also see: predicting scheduled starts for surgery
For this OR scenario, the PreOp and PACU graphs are particularly useful if the three surgeons are working at the same facility. However, you might want to consider the situation in which each has his own private OR. Under what conditions (type of contract and type of OR scheduling) would each of these surgeons be more likely to leave private practice and become an employee of a hospital?
With minimal modifications, the techniques used below would also be applicable to the ER, OB, Radiology, Radiation oncology, GI suite or Floor units. It’s adaptable to hundreds of unique factors -each with multiple effects- for each case during a typical OR schedule. It can easily be applied outside the field of healthcare.
I have a separate concept graph that makes most of this quite clear and self-evident that’s part of a poster presentation at the 2013 Healthcare Systems Process Improvement Conference in New Orleans. Real time simulations with case delays, case add-ons, 50 modifiers for a single case, real options, portfolio scheduling (right surgeon…wrong place or time?) and fitting emergencies into the on-going schedule will be simulated.
The first graph gives the baseline for 3 different constraints using identical case loads.
PreOp and PACU shows the patients in those areas. The colors and names of the surgeons show which patients are related to which case. The horizontal bars below PACU are labeled to indicate which surgeon did which cases. The large vertical bars are each 60 minutes wide. Each surgeon within a horizontal bar goes left-to-right from big blue “O” to big blue “O” in sequence, sometimes flipping rooms to do so.
‘Schlicter’ (1,2,2) = 1 surgeon, 2 anesthetists, 2 rooms with nurses (flipping rooms, surgeon is the constraint)
‘Smith’ (1,1,2) = 1 surgeon, 1 anesthetist, 2 rooms with nurses (flipping rooms, anesthesia is the constraint)
‘White’ (1,1,1) = 1 surgeon, 1 anesthetist, 1 room with nurses (no flipping, room is the constraint)-normal)
Notice that for this baseline case Schlicter finishes @14:00, Smith @16:00, and White @ 17:00
<click on graph to enlarge>
The following graphs will show how intra-case changes (potential for lean improvement) can have major or no effect on the case durations and significant effects on the different agents involved (surgeon, anesthetist, or nurse). The purpose is to emphasize the use of constraint theory and simulations to improve the focus of lean activities (and data analysis) to where they have the most desired effect to optimize personnel, salaries, effort and time.
This next graph increases the surgeon’s operating time from 60 minutes to 75 minutes. Regardless of the constraint (surgeon, anesthesia, room) the time till cases are finished have all been extended by 1 hour. The nurses and anesthesia have an additional 15 minutes between cases for Schlicter (1,2,2). For Smith (1,1,2), only the nurses have an additional 15 minutes since anesthesia is the constraint. For White (1,1,1) there is no change in downtime for anesthesia or nurses.
In the following modification, surgeon’s time has been restored to 60 minutes. Closing time, by the surgeon’s PA, has now been increased from 30 minutes to 45 minutes. For Schlicter (1,2,2) the cases are all finished by 14:15 instead of 14:00 (15 minutes later). For Smith and White, the end times are 60 minutes later. The nurses and anesthesia have 15 minutes less time between Schlicter’s (1,2,2)cases. The nurses have 15 minutes more between Smith’s cases (1,1,2). Since anesthesia is the constraint in Smith’s (1,1,2) cases, there is no change in down time for anesthesia. For White’s cases (1,1,1) there is no change in down time for anesthesia or nurses.
Let’s go back to the baseline…. This time, anesthesia is slow for whatever reason: We increase the anesthesia patient to sleep time by 15 minutes, and the anesthesia wake up time by 15 minutes.
As you can see, Schlicter’s (1,2,2) case load time to finish increases by only 15 minutes. Where’s as Smith’s (1,1,2) are now almost as long as White’s (1,1,1). In this scenario, there’s no reason to try to run a second room of nurses unless you have 2 anesthetists (1,2,2).
This time, along with slow anesthesia, we’ll prolong nurse setup time by 15 minutes. This can happen when anesthesia and nursing are working in series instead of in parallel. As you can see, Schlicter’s (1,2,2) cases don’t suffer much with only a lengthening of the day by 30 minutes more than baseline. Smith’s cases are extended by 75 minutes, and White’s by 2 hours.
This last scenario has normal anesthesia, but retains slow nurse setup.
The last graph is the same as above with slow nurse setup, but cut time is at 7AM instead of room setup time at 7AM. Patient times in PreOp and PACU are also included.
There are hundreds of causes and many interactions in delays. Some can be minimized by proper scheduling. There are also opportunities to take advantage of faster than normal surgery, anesthesia, nursing, PreOp, and PACU. Knowing how to avoid the consequences, where to focus on improvement, and how to take advantage of changing events will result in spending less time and effort on ineffective lean initiatives in and outside the OR.
Scheduling systems should be an important consideration when buying a surgeon’s practice. Dr. Schliter is much more productive than Dr White and would require a higher buyout and salary. However, if the hospital can’t schedule (flip rooms) as efficiently as Dr. Schlicter’s private OR, then the hospital will have payed extra without realizing the gain.
Conversely, if Dr White becomes a hospital employee and expects to spend 07:00 until 17:00 in the OR with long breaks between cases during which he surfs the web and eats donuts, he’ll be upset if he has short turnover time (TOT), finishes by 14:00 and has to spend the next three hours in clinic.