John Dover had just completed an intensive course, "Statistical Thinking for Continuous Improvement," that was offered to all employees of a large health maintenance organization (HMO). There was no time to celebrate, however. Dover worked as a pharmacy assistant in the HMO's pharmacy and he was under a lot of pressure, because his manager, Juan de Pacotilla, was about to be fired. Pacotilla's dismissal appeared imminent because of numerous complaints and even a few lawsuits over inaccurate prescriptions. Pacotilla now was asking Dover for his assistance in trying to resolve the problem. "John, I really need your help," said Pacotilla. "If I can't show some major improvement or at least a solid plan by next month, I'm history." "I'll be glad to help," replied Dover, "but what can I do? I'm just a pharmacy assistant." "Your job title isn't important. I think you're just the person who can get this done," said Pacotilla. "I realize that I've been too far removed from day-to-day operations in the pharmacy, but you work there every day. You're in a much better position to find out how to fix the problem. Just tell me what to do, and I'll do it." "But what about the statistical consultant you hired to analyze the data on inaccurate prescriptions?" asked Dover.

"To be honest, I'm really disappointed with that guy. He has spent two weeks trying to come up with a new modeling approach to predict weekly inaccurate prescriptions. I tried to explain to him that I don't want to predict the mistakes; I want to eliminate them. I don't think I got through, however, because he said we need a month of additional data to verify the model before he can apply a new method he just read about in a journal to identify 'change points in the time series,' whatever that means. But get this, he will only identify the change points and send me a list. He says it's my job to figure out what they mean and how to respond. I don't know much about statistics. The only thing I remember from my course in college is that it was the worst course I ever took. I'm becoming convinced that statistics really doesn't have much to offer in solving real problems. Since you've just gone through the statistical thinking course, maybe you can see something I can't. I realize it's a long shot, but I was hoping you could use this as the project you need to officially complete the course." "I used to feel the same way about statistics, too," replied Dover. "But the statistical thinking course was interesting because it didn't focus on crunching numbers. I have some ideas about how we can approach making improvements in prescription accuracy. I think it would be a great project. But we might not be able to solve this problem ourselves. As you know, there is a lot of finger-pointing going on. Pharmacists blame the doctors' sloppy handwriting and incomplete instructions for the problem. Doctors blame the pharmacy assistants, who do most of the computer entry of the prescriptions, claiming that they are incompetent. Pharmacy assistants blame the pharmacists for assuming too much about their knowledge of medical terminology, brand names, known drug interactions, and so on." "It sounds like there's no hope," said Pacotilla. "I wouldn't say that at all," replied Dover. "It's just that there might be no quick fix we can do by ourselves in the pharmacy. Let me explain what I'm thinking about doing andhow I would propose attacking the problem using what I just learned in the statistical thinking course."

 

Case Question: 

How do you think John should approach this problem, using what he has just learned? Assume that he really did pick up a solid understanding of the concepts and tools of statistical thinking in the course.