Measure for Measure

September 2, 2010
Steven Zuckerman, MD

Dr. Steven Zuckerman explains why he feels quality measures are missing the mark.

I have reservations regarding what are being accepted as “quality measures,” which supposedly reflect how well we are taking care of our patients. The government is determined to purchase “quality” of medical care and not “quantity” (ie, pay for service) and require that we show the performance of given quality indicators as proof that high-quality medicine is being practiced. Most of these measures are surrogate pieces of information that do not really demonstrate whether a patient with a given condition actually has had a favorable or unfavorable outcome due to the medical interventions provided. Therefore, to make our reimbursement dependent upon demonstration of completion of these measures seems irrational. The performance and documentation of completion of quality measures becomes an end in itself—and is even being required to prove that your EMR is being used “meaningfully.”

I would love to be able to integrate decision support into my Electronic Record. This has been touted as a large benefit of utilizing technology in the office, improving the safety and quality of medical care because of the potential for clinical decision support. However, besides the mostly annoying and meaningless drug—drug interactions displayed, there is little that the EMR currently provides that has any clinical relevance.

The AAN has produced many evidence-based guidelines, but few (if any) are written in such a way that they can be incorporated into your EMR to help you improve the quality of care you provide. Several neurologists who use one popular EMR have tried to adapt some guidelines to enable the EMR to document the application of best medical practices at the point of care. After reviewing all of the guidelines, we really could not find any that made treatment recommendations that we could incorporate.

The AAN currently has three sets of quality measures: for stroke, epilepsy, and Parkinson’s disease. Stroke guidelines are relevant for hospital care, but not really applicable to the outpatient setting. The measures for epilepsy and Parkinson’s disease contain multiple historical pieces of data but seem to lack a crucial element in determining the quality of medical care. For example, in epilepsy, the measure calls for documentation of seizure type and frequency, as well as querying about medication side effects. If you record the data, you get credit for being a great physician. However, what you do with the information you collected is never assessed. Whether you make an appropriate dose adjustment or medication change is never factored into the assessment of the quality of your care. How many of your patients become seizure free with minimal to no medication side effects is not currently measured. Likewise, despite all of the published epilepsy guidelines, there is none written in such a way as to provide a clinician with treatment recommendations. Somehow, we are missing the point.

Similarly, the Parkinson’s disease quality measures require a great deal of historical data—about possible atypical features, psychiatric problems, sleep disorders, falls, motor complications, and autonomic disturbances—but does not address how these features should be treated. In the treatment of non-motor symptoms of Parkinson’s disease, only three recommendations were made:

  • Sildenifil citrate for erectile dysfunction
  • Polyethylene glycol for constipation
  • Levodopa/Carbidopa for periodic leg movements of sleep

I suspect that many physicians may have been able to anticipate these interventions before the evidence-based guidelines were published.

In fact, the quality measures do not call for a listing of the medical regimen and how it may relate to the patient’s symptoms. When does the patient actually take their medications? No estimation of on/off times is requested. It would seem that knowledge of these critical factors would be necessary to determine the quality of their PD management, but none of these are considered. Like Epilepsy, no decision support is available to be incorporated.

Maybe it is me that is missing the point. Or maybe I am expecting more information than is really available from our current state of medical knowledge. What do you think?