How much health did you make today?
Two articles in the August 12 issue of the New England Journal of Medicine highlight the difficulty in assessing the work product of physicians. The first article, “Quality Measures and the Individual Physician,” describes the difficulties encountered by a primary care physician, Danielle Ofri, MD, PhD, in achieving desirable scores on process measures, such as glycohemoglobin and blood pressure.
The second article, “Accountability Measures—Using Measurement to Promote Quality Improvement” (by Mark Chassin and others) outlines a method for ensuring the validity of measures used by regulatory agencies. Many of us find ourselves in a predicament similar to Dr. Ofri—measured against processes that she is seemingly unable to influence. Incentives are proliferating in hospital medicine and most are based on compliance with various aspects of evidence-based medicine. Does this actually make patients healthier? I do not know.
Our ability to measure health status in an individual patient is crude and primarily reflects the absence of symptomatic disease. High levels of blood pressure, glucose, and cholesterol all portend early death when population statistics are considered, but many individuals blithely ignore these warnings yet manage to achieve threescore and ten without major impairment. Outcomes are what matter to patients and provide the best basis for measurement. The outcome of an intervention against acute symptomatic disease can usually be determined within a short time. The treatment provided by hospitalists can be deemed successful if it allows the patient to be discharged with an acceptable functional status.
This is complicated by the fact that what doctors consider acceptable status may not be acceptable to the patient. More disquiet comes from the fact that about 20% of patients discharged from the hospital today will be back in the ER within 30 days. It is true that we see many patients whose major organs are in a race to the graveyard, and one or another may be producing excessive symptoms on any given day. Still, I am concerned that our ability to assess overall health status for a hospital patient in “stable” condition is crude and needs a breakthrough if we are going to reduce readmission rates.
What about process measures? It is usually a good thing to give heart patients aspirin and add an ACE inhibitor if the ejection fraction is low, but there are some patients for whom these interventions will be harmful. Antibiotics should be given promptly to pneumonia patients, but Chassin notes that this does not help much if the infiltrate is really CHF. Not only is the process of measuring guideline compliance cumbersome (because of exceptions like the ones noted above), but the available evidence-based accounts for only a small portion of the variation in patient outcomes. The Dartmouth Atlas tells us that most of what doctors do, especially in the last two years of life, does not seem to affect outcomes very much.
Where does that leave those of us who must judge the work output of our employed physicians? We must recognize that most of what we measure, including RVUs, core measure compliance, and committee participation, describe the physician as an employee and tell us little about his or her ability as a healer. Achieving discharges with moderate use of resources is a fair basis of comparison. Readmission rate is probably the strongest criterion of healing success that we have for hospitalists, though we understand next to nothing about it. Patient satisfaction is clearly a component of customer value but, remember, that charlatans invariably have great satisfaction scores. Measuring health output will likely remain the holy grail of medical managers for many years to come.
About the Author: Richard Rohr MD, MMM, FACP, started one of the first hospitalist programs in Connecticut in 1996 and is a charter member of the Society of Hospital Medicine. He is a graduate of Columbia University, New York Medical College, and received the Master of Medical Management degree from Tulane University. Dr. Rohr is certified in internal medicine and pulmonary disease. He began his career practicing pulmonary and internal medicine for two staff-model HMOs in MA and CT, and then directed a long-term care facility specializing in medically complex patients before joining Milford Hospital to establish the hospitalist program. Dr. Rohr spent a year as vice president for medical affairs at a small hospital in upstate New York before assuming his current position directing hospitalist programs for the Guthrie Healthcare System in Sayre, PA.
richard scott | Sep 13, 2010 | Reply
thanks. You have voiced well many of my ongoing concerns as we work with credentialing and privileging and have hired Crimson to do focused evaluations. How to set up the appropriate measuring criteria is problematic.