiency and deliver better patient care.