“Healthcare systems face a unique paradox: Scarce resources such as operating rooms and infusion chairs are overbooked yet underutilized on the same day – every day. Even healthy “average utilization metrics” mask the reality that the 80 percent utilization metric is achieved by 40-50 percent utilization during a portion of the day counterbalanced with 100 percent utilization for some hours. This problem is only likely to get worse given the demographics of an aging population and the Affordable Care Act adding millions of patients into the system,” says Mohan Giridharadas, Founder & Chief Executive Officer, LeanTaaS.

LeanTaaS is a solution provider at the marcus evans National Healthcare CXO Summit Fall 2016, in Los Angeles, California, October 16-18.

What are the complex operational problems at hospitals today? And how would you address them?

The number of patient encounters is likely to increase by 30-40 percent over the next 5 years. At the same time, value-based reimbursement will force hospitals to do more with their existing resources – instead of adding staff or facilities.

Units that operate near 100 percent utilization enter an unstable state – wait times climb, more patients are in the system and nurses get swamped. The core reason this occurs is that patients are scheduled in a way that makes it impossible to optimize. Each appointment is a one-on-one negotiation between the scheduler and the patient – without any attempt to optimize the placement based on the mix of patients likely to be in the unit at the time being negotiated. This is analogous to making a stock recommendation without looking at the other stocks held by the investor. Using lean and data science, healthcare providers can optimize the fit between the demand curve and the available resources at a fine level of granularity in order to improve patient flow while lowering wait times.

Conventional process improvement methods (e.g., lean and Six Sigma) tend to use spreadsheets and “averaging arithmetic”. Healthcare operations are incredibly complex and variable and therefore require advanced mathematical algorithms for forecasting, simulation, optimization and machine learning in order to drive actual impact. The trick is to use the best of lean thinking combined with real data science to improve capacity and flow.

How can the use of analytics or data science help reduce healthcare costs?

Most Business Intelligence (BI) tools visualize prior performance – while helpful, it cannot drive performance.  In hockey, skating to the location of the puck from 5 seconds ago is futile – you have to skate to where it will be 5 seconds from now. Using predictive analytics and optimization, hospitals have to anticipate bottlenecks that will occur over the coming hours/days and take steps to address it.

How can operating room utilization improve? What financial implications can that have?

Operating rooms are the most valuable resources in a hospital, generating half the revenue. Most hospitals use block scheduling - while convenient, allocating a block of the right length to the right surgeon (or service line) is a complex problem. Block scheduling repeatedly requires finding a replacement for the surgeon who is not available to use a future block. Reassigning the released block to another surgeon often results in an abandoned or underutilized block. The value of optimizing block schedules is enormous – a hospital with 20 or 30 operating rooms could generate an additional USD
4-6 million in revenues each year through a 2-point improvement in utilization.

What other resources do hospitals underutilize?

There are hundreds of scarce assets in a hospital including operating rooms, inpatient beds, infusion suites, imaging assets and specialists. Most assets have a very “peaky” profile – either a peak in the middle of the day or the middle of the week. Peaky profiles create a “rush hour effect” that is susceptible to the smallest thing that goes wrong – a minor event delays thousands of people. Patient care is highly variable with a risk of delays in lab results, pharmacy, clinics, doctors, or sick calls. At “rush-hour”, this creates a disruption to the experience of all of the other patients.

Imagine if you could persuade a small percentage of drivers to leave a few minutes earlier or later than their usual departure time. Without requiring a dramatic shift in behavior, it is possible to achieve an enormous impact on the overall flow of traffic (similar to what we all experience during a bank holiday or a school holiday). Queuing systems operating at the edge of peak capacity are unstable– a small shift (of even 15-30 minutes) in the timing of the demand pattern can unlock dramatic improvements in wait times.

Hospitals can (and should) exercise their ability to optimize the placement of appointments to achieve these outcomes.