Tuck professor Laurens Debo finds that well-calibrated wait-time announcements improve the patient experience.
Of all the occasions where you have to wait to be served—say, at a restaurant, on a customer-service call, or in line for a ride at an amusement park—the most unpleasant is probably the wait to be treated in the emergency department (ED) of a hospital.
At Tuck, Laurens Debo teaches the Management of Service Operations elective.
Not only are you likely in pain and feeling anxious about your condition, but other people with more acute injuries who come in after you will likely be moved ahead of you in the queue. This situation is made even worse by the current state of the health care system, where hospitals are increasingly struggling with overcrowded EDs and long wait times. Unsurprisingly, these underlying factors have negatively impacted patient satisfaction. This is not only a problem for patients but also for the finances of hospitals, as the Centers for Medicare and Medicaid Services adjusts its hospital reimbursement rates according, in part, to patient satisfaction scores gleaned from patient surveys.
One way to improve patient satisfaction in the ED is to reduce wait times by expanding the capacity of the hospital so it can treat more people in a given amount of time. A less expensive option is to engineer the waiting experience so it’s a bit less anxiety producing, by giving patients more information and finely calibrating their wait-time expectations to the conditions in the ED. The latter route is basically a queuing problem, which is one of the main research areas of Laurens Debo, professor of operations management at Tuck.
In a paper that was the runner up for the INFORMS Best Working Paper Award in Behavioral Operations Management, Debo and a group of co-authors conducted an experiment in an urban ED where they gave patients personalized wait-times upon checking in at the triage desk, and then studied how those wait-time announcements impacted patient satisfaction scores.
The foundation of Debo’s experiment rests on Prospect Theory, which was developed by the renowned social scientists Amos Tversky and Daniel Kahneman. Prospect Theory posits that humans evaluate experiences relative to a reference point (i.e., their expectations) and that losses compared to the reference point are counted heavier than gains. For example, if a patient entering the ED is told the wait will be one hour and it’s actually 90 minutes, their disappointment will be felt greater than their happiness if the wait is actually only 30 minutes.
The losses count more heavily and negatively than gains of the same duration count positively for the customer experience, Debo said.
The losses count more heavily and negatively than gains of the same duration count positively for the customer experience.
To test how Prospect Theory might play out for patients in a real ED setting, Debo and his co-authors devised a way to estimate wait times, via an algorithm, and then announced those wait-times to patients. But they varied the accuracy of the wait-time announcements, with some patients receiving the base-condition estimated wait-time (50th percentile) and others receiving estimated wait-times that were at the 70th or 90th percentile (in effect, overestimations). The idea was that the announced wait-time would become a patient’s reference point, and that their satisfaction would be tied to how the outcome compared with their reference point. The co-authors theorized that the announcement would have an “initial effect” on patients of reducing stress by reducing uncertainty, but it could also potentially create disappointment if the announcement is longer than what they expected before they received the announcement. The announcement would also lay the groundwork for the “end effect,” which is their happiness or disappointment with the actual wait-time.
They found that patients reported a 21 percent higher satisfaction score, on average, when they were informed of their wait-time, but that this effect varied widely by the announcement type. Interestingly, the patients who were given the 50th percentile wait-time estimation did not report increased or reduced satisfaction. On the contrary, the patients whose expected wait-time was moderately overestimated (70th percentile) reported the highest satisfaction. These results are nicely summarized in the title of their paper: “Under-Promising and Over-Delivering to Improve Patient Satisfaction at Emergency Departments.”
Under-promising and over-delivering is actually not a new strategy in the service sector. For a while now, airlines have tried to land 15 minutes ahead of schedule, to make their customers feel good about the flight. But this approach is novel in the ED setting, where it’s both challenging to issue an accurate announcement and yet very important to balance the interests of keeping patients informed and not disappointed. Now that Debo and his co-authors have a sense of how patients respond to padded wait-times, they are expanding their research to look into wait-time announcement strategies without padding because, someday, patients will figure it out.
If we are always padding the wait-time announcement by 30 minutes, Debo said,
when patients hear the announcement, they can just subtract 30 minutes and the strategy becomes moot.