How do large businesses solve very hard day-to-day operating problems in order to assure the lowest costs, the best possible revenue outcome, and the best possible customer satisfaction?

Major organizations are attacking these complexities by applying higher levels of analytics all the time. Consider just three examples.

Walmart uses cutting-edge technology and analytics of inventory management to research, establish best practices, monitor, and adjust finished goods inventory, transit inventory, buffer inventory, and anticipation inventory across 10,623 retail stores and 380 distribution facilities worldwide.

Airlines use cutting edge technology and analytics for scheduling to research, establish best practices, monitor, and adjust which individual pilot will make up each two- or three-pilot team in each cockpit for each flight, considering rules and regulations, seniority, fixed schedules, flexible schedules, standby schedules, days off, early standby status, pilot fatigue, and flight delays and cancellation.

Apple’s cutting edge supply chain technology and analytics have been so successful in researching best practices and monitoring timely, efficient, cost-effective access from suppliers in 43 countries across six continents that Gartner, after awarding Apple the top supply-chain ranking five years running, created a special Masters Award for Apple.

This is how some of the most sophisticated companies in the world use operations research and analytics to continually monitor and improve their most complex, mission-critical, and resource-intensive operational challenges.

How do America's hospitals relate to this level of technical capability?

Wicked Problems and Solvable Challenges

Coming out of Covid, hospitals have struggled to achieve sufficient levels of financial performance. Forty percent of hospitals have negative operating margins. Expenses continue to rise faster than revenues. The underlying causes of these financial struggles are some of healthcare’s most wicked problems, including labor disruption, payer chaos, the continuing shift from inpatient to outpatient services, alternative care options, and pervasive difficulties in patient access.

These issues are appropriately characterized as wicked problems: problems that have not been solved with technical capability, problems that have no clear solution, and problems for which a solution intended to fix one aspect may well make other aspects worse.

But given hospital financial performance coming out of Covid, many complex operational issues also afflict the hospital business—issues that look very much like the operational challenges we cited that Walmart, airlines, and Apple solve with sophisticated research and analytics.

Applying Operational Research and Analytics to the Hospital Setting

Take, for instance, staffing an operating room. In a 2023 article in the Annals of Surgery, researchers from the University of Chicago identified numerous barriers to staffing to optimal levels of familiarity and stability, including the frequent need for morning adjustments due to unexpected circumstances, scheduling limitations when a single OR is scheduled for back-to-back cases in different specialties, variation in technical skill among all team members, and lab or unionization rules, among many others.

Here is how advanced analytics of the type used by large commercial businesses could be used to fix this long-time issue affecting hospital revenues, expenses, and outcome. 

Volume Predictions

Before determining OR staffing, hospitals need to determine expected volume.

Hospitals traditionally struggle with accurate predictions of volume, especially during times of change. Analysts often over-rely on rolling forward the previous year’s performance, coupled with gut assumptions about volume inflation.

Operations research approaches forecasting differently, creating models that identify specific drivers of historical data that can be incorporated into a forecast to more accurately predict future demand, including, for example, flu data, school calendars, holidays, staff vacations, and weather. For each of these factors, and combinations of these factors, the model then describes and uses the factors that most strongly influence volume predictions.

Such a model can provide a level of historical accuracy that gives hospitals volume figures in which they can have a high level of confidence. Models can generate forecasts at an hourly and department level all the way up to a multi-year system level, which creates consistency across various planning efforts. These models can adapt in real time to changing conditions to provide better estimates without added effort.

Perioperative Scheduling

With more accurate forecasting, an organization can use operations research to better align resources with anticipated demand, for example, in perioperative scheduling.

What are often unknowns can be assigned quantitative meaning: what types of procedures are needed, how long do these take, how much time between procedures is needed to turn rooms, how many rooms are available, what equipment in those rooms are needed (i.e., robotics), how many surgeons are available, how many anesthesiologists are available, what do block schedules look like, how long has a patient been waiting, what does physician productivity look like, and so forth.

Queuing theory and optimization can be used to understand how to assign each surgical case and to achieve objectives like minimizing cost, maximizing throughput, maximizing utilization, and others. Constraints are placed on these objectives to ensure safety, patient satisfaction (like keeping wait times to a minimum period), and availability of resources.

In these ways and others, advanced analytics can lead to expense savings, revenue generation, and, importantly, improved patient outcomes and satisfaction, along with improved employee/surgeon satisfaction.

A Necessary Strategy

Faced with wicked problems on both the revenue and expense sides of financial performance, resulting in pervasive and ongoing insufficient margins, hospitals must identify the areas of performance that they can control, and take the necessary steps, even when those steps are novel and unfamiliar.

We take for granted the enormous operational complexity involved in an everyday activity such as taking a flight. From the passenger’s perspective, we simply get our boarding passes, drop off our luggage at a kiosk, make our way to the gate, and find our assigned seats.

From the airline’s perspective, however, what happens seems almost miraculous. American Airlines has more than 6500 flights per day; for each of those flights, American tracks, manages, and adjusts as appropriate pilot and crew staff, supplies, maintenance, flight plan, atmospheric conditions, planned and actual departure and arrival times, any incidents, and interdependencies among flights. American Airlines serves more than 500,000 passengers per day; for each of those passengers, American tracks, manages, and adjusts as appropriate ticket type and associated benefits, baggage number and location, seat assignments, and flight connections.

In healthcare we have absolutely the same kinds of operational challenges, only on a much more personal level. If your flight is late or your baggage is lost, you will probably be annoyed. But if your surgery does not get done when it is scheduled, you will be far more upset and for good reason.

There is no doubt that improvement in the most intricate areas of hospital operations can make a very big difference in revenues, expenses, patient experience, provider experience, and clinical outcomes.

Understanding the kind of advanced analytics techniques used in other complex businesses, and bringing these techniques forward, is an absolute necessity for properly managing today’s hospital organization in order to compete, and in order to overcome difficult economic and operating environments.