Background:
We identified a number of quality, patient satisfaction, and staff satisfaction issues that all seemed to be linked to throughput challenges. We had trouble getting patients out of the hospital on discharge day, which set off a number of downstream effects including backups in the Intensive Care Unit (ICU) and Emergency Department (ED). These challenges created frustration and safety issues for patients and agonized staff. We took on this challenge to improve hospital throughput. A natural challenge for our throughput issues was late inpatient discharges. Our idea of a simple project and with no additional monetary investment would target a large reduction in inpatient discharge times by requiring the smallest group of people to make the fewest changes as possible. Through a deep dive into our own discharge data, we found an opportunity that fit our concept of an ideal project.
Methods:
First, we greatly simplified the problem by narrowly defining our objective to improving the discharge order time rather than the larger and more complex patient departure time. For the hospitalist group, it meant that physicians would be measured on something they nearly completely controlled. The more significant innovation was in how we measured the percentage of discharges before 11 a.m. For feedback and motivational purposes, we only considered “expected discharges,” defined as discharge orders occurring between 6 a.m. and 2 p.m. The reason for limiting the timeframe was to eliminate any incentive physicians might have to hold onto unexpected discharges late in the day until the next morning. Hospitalist leadership felt that virtually all of the expected discharges were being rounded on by 2 p.m. each day. Therefore, we assumed that discharge orders occurring after 2 p.m. were primarily unexpected discharges. A monthly report was sent to hospitalist leadership detailing how each individual hospitalist performed for each of the previous six months. The report included their group performance as well as the group’s mean discharge order time and the mean patient departure time. Hospitalist leadership held discussions with colleagues who weren’t making progress.
Results:
We tracked detailed results of this project for a seven-month period from February 2015 to August 2015. In February, only 37% of hospitalist discharges occurred before 11 a.m. By the end of August, that number had risen to 70%, an 89% improvement over February. The impact of this improvement can best be seen in the mean change in both discharge orders and patient departure times over that seven-month period. There are significant cost savings that stem from this project. On the inpatient side, we use an estimated cost to the hospital of $400 per patient day. That equates to $16.66 per patient hour. The 2,623 inpatient hours saved over this seven-month period are valued at $43,718.
Conclusions: In contrast to earlier studies showing no impact of earlier discharge order times, we found substantial impacts. Prioritized rounding had significant and reverberating effects throughout Hospital in terms of new efficiencies both on the inpatient and ED side. The improvement in patient satisfaction and cost savings was a huge bonus. The project was well worth the effort, and we highly recommend it for other hospitals struggling with throughput challenges that are similar to those we faced at our Hospital.
