Medicare patients with heart disease in Buffalo are one-third as likely as patients in Brooklyn to receive procedures to unclog arteries.
But degenerative arthritis patients in Buffalo are twice as likely as those in the Bronx to undergo hip replacements.
The large differences in procedures reflect the many variations in care here and across the country in how patients get treated for similar conditions.
And whether a patient gets surgery or nonsurgical treatment often depends on physician recommendations, perhaps neglecting patient preferences, according to a new report by the Dartmouth Atlas of Health Care.
“The variations show that where you live is what you get,” said Shannon Brownlee, lead author of the report, which for the first time offers a detailed comparison of elective procedures in Buffalo and other cities.
The results also indicate that the medical community should embrace shared decision-making in treatment decisions, especially for elective procedures and tests, to ensure that patients get the care that is right for them, the authors said.
Patients should ask questions about their treatment choices instead of routinely delegating decisions to their caregiver in the belief that “the doctor knows best,” the authors said.
“It’s our hope to better understand and highlight the major cause,” Brownlee said of the variations in treatment. “That is that patients’ preferences often are not taken into account, and that physicians vary widely in their opinion about the best treatment for particular conditions.”
The report is based on rates of surgery per 1,000 patients with Medicare, the federal health program for individuals 65 and older, for the years 2008 to 2010.
The Dartmouth Atlas has been an ongoing project for many years at Dartmouth College’s Institute for Health Policy and Clinical Practice to document variations in medical care and their links to cost and quality.
The group, which is funded primarily by the Robert Wood Johnson Foundation, released four reports focused on different regions of the country.
In the mid-Atlantic states, for instance, if you have back problems in Lancaster, Pa., you are more than four times likely to get surgery than if you live in Brooklyn.
If you develop breast cancer in Trenton, N.J., you are three times more likely to have a mastectomy than in Syracuse.
And, compared with Buffalo, prostate cancer patients in Rochester are five times more likely to receive an operation to remove the prostate gland.
The greatest variation was seen in the use of prostate-specific antigen testing, or PSA, to screen for prostate cancer. Among the 30 largest communities in the mid-Atlantic states, rates of PSA testing ranged from 6 percent of male Medicare beneficiaries in Binghamton to 60 percent in Manhasset. Buffalo’s rate was 35 percent, near the national average.
“We know that the variation we see is little explained by differences in population,” said Dr. David Goodman, co-author of the study, which adjusted for the sex, age and race of patients. “It’s largely the result of differences in the way doctors practice.”
It’s not clear what the right rates are, Goodman said. However, sharp differences influence the quality and quantity of the health care patients receive, he said.
What accounts for the variation?
Among other things, there isn’t agreement on how best to treat some medical conditions. Often, it’s not clear, based on current medical knowledge, which treatments are the most effective for a particular patient.
Back surgery, for instance, is an example of the lack of consensus over how to care for back pain. With no agreement, the rate of back surgery greatly varies from city to city, depending on physician preferences, the report concluded.
About 3 percent of patients received back surgery in Buffalo. But the rate ranged from 1.6 percent in Brooklyn to 7 percent in Lancaster, Pa. The national average is nearly 5 percent.
The PSA test offers another example.
Doctors use the PSA test to screen for prostate cancer. But prostate cancer is often slow-growing, and the test can’t tell which tumors will cause a problem, leading many patients to undergo unnecessary biopsies and therapies.
Some medical authorities argue the test does more harm than good because of the risks of treatment, including incontinence. The U.S. Preventive Services Task Force – independent experts authorized by Congress to make treatment recommendations – this year recommended that PSA screening should no longer be done routinely in men with no signs of illness. The task force advised doctors and patients to make testing and treatment decisions based on a full discussion of the benefits and risks of the options.
Other medical experts counter that the PSA may be imperfect, but that it saves lives.
The result is confusion. Patients struggle with conflicting advice or may receive the test without a chance to weigh the options.
“The lesson of the PSA test is there isn’t as much agreed upon in medicine as you would think. Often, the right care requires a nuanced understanding of the options for each individual patient,” said Dr. Brian D’Arcy, senior vice president of medical affairs for the Catholic Health hospital system.
D’Arcy points to other causes of variation, including differences in access to care, such as in smaller cities or rural areas. In addition, physicians often have legitimate differences of opinion based on their training, experiences and the practices of their peers in the community, he said.
The issue is more complicated than simply choosing which treatments are best and standardizing that care, according to experts.
Medicine also is faced with the challenge of standardizing how procedures are done, said D’Arcy. For instance, he noted the importance of instituting agreed-upon safety checklists for surgery and protocols for choosing the most appropriate patients for such procedures as carotid endarterectomies.
The rate for carotid endarterectomy – a procedure to clear deposits from the carotid artery to prevent stroke – ranged from 0.8 per 1,000 patients in Manhattan to 3.2 per 1,000 in Utica, with Buffalo at 1 per 1,000.
Variation data highlighted by the Dartmouth Atlas is useful for identifying the outliers, according to Dr. Irene Snow, medical director of Buffalo Medical Group, a large multispecialty physician group.
“If you are at one extreme or the other, you want to drill down and find out why, and see if it’s an actionable item,” she said.
More than anything, she and D’Arcy said the documented variation in care calls for the continued movement in health care toward arrangements with physicians that stress value and quality of care, over volume, and greater attention on informed decision-making.
“Historically, medical decisions have been about the preference of the physician,” she said. “We’re moving to shared decision-making, and it shows that there is value to having a second opinion.”
Care varies greatly by location
The treatment you receive often depends on where you live
Percent of patients who receive the PSA test to screen for prostate cancer
U.S. average: 34.5%
Procedures to unclog heart arteries per 1,000 patients
Toms River, N.J.: 11.8
U.S. average: 8.2
Hip replacements per 1,000 patients
U.S. average: 3.8
Source: Dartmouth Atlas
PSA percentages from 2010; heart procedure and hip replacement rates from 2008-2010. Rates adjusted for age, sex and race of Medicare beneficiaries.