This article is more than 5 years old. Information may no longer be current.
Physician decision making viewed with growing mistrust
Click Here to Manage Email Alerts
There is a growing mistrust of physicians by health policy makers and others who say some of our medical decisions are overly influenced by nonpatient care factors.
During the slow, stepwise progress towards evidenced-based medicine, expect outliers and alternative treatments to be questioned and rewarded differently. We should continue to closely monitor and analyze the design and data interpretation of the new studies that will be used to determine — and restrict — treatment and reimbursement options for our patients and profession.
Patients or profits
|
I will start with an example from outside of orthopedics. It involves a study that looked at the dosage of Epogen (epoetin alfa, Amgen Inc.) used in 134,000 patients treated at dialysis centers in 2004. An article in the Journal of American Medical Association reported that at for-profit dialysis chains, dosages for some patients taking anemia drugs were up to three times higher than at nonprofit centers. The basic issue raised: Do doctors make Epogen dosage decisions on individual patient needs and outcomes, or are they more related to profit incentives? This and similar articles fuel the ongoing debate over the need for more oversight of physician decision-making and behavior.
In reaction, Congressman Fortney “Pete” Stark, chairman of the House Ways and Means Health Subcommittee, is quoted as saying, “This study validates that for-profit dialysis chains are putting profits before patient care.” He called for changing Medicare reimbursement “… to eliminate financial incentives for overdosing.”
Based on this interpretation of the study, health policy may be influenced further in the future, just as we have we have all seen our practices influenced by other Stark-like laws and recommendations in the past.
My interpretation of that particular article was not as analytical as it could be, because the topic is outside my area of expertise. However, let us look at a study within orthopedics — the Spine Patient Outcomes Research Trial (SPORT), a large-scale, 5-year, multimillion-dollar study sponsored by the National Institutes of Health (NIH). This study, and others like it in the future, could have significant repercussions for our specialty.
The SPORT illustration
|
The SPORT study involved 11 U.S. centers and enrolled more than 3,000 patients, with some groups randomized and some nonrandomized. It looked at three major, common spinal conditions: herniated disc, spinal stenosis, and degenerative spondylolithesis. SPORT illustrates some of the potential problems for complex studies designed to compare the efficacy and cost-effectiveness of spine surgery to nonsurgical management of back and leg pain patients.
First, let me acknowledge SPORT for being able to come up with a study design that met some type of consensus among the interested societies. Certain aspects in the statistical design continue to be questioned by surgeons with concerns for bias favoring nonoperative treatment. There were also concerns over the definition of failure and successes in the surgical and nonsurgical goups (ie, some failures in the nonoperative group became surgical candidate-crossover cases). Also, the randomization was felt to be less likely to select the ideal surgical group.
In addition, there are always the groups of patients who will not agree to be randomized (would a busy person like you having significant pain agree to randomize your back treatment?). Whatever the questions and criticisms of the study, everyone lauds the effort to increase our understanding of the cause and treatment of back pain, and recognizes it as an enormous undertaking.
Time warps
The rolling changes of time present problems impacting the patient enrollment in studies like SPORT. Enrollment ended in 2002. Since then newer spinal technologies, already in widespread use, may present a different standard of surgical care. Many less-invasive treatments and some biologic options were unavailable in 2002.
Assessing cost-effectiveness of nonsurgical vs. surgical options for treating low back pain also raised the question of treatment over a lifetime. Does it mean taking medication and subsequent periods of disability vs. a return to full activities for a lifetime?
Bias for lower reimbursement
The potential long-range impact of all these studies is huge. Third-party payers and even Medicare will have a willingness to accept biases in their favor regarding reimbursements. So expect study data to be used to restrict patient access to various treatments and to have reimbursement implications.
This whole field of database and “practice-base disparity” will be increasingly important in medicine. How best do you monitor and measure factors influencing human behavior in the delivery and financing of health care?
Others areas of business are well ahead of us in this area. The casinos track how much individual players put at risk and this determines many of their business decisions. Other sophisticated models work well too — airline rewards programs modify behavior, statistical analysis by large police departments can identify likely times and locations of crimes, and even Netflix has models that track individual and group behavior to aid planning. The real challenge in medicine is developing software for rolling data collection and analysis for physician decision-support.
Changes are coming — with or without us. We need to continue to be part of the data gathering and analysis used for the decisions.
Douglas W. Jackson, MD
Chief Medical Editor