"If a portion of my compensation is based on patient satisfaction then to maximize my reimbursement I must provide to my patients everything they want. And if that means services they don't actually need, be it an antibiotic or x-ray, so be it. If I say no they are mad (unsatisfied) and it will cost me money. These are the rules of engagement the system has created and thus these are the engagement rules I will follow." 
This is an example of many conversations I have had with physicians working within our broken healthcare system, and this is an example of an unintended consequence associated with protocols implemented by well-meaning individuals who are trying to improve the system but without understanding the complexity of that system.
When a physician is triple booked every 15 minutes and required to generate 30 relative value units (RVUs) per day, and on top of that is being assessed on patient satisfaction and other measures, they find it easier to simply provide the answers the patient wants to hear and move on to the next patient.
But at what cost?
At the cost of patient's lives, patient safety, increased healthcare costs, decreased access for those who truly need care ... all while the system is incentivizing and striving for improved patient satisfaction.
Shannon Brownlee has written extensively on the cost and dangers of overtreatment:
"Overtreated debunks the idea that most of medicine is based in sound science, and shows how our healthcare system delivers huge amounts of unnecessary care that is not only wasteful but can actually imperil the health of patients."
We can do far better.
- When a physician (or other clinician) has limited time with a patient
- When there is little connection, no relationship, no trust
- When we as healthcare leaders create systems that do not allow the patient to tell their whole story to a physician they trust
- When we create models where a patient doesn't know who their doctor is and lacks an authentic relationship with their physician
- When we create models where teams have ever-rotating physicians, nurse practitioners, physician assistants, and other practitioners all in the name of creating a team-based model (see patient-centered medical home)
- When all of this happens, then we are the ones who must look in the mirror and recognize we are placing our patients, our families, our communities in danger and wasting limited resources in doing so.
As healthcare leaders we need to create models where physicians and patients have the time to develop relationship and trust. We must create a model where the patient's whole story is heard and understood. We must create a safe place for the patient and physician to share in decision-making and dialogue about treatment options.
In such a place, when the patient asks the physician a question or for a specific treatment option the answer can be: "I don't believe that would be in your best interest, here is why, and here are other options that we should consider. Let's discuss and together determine the best path to follow."
This discourse would lead to improved patient satisfaction (because of the authentic relationship and trust) and improved patient engagement--not to mention increased physician satisfaction, decreased cost of care, decreased utilization, better patient adherence, and improved patient safety and outcomes.
If we want to improve patient satisfaction without the negative side-effects then we must change the system. If we want to change the culture of healthcare then we must change the model.
To honor our commitment to improving the health of our patients, families and communities, we must be mindful of the complexity of the system, understand the unintended consequences, and focus on those things that bring physicians and patients closer--and rail against anything that hurts the physician-patient relationship.
Thomas H. Dahlborg, M.S.M., is chief financial officer and vice president of strategy for the National Initiative for Children's Healthcare Quality (NICHQ), where he focuses on improving child health and well-being.