I am of the belief that every ACO must be physician-led. We must depend on them not only for clinical improvement, but also for developing a culture of improvement. Culture is vitally important. Culture trumps dollars, technology, data, and about anything else you would use in clinical medicine.

If I was getting into the ACO business, I would start recruiting clinicians that embrace these characteristics:

1. Team leadership. Every doc is a leader to some degree, but we are looking specifically for physicians who enjoy leading teams. Start with a primary care doc who has done a couple of improvement projects in the hospital or clinic. Pay her for the time she dedicates to the project. Your success will depend on shared baselines or other models of care.

2. Willingness to change. There are plenty of physicians who are skeptical and worried, but will participate. Put together a great message of change and how it will be better for the patients. Then, deliver the message over and over until they internalize it. Don’t waste too much time trying to change the docs who are defiantly resistant; some people are just unwilling to change.

3. Passion for patients. This person is who I really want in my ACO: somebody who is a strong patient advocate and takes a positive stance on change. They likely understand it when you start talking about how quality controls costs and how variation in care is the enemy. She may turn out to be great leader.

4. Don’t fight the data. It is good for clinicians to question the data in the early stages. The job of the data architect is to reconcile their concerns and tweak the platform so that the data is relevant, reliable, and timely. That can take a while, and you will constantly refine your data. You need docs who understand this process and will engage the data when it becomes available. Data is a strong tool for physician engagement.

5. Use the right technology, and only the right technology. Technology is costly, and you are in a game of low-cost. You must have information systems, but you may not need three robots. Telemedicine may be a good idea, but if you can’t easily determine which of your diabetics are late on their A1c, you need some help. The key is to prioritize. Start the messaging early on: “We are making investments that help you improve quality and reduce cost.” That is your mantra. Own it.

Tommy Prewitt is director, Health Care Delivery Institute, HORNE LLP.