May 1997

Benchmarking the perioperative process.

Rotondi AJ, Brindis C, Cantees KK, DeRiso BM, Ilkin HM, Palmer JS, Gunnerson HB, Watkins WD; Journal of

Clinical Anesthesia 1997; 9: 159-169.

[ see abstract below ]

This paper presents the development of a patient routing system that tracks patients as they move through the perioperative experience.

You may notice that the senior author on the paper being reviewed is one of our esteemed board members, Dr. W. David Watkins. The work of Dr. Watkins' group in Pittsburgh has generated much discussion among those who understand what an incredibly complex process perioperative patient flow really is.

You can only positively impact flow if the problem you address is the primary bottleneck (i.e. the current rate-limiting step) and you do not create new bottlenecks either upstream or downstream of the problem you are addressing. And that is often what happens. Until now, we had no way to routinely measure those effects. Dr. Rotondi et al. have described a system that identifies where delays currently exist, and by which we may monitor the impact of any operational change in the OR on the entire complex process of perioperative patient flow.

The concept is actually deceptively simple, and that accounts for the average reported compliance in using the system at 95%. Each patient has a bar-coded card that is scanned at one of 17 time points (see table 1) in a variety of venues. The system tracks a same day surgery patient as he enters admission, goes to preoperative holding, into the OR, out to the PACU, and on to the ward or ambulatory discharge area. At any point, the system may inhibit efficient flow. This paper primarily deals with the concept of system problems and analyzes the data so that those problems can be addressed.

The greatest variance is thought to indicate the areas most likely to benefit from investigation (i.e. how can one patient go through here fast, but not all patients?) However, it important to recognize that systematic problems may be present as low variation.

For instance, if the minimum stay in the phase 1 PACU is 45 minutes for ambulatory patients, but they are ready from 5-44 minutes routinely, there would be no variance as a result of a system that keeps all patients in phase 1 PACU longer than they need to be there (a not uncommon problem in many hospitals). In the authors' institution, the time intervals from same day surgery admission, to patient ready for transport, to preoperative holding was a long time, with great variation. They recognized that the lack of a preoperative admission clinic was slowing down the process.

Since they have now begun such a preoperative assessment clinic, they can measure the before and after times, and see exactly what efficiencies this new service has generated. The particulars of what needed to be fixed at the authors' institution are less important than the description of a methodology that can be universally employed. And universally employing this system at many institutions will allow benchmarking among like institutions so they can know if their performance has room for improvement.

The system is computerized, and one could easily envision using the simple system described in this paper to further analyze perioperative patient care. Delays can be related to particular types of patients, or related to particular physicians and particular procedures. The ability to identify groups of patients for whom the process can be improved, or groups of surgeons/anesthesiologists who can improve their performance, is an important and obvious extension of this current work.

Professionals will improve their performance if the data fed back to them is accurate and timely. Until now, even advanced management groups in hospitals did not dissect out the details of patient flow as this system does. Now that Dr. Watkins's group has shown us the way, will we take the trouble to implement his simple solution?

Return to the Current Literature Review Front Page, or read the abstract:

ABSTRACT