Bob Luttman, Robert Luttman & Associates
Introduction
(Briefly) Revisiting VMS Design Issues
Gateway Documentation - Normal Days
Gateway Documentation - Insert Days
Introduction
After weeks of preparation, pathway development, endless meetings, and a dozen format drafts the clinical pathway system was launched. The first pathway (uncomplicated cardiac surgery) was implemented, first as a pilot for up to 50 patients. Those 50 patients cam and went, the variance data was collected, the numbers crunched, and the first variance report generated.
Disaster.
LOS was on average 1 day longer than planned as only 40% of the patients met the LOS goal. Yet the variance data (only critical variances and presented in a nice pareto histogram) showed nothing meaningful. No causality was apparent. The correlation matrix contained the first glimmer of an answer: multi-collinearity.
Worse: the nurses hated the damn thing. The documentation was too complicated, too time consuming, and - by the way - what do we do when the patient stays past the pathway?
Those issues we discussed last week smacked us right in the teeth. Back to the drawing board. Quickly, the other pathways were scheduled for piloting in a couple weeks.
Out of this disaster arose the Gateway variance management system. The traditional model was abandoned and the system designed from scratch, with heavy physician and nurse input. In this week's session we will present the Gateway model and critique it with respect to the issues we discussed last week.
As always, we will solicit your comments, questions, and feedback at the end.
After you finish your homework.
