Running an Operating Room

We had a grand rounds on operations management, and the discussion helped me realize how complex it is to "run the board" - that is, schedule and manage the surgical flow of 21 operating rooms. Even simple decisions become convoluted. When should the day end? Should there be an end time? Having an expected end-time simplifies staffing; shifts can end at 6pm except for a skeleton crew to accommodate emergencies. Setting and meeting expectations helps well-being, allows planning, and keeps operations organized. But this must be balanced against the increased revenue that can be obtained by letting surgeons book as many cases as they want. The more cases a hospital does, the more money the hospital makes. And if a surgeon is willing to operate, why restrict him to 6pm? Yet if there is no set end-time, how do you manage staffing of anesthesiologists, nurses, and technicians far in advanced? How can we plan our lives if surgeons can add on a non-emergent case, and we have to be available to do it?

Then there are problems at a more micromanagement level. Say you have one room that has 4 cases scheduled: one very long, two short, and one of variable duration. Do you schedule the longest case first? This would allow the other cases to be shuffled to different rooms if other rooms finished earlier. It would also mean that if a case had to be canceled because the room ran over, it would be a short case and not the long one. But there are risks; what if the long case is canceled, and none of the other cases are ready to go because they expected to be in the afternoon? How do you fill the gap?

Lastly, there are a lot of problems specific to medicine. As opposed to a production line or a senator's schedule or the fast food rush, surgeries can be incredibly variable in length of time. The same operation can take much longer or much shorter than expected, and it's hard to accommodate for this. The most extreme example is pancreatic cancer surgery. Despite a comprehensive workup, a surgeon can open up the belly expecting to take out the cancer and find that it is much more widespread than he thought. In this case, the best decision for the patient is to close the abdomen without doing the surgery because surgery will not help. But from an operations standpoint, this changes a 6 hour case to 30 minutes. How do you deal with that?

Intuitively, I favor a computer algorithm or calculation that can look at past cases, take into account all these rules and preferences, and spit out an answer. But after listening to the ground rounds, understanding the pros and cons of different approaches, and seeing the practical result of our daily case scheduling, I'm not sure it's that easy.