Clinical Benchmarks and Their Implicit Social Pressures

It appears as if the concept of quality has implanted itself firmly into the medical superstructure. depending on whom one listens to, widespread adoption of quality improvement measures will have effects ranging from moderate improvements in delivered care to a huge cost savings coupled to enormous improvements to the entire healthcare delivery pathway. Presently, the Center for Medicare and Medicaid Services (CMS) and the Congress have mandated the adoption of quality measures be tied in some way to reimbursement to hospitals and providers. Regardless of whether this notion is sound, and reasonable people have cogent and sound arguments about both sides of such a question, the momentum is squarely behind this vision. There are two basic components to what has been termed Pay For Performance or Clinical Benchmarking. First, the “value” of medical care delivered must be quantified and second, this “quality” would be measured and physicians and hospitals would be ranked based on some scale. This ranking will be tied in some manner to payment for medical and hospital services.

Again, although there are cogent and rational arguments that are being made for and against such an “overlay” onto medical care, there has been little attention paid to the social implications of what widespread adoption of clinical benchmarking parameters might be like. Such latent social dangers are best illustrated by three “cases” as we call them in medicine. The first occurs in hospital A. Here, a 76 year old man with kidney failure undergoes surgery for the placement of a permanent dialysis access site. He subsequently develops post-operative pneumonia and remains in the intensive care unit for two months. The patient is then discharged to a chronic care facility because his family is unable to afford care for him. The second case occurs in hospital B. Here, a 27 week premature infant is delivered via cesarian section. She spends the first 28 days of her life in an intensive care unit and then dies due complications related to her prematurity. The third case is of a patient at hospital C. There, a 41 year old man is admitted for chronic low back pain where he undergoes surgery to remove a herniated disk. He is discharged that night without any complications. He returns to work seven days later with marked improvement of his symptoms.

To further our exercise, let us ask what the “value” of the care delivered in these cases is, and how would they rank in the “quality scale”? The three cases are common and many people would agree at the outset that the patients in hospitals A and B drains limited resources. Taken a step further, those people would go on to argue that better “value” could be attained for those expenditures if the ages of the patients in question were 46 years old and 39 weeks, respectively. Here then lurks the demon: because we as physicians, and we as a society have not yet deemed such issues like age, likelihood of survival and comparison to others in need of the same care as factors in deciding to treat, these people end up receiving care, and thus, counted in our statistics. Hospital A has a higher mortality rate than their neighbor hospital C, and spends more money per patient that hospital C. That is, Hospital A spends more, but delivers less quality care. For hospital B: their infant mortality rate is higher and their cost of care per infant is higher than another hospital that does not manage high risk obstetrics patients C. So in the end, the same statistics are evident: more money spent, less quality care delivered.

It has been said that data speaks for itself. This is true. What is often not widely understood is that data speak in a language all of their own. What that language translates into are conclusions. The best conclusions are the ones that are drawn directly and precisely from the data. The more loose the translation of the data, the weaker the message proffered in the conclusion. That conclusions can vary so widely from similar data sets reflects the fact that human beings derive the conclusions and the underlying reality that make up those data sets is complex. More importantly, this reality is comprised of scenarios that should inform widespread debates we as a society should be having. I am all for quality measures, as long as they engender such discussions. Until that time, such measures should not be used as a blunt tool for allocation of resources.

Lastly, to the possibility that widespread application of quality measures can help to regain some cost stability in the healthcare arena.  That the US spends more as a percentage of GDP on health care than most industrialized countries and has one of the highest infant mortality rates is an oft quoted pair of statistics that is supposed to shock listeners into resigned submission. I suggest however that there is a great way to to fix these statistics instantly: tell Hospital A and B to reject their patients, in favor of ones like those at hospital C. Choose newborns that are a bit older, or aged diabetics that are a little less aged, and perhaps a little less ill. If this makes sense, now you begin  to understand the origin of another massive current problem in medicine: the untenable annual cost increases in the medicare/medicaid programs and the access issues for the programs patients. For those hospitals, currently mostly in major cities that do offer access to the poor, indigent and aged, “quality stats” show high expenses, low quality (hospitals A and B). If the government ties reimbursement to these numbers, of all possible outcomes, one is certain: these patients will be denied access across the country.

Holding hospitals and doctors accountable for care delivered is a fine idea– but one much more complex that the persons behind the benchmarking movement currently appreciate. Unless we as a society are prepared to get our hands a little dirty and confront some of the underlying social issues behind various “quality” measures, we as a society will not be able to strive for the reasonable and laudable goal of improved care for all of us.