In our field, there is a continuous focus on creating an environment of accountability. Thus, if we are judged on how well we perform, outcome measures are essential to determine if success occurs. Unfortunately, often in healthcare, we focus on outcome measures that are not providing “the full picture,” or we attempt to succeed in areas that might create unintended behaviors in our patients.
In all fairness, using a surrogate marker is one way we can define success and therefore accountability. However, when we elevate our learnings concerning such a methodology, or if the marker becomes less relevant, we have to find a different solution. As newer treatments evolve, using the way a surrogate end point responds instead of actual outcomes may lead us to inappropriate conclusions. An example is controlling diabetes and HgA1c. For improvement in diabetes outcomes such as decreased micro and macro vascular events, there is a focus on glucose control. Hence, as the HgA1c level correlates to glucose level, the logic is that if an intervention lowers this number then fewer events will occur, which has been shown historically. However, with newer treatments, there is a disconnect between outcomes and blood levels. In other words, there is an improvement in vascular events, regardless of HgbA1c reduction. From a scientific and treatment perspective, this is terrific news. However, when basing the definition of quality treatment on a reduction of solely a specific number, and then tying reimbursement to it, we are essentially separating our accountability measurement from the outcomes desired.
Another aspect of measurements that can be problematic is relying on patient-reported outcomes concerning their interactions with physicians and hospitals. I can recall sitting in a board meeting discussing how to improve satisfaction scores of our physicians. One brave physician and board member stated, “Regardless if it is the right treatment, let’s just direct all of our doctors to fulfill whatever the patient wishes.” Though we all chuckled, he introduced an important dichotomy. When we are managing multiple dynamics simultaneously, we must also remember to examine the unintended consequences. For example, when a person has a common cold, frequently they request antibiotics, as they believe that is the best treatment. When we say “no” and explain the medical reason, we run the risk of being correct clinically, while at the same time creating customer dissatisfaction, therefore offering better health, yet a poor satisfaction score. This scenario plays out in the hospital as well when we focus on decreasing the length of stay when patients wish to stay longer. Although this frequently occurs for myriad reasons, unfortunately, not all relate to medical needs.
An additional example is the current concern of the opioid epidemic as it relates to the overuse of opiates to decrease pain. Since there is a necessary and often immediate focus on reducing pain, the management of this condition became an outcome measure. Stepping back, however, sadly, the tools we implement, can lead to addiction. Once again, resolving one medical challenge while adding a secondary issue. Instead, we must treat pain while reducing dependence. Therefore, if we are going to measure accountability in these instances, the measurements need to be two-fold, getting pain under control and decreased addiction rates.
To measure accountability properly, the key is not to discount specific metrics or outcomes, but to look at the bigger picture and to always focus on adjusting our metrics to support continuous improvement. In addition, we have to be agile and understand that changing our measurements is also important to better support those we serve, instead of doing what we have always done for the sake of avoiding change.