When I did some work with a major hospital in the Netherlands and was preparing my first intervention, I reviewed some of the earlier work I did in the health care industry, that included working with executives from the Alexandra Hospital in Singapore. To update myself I also explored their website and found out that they are using a publicly accessible webcam for their Emergency Services waiting room. Why would you want to do that?
Before I explain that, I should tell you that Alexandra’s management team, under leadership of Mr. Teng Lit Liak, was always serious about continues improvement. They had done some amazing things including analyzing and improving the flow of patients from the moment they leave their house for a same-day procedure to arriving at the hospital and going back home again. This was in 2002. When I visited Mr. Liak a few years ago, he was still at it with his management team, now using insights from Toyota, very carefully translating them with his people into possible applications that would help improve the patient’s experience at Alexandra Hospital.
Apparently one of the ideas that emerged was putting up a web-cam in the emergency room that would provide every 5 minutes an update on the average waiting time for patients and the absolute number of patients waiting to see a doctor. The hypothesis is that people who have a minor injury and would like to be treated in a hospital would first check the waiting time on the internet before they would go to Alexandra hospital. Based on what they see on the webcam they might make a more informed decision, e.g., go to Alexandra Hospital, or go to another hospital or clinic. This idea is based on the so called “Hejunka” principle derived from the Toyota Production System. “Hejunka” means to level or make smooth and in the emergency room context could potentially lead to better management of the relative unpredictable peaks in waiting time that occur. Possible advantages might include that patients are served more timely and a better match of people resources (number of doctors and nurses to patients).
The whole idea is based on a self-organizing patient flow, i.e. the patient making decisions based on one extra piece of information that could influence the actual demand of patients at Alexandra hospital.
The question is, will it work? It was decided to just try it. In October 2005, the average waiting time was 1 hour and 43 minutes and in October 2006, a month after implementing this system, the waiting time for 95% of the patients had fallen to 1 hour and 12 minutes. The implementation team agreed that it was not clear that the webcam contributed to this result, in other words the cause-and-effect relationship was not apparent, but they decided to continue the use of the webcam. To their surprise, other hospitals copied their idea which they had not predicted. But the usual publicity in newspapers and other media led to an emerging system that started to influence where patients would go for emergency treatment. While the who came up with the idea could not scientifically prove this experiment had led to lower patient waiting times they believed it.
I think this is a classic example where cause-and-effect relationships are unclear, the adoption by other hospitals was unpredictable and a new system of patients making more informed decisions about where to go was emergent. The outcome is unpredictable. You are dealing with a complex system with so many actors that your best bet is to do a safe-to-fail experiment at minimal cost of S$ 400 for a webcam! The moral of the story: Just do an experiment and watch what happens and start believing.
This is a revised version of a blog contribution for Cognitive-edge.