It's a good question. If medicine is founded upon scientific evidence, why are there still variations in treatment protocol, even in well-understood circumstances?
While I am a physician, I also enjoy a reasonably social life. It’s not uncommon for me to share an evening out with friends or colleagues. Over the course of casual conversation, I am often asked questions related to clinical practice or for my professional perspective on a variety of topics. Recently, I fielded a question that went something like this: “When I was diagnosed, my doctor told me to do X. However, when my neighbor had the same situation, his doctor told him to do Y. Which approach is the right one, and why are they different?”
It’s a good question. If medicine is founded upon scientific evidence, why are there still variations in treatment protocol, even in well-understood circumstances?
Of course any physician will answer that some variation is appropriate and even necessary, based upon differences in physiologic, emotional, cultural, religious, and other preferences between patients. Acceptable variations in care may relate to pharmacotherapy, for instance. Every medication has possible side effects, which can vary widely, as can patients’ tolerance. Some people’s preference is to alleviate symptoms regardless of side effects, while others are more willing to live with symptoms than the impact of the medication. Other appropriate reasons for variation relate to personality. For instance, patients who can’t tolerate uncertainty need their provider to pursue an iron-clad diagnosis with vigorous testing. In contrast, a patient with a higher tolerance for uncertainty may prefer to take a wait-and-see approach in a low-risk situation before undergoing expensive or invasive testing.
In fact, if variations in care were never appropriate, then doctors could easily be replaced by robots or apps that would provide patients access to accepted guidelines and established care pathways. Clearly that’s not a viable option.
As physicians, it’s our imperative to respect informed patients’ individual needs and values as long as they understand the potential consequences of their choices. We shouldn’t impose our own preferences or biases outside of science that can cause or perpetuate variation in care. When I’m asked about the distinction between warranted and unwarranted variation I say that warranted variation is driven by differences among patients, including their preferences and values, and not on the preferences of their physician.
The problem is that unwarranted variation is still rife within our industry. And unfortunately, it’s often due to gaps in professional knowledge or organizational lapses in communicating and institutionalizing that knowledge throughout the delivery system. And where there's variation in care, in which clinical best practices are only episodically employed, there's inconsistency in outcomes.
Public Health England (PHE), NHS England, and NHS Right Care recently launched the NHS Atlas of Variation in Healthcare 20151 , which identifies opportunities to address unwarranted variation by revealing possible over-use and under-use of different aspects of healthcare. The report shows wide variations in common care situations including the prescription of antibiotics, monitoring of diabetes, and assessment of dementia. According to John Newton, Chief Knowledge Officer at PHE: “Variations are not always bad. Some can be explained by local circumstance or patient-centered care, but unwarranted variation is very different. While some patients are missing out on the right care, others are being given care they don’t need.”
While reports such as the Atlas of Variation in Healthcare can help identify areas to be addressed, we also need to take into account the shifting role of the clinician as well as shifting cultural attitudes. An unprecedented degree of choice is being introduced into the system, putting clinicians in the perhaps unexpected role of consultant-partner for their patients. Some of this change is generational, with younger patients increasingly expecting choices. Older patients, on the other hand, often say: “You’re the doctor, you tell me what to do,” when presented with a variety of options. The result is that physicians need to learn how to deliver variable care that still helps patients make informed choices that follow best practices and guidelines. In this new system, evidence-based clinical guidelines are not a set of black-box instructions. Instead, they are tools to help reduce unwarranted variation, improve systemic care delivery, and promote better individual outcomes.
Of course, this also begs the question of what we do when there is more than one guideline, evidence is in conflict, or new guidelines are being reviewed and codified. We’ll talk about that in our next blog.