by Thomas Dahlborg (originally posted on www.hospitalimpact.org)
Lowering elevated blood pressure and elevated cholesterol levels are two key metrics associated with improving and maintaining health.
More and more focus is being placed on these (and other similar) metrics. Reimbursement is being tied to individual linear metrics via Pay for Performance and other programs. In 2015, CMS, through its Physician Quality Reporting System (PQRS), intends to begin penalizing physicians for not reporting data on these types of metrics. Awards and recognition are being handed out for achieving them. Physicians and physician practices are being rated on them. Consumers (you and I) are being "educated" to select our preferred provider based on these metrics.
And yet ... is such a linear approach to measuring healthcare quality truly benefiting individuals and communities?
Rodney Hayward, a professor of public health and internal medicine at the University of Michigan, has made note in a recent article that encouraging physicians to get all their high-risk patients' blood pressure down below 130/80 could actually hurt people and even lead to deaths due to physicians prescribing more and more medicines to achieve the metric. For the doctor, "they're following the quality measure they're judged on, but they're over-treating the blood pressure."
So based on that one linear metric, a physician could be "rewarded" via a Pay for Performance reimbursement program, receive recognition and awards, receive a blue ribbon on a physician rating website, consumers would be informed that this physician provides high-quality care and thus we would be more likely to select the physician as our physician. However, the physician (in this example) actually may have done more harm than good to specific patients while achieving the performance measurement.
Now let's look at the lowering elevated cholesterol levels metric. It is well-known that lowering cholesterol levels is essential for better health. No brainer right?
However, according to the Harvard Medical School Harvard Health Newsletter, "the relationships among cholesterol levels, psychological function, and neurologic disorders are complex and sometimes controversial."
Controversial?
From Discovery Health: a Duke University Medical Center study (late 1990s) found that healthy women with cholesterol levels below 160 mg/dL were more likely to show signs of depression and anxiety than women with normal or high cholesterol levels. Researchers in the Netherlands published a study in 2000 showing that middle-aged men with low cholesterol are more likely than other men to have symptoms of severe depression.
A seven year epidemiological study, shared by Dr. John Briffa on his blog The Cholesterol Truth, found that in men, low levels of low-density lipoprotein (LDL) cholesterol (supposedly "unhealthy" cholesterol) were associated with an increased risk of depression. In women, low levels of high-density lipoprotein (HDL) cholesterol (supposedly "healthy" cholesterol) were associated with an increased risk of depression.
And yet the individual linear metric we focus on is lowering elevated cholesterol levels, and in our siloed system the mental health connection is not even assessed.
So again, a physician can do an amazing job and help his or her patients achieve lower cholesterol levels and thus achieve the metric and receive the same adulation as shared above. But that physician also may be doing more harm than good when considering the bigger health picture (in this case, the mental health impacts).
Jennifer Eames Hoff, of the consumer advocacy group Pacific Business Group on Health, has said "It's the tension between how good is good enough. Do you really want to get to this super-scientific rigor that will take years and years? Or can you live with a measurement system that is really, really good and could usher in dramatic improvements in patient care?"
I don't necessarily disagree with Ms. Hoff. I simply believe we are not yet at the "really, really good" stage, and what we currently consider "really, really good" can be dangerous.
Yes, individual linear metrics can be quite helpful in context, as they are snapshots of improvement steps on the pathway to a larger health goal. To truly assess quality and outcomes, a comprehensive 360 degree view that includes individual linear metrics is essential. Anything less than that can incentivize wrong behaviors and/or incorrect approaches to care and can be dangerous. I am okay with sacrificing PERFECT for GOOD. I am not okay with sacrificing PERFECT for DANGEROUS.