Significant variations in costs generated by surgeons under controlled circumstances

Surgeons incurred a wide range of hospital costs even when their procedures were controlled for complexity and patient risk, according to one of the first studies of its kind to track the cost performance of individual surgeons.

According to findings from the study, which was published today in the April issue of the Journal of the American College of Surgeons, surgeons differed significantly in overall costs, and the dollars-and-cents cost differences were substantial. For example, one surgeon's costs were 45 percent lower than those of an arbitrarily selected reference surgeon, and the costs generated by six surgeons were at least 39 percent lower. These differences were highly statistically significant, according to Bruce L. Hall, MD, PhD, FACS, assistant professor of surgery, Washington University, St, Louis, MO, and the lead author of the study. Some surgeons also had highly volatile cost profiles, Dr. Hall said. The estimated ranges of costs for these surgeons varied extensively, even for surgeons who saw similar mixes of patients.

Unlike other studies of surgical costs, which have analyzed the cost performance of all surgeons in a hospital as a whole, this study focused strictly on the performance of individual surgeons. The study also used a statistical model to eliminate fluctuations in cost due to the severity of illness of patients or the nature of the surgical procedure. In essence, the study gave surgeons an identical set of patients and tested how much it would cost for the surgeons to perform an identical set of surgical procedures.

"Surgeons have been provided with information about their total costs, but rarely are those adjusted for the sickness of patients, or the complexity of the operation. I don't know of any instance apart from this publication where surgeons were provided with their risk-adjusted and procedure-adjusted estimates of costs. This process enables surgeons to examine their estimated costs and determine how they utilize resources and whether they do things efficiently," Dr. Hall said.

The costs calculated in the study were not solely related to patients' outcomes. "People may feel that a higher cost surgeon gets better results. This study is standardized by surgical procedure and patient. So the question is not, does a more expensive surgeon have better results? We're asking whether surgeons differ in their costs even when they are performing the same kinds of procedures on the same kinds of patients," Dr. Hall explained.

A corollary question is whether efforts to reduce the variability of surgical costs may improve the use of resources. "We may be able to generate cost efficiencies without sacrificing the quality of care in this area and use the monies saved for other things, such as prenatal care or the provision of care to the uninsured or the poor," according to Barton Hamilton, PhD, Robert Brookings Smith Distinguished Professor of Entrepreneurship at the John M. Olin School of Business of Washington University, St. Louis, MO, a coauthor of the study.

Such efforts also may help increase the efficiency of delivering hospital care. "Our finding that there are differences not only in the average cost across surgeons but that there is a lot of variation in terms of volatility suggests a management challenge," Dr. Hamilton said. "From a health system perspective, you would like processes and procedures to be fairly reproducible, to have predictable ranges of variance. We're showing that there is quite a bit of variability in performance, which can make planning, budgeting, and other management-related processes that much more difficult," Dr. Hamilton added.

The study represents a unique collaboration among the departments of surgery at the University of Michigan, Ann Arbor, MI, and Washington University, St. Louis, MO; the John M. Olin School of Business at Washington University; and the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP), Chicago, IL. The ACS NSQIP is a nationwide effort provided by the American College of Surgeons to private sector hospitals that wish to assess the quality of care delivered to patients who have undergone general or vascular surgery. The program was originally supported by a grant from the Agency for Healthcare Research and Quality (AHRQ) until September 30, 2004, with subsequent funding and management provided by ACS NSQIP.

The researchers used raw data that had been gathered through the ACS NSQIP on patients who had undergone general or vascular surgery at the University of Michigan between 2003 and 2004. Data on 785 patients who were treated by one of 29 University of Michigan surgeons were analyzed by members of the department of surgery and school of business at Washington University, who applied several statistical algorithms. To control for variations in cost that may be related to treating more seriously ill patients, the researchers included a statistical algorithm intended to create a "typical patient." The algorithm adjusted for the degree of sickness or acuity of patients on the basis of their age, gender, race, number of comorbid conditions, need for a ventilator, anesthesia risk, type of wound, and the priority of the case.

To control for differences in costs associated with performing more extensive or complicated procedures, the investigators used an additional algorithm that reflected a "standardized" surgical procedure. This algorithm adjusted for the complexity of the surgical procedure based on the number of relative value units of work involved. Relative value units comprise a method of cost accounting that takes into consideration relative differences in the consumption of resources.

The next step for the researchers is to learn why costs differ among surgeons. According to Dr. Hall, "Here is a process that is showing variability, but it's not clear why. If we can target this process and examine it more closely, we believe we will be able to improve the quality of the outcome. We want to learn whether there are differences here that do not need to exist, and if they do not need to exist, how can we improve the efficiency of our system?"

Other participants in the study on variations of surgical costs included Darrell A. Campbell Jr., MD, FACS, professor of surgery, and Laurel R. S. Phillips, RN, MSN, both from the department of surgery, University of Michigan Hospitals and Health Centers, Ann Arbor, MI.