Balancing Patient and Payer Preferences in Medical Decision-Making
Applying maintenance models to medical decision-making in the health care industry can benefit patients as well as insurers.
ANN ARBOR, Mich.—Manufacturing companies traditionally use maintenance models to determine the actions needed to keep production equipment in good running order and to prevent costly system breakdowns. A researcher at the University of Michigan's Ross School of Business suggests a similar maintenance framework might be applied to medical decision-making in the health care industry.
Julie Simmons Ivy, assistant professor of operations and management science at the Ross School, has developed a multi-stage model that can be used by physicians, patients and insurers to determine cost-effective policies for monitoring and treating breast cancer.
Her model considers the uncertainties associated both with limited observation of the disease progression and with various treatment outcomes for lumpectomies and mastectomies. The model uses a dynamic efficient frontier to balance the patient's and payer's conflicting preferences on screening policies.
Today, breast cancer is the most common noncutaneous cancer among American women. In 2003, an estimated 211,300 new cases of invasive disease and 55,700 cases of in-situ disease (not metastasized) were reported, with 39,800 deaths. Mammography, which detects breast cancer at its earliest, most treatable stage, has been shown to significantly reduce mortality.
However, there is disagreement over when and how often to screen women who may or may not be prone to developing breast cancer, Ivy said. Additionally, the cost of annual mammograms must be weighed against the diminished quality of life endured by patients and the cost associated with medical care for a diseased patient borne by insurers, if breast cancer develops and spreads.
"Despite the current recommendations of the American Cancer Society, we may not be screening women at the right time—either too infrequently or too often," Ivy said. "Our model identifies decisions that minimize the total anticipated medical costs while maximizing the effectiveness of screening and treatment over the lifetime of the patient.
"This approach realistically portrays breast-cancer monitoring as a dynamic decision-making process and helps decision makers develop an optimal policy based on beliefs about a patient's perceived condition (e.g., the probability she has non-invasive cancer) rather than simply her age."
To illustrate how her model works, Ivy cites the example of a 30-year-old patient for whom it is possible to perform mammograms every year until she is 80. In this case, the cost-minimizing decision is to pay for a mammogram, unless there is a 50 percent or greater probability that the patient currently does not have breast cancer and no more than a 10 percent probability that she has in-situ breast cancer.
"Our cost-minimizing policies, which recommend screening for women under 40 when there is a belief the patient may have either in-situ or invasive breast cancer, result in savings of 35 percent to 87 percent," Ivy said. "Our conclusion differs from the commonly held belief that screening is never cost minimizing. On the contrary, there are costs of patient care attributable to the incidence of breast cancer and death from the disease. Our analysis suggests the current screening policy can be improved."
For more information, contact:
Phone: (734) 936-1015 or 647-1847