The most likely determining factors for healthcare spending and physicians tendencies to order tests, treatments or referrals appear to be location, location and location, according to a report in the Oct. 24 issue of the Archives of Internal Medicine.
In fact, the report suggests that the geographic location -- and not necessarily the health of their patients -- is a key factor in predicting the level at which primary-care physicians recommend interventions.
Researchers with the VA Outcomes Group in White River Junction, Vt., and the Center for Evaluative Clinical Sciences at Dartmouth, in Hanover, N.H., compared how 5,490 family physicians and internists in different regions answered the Center for Studying Health System Change's 1998-99 Community Tracking Survey. The telephone survey included six vignettes describing different patient scenarios and, not surprisingly, it was found that physicians practicing in regions with high per capita Medicare spending recommended the most interventions.
For example, 32% of the physicians in high-spending regions said they would refer a hypothetical 60-year-old patient with symptoms of an enlarged prostate to a urologist while only 23% of the physicians in low-spending regions said they would do so. In another scenario, a 35-year-old man presented with prolonged back pain and foot weakness. In high-spending regions, this patient was recommended for a magnetic resonance imaging scan 82% of the time. In low-spending regions, an MRI was recommended 69% of the time.
"Physicians respond differently depending on the where they practice, and the higher the spending in the area, the more likely they are to intervene," said the report's lead author, Brenda Sirovich, M.D., a staff physician with the VA Outcomes Group and an assistant professor of medicine at the Dartmouth Medical School. "Healthcare spending does happen to reflect the decisions physicians make on individual interventions."
She said the survey's vignettes were designed so that they would have no obvious right or wrong answers. In fact, there was a general consensus among only one out of six vignettes and that involved referring a 50-year-old man with chest pains to a cardiologist. With the other five scenarios, it was calculated that the high-spending regions recommended interventions 10% more often than the low-spending regions.
"They were intended to be about discretionary decisions," Sirovich said of the vignettes. "Many, many medical decisions lie in that gray zone, and that's also probably where the majority of healthcare spending and utilization decisions lie."
Sirovich and colleagues used the 306 hospital referral regions mapped out by the Dartmouth Atlas Project. Some of the regions in the top tier of spending included: Miami, New York City, Philadelphia and Orange County, Calif. Regions in the lowest tier of spending included: Appleton,
Wis.; Binghamton, N.Y.; Columbus, Ga.; and Rapid City, S.D.
Sirovich said the results of their study were not surprising, but said the strength of the study was that it isolated the role of physicians in explaining the wide variations seen in practice and spending across different regions.
"There is a local culture about how medicine is practiced, but its origin -- how it got to be that way -- is less clear," said Sirovich, explaining that variations could be due to the local medical-legal climate or to the number of specialists in the area who are available for referrals. "In areas of high spending, there tended to be a large supply of specialists, so that probably does influence how often physicians refer."
American Academy of Family Physicians President Larry Fields, M.D., who practices in Ashland, Ky., said the study's results were interesting, but not surprising. Fields noted that primary-care physicians are seeing patients with increasingly complex conditions, so ordering more tests is often prudent. Nevertheless, he added that, in areas with intense legal climates, there will be a tendency to order tests or refer patients to specialists even in cases where these actions may be seen as having a "low yield and high cost."
"Family physicians are seeing patients that have an average of four problems, so it doesn't surprise me that they order more tests -- the less clear-cut something is, the more you'll have to work it up," Fields said. "Also, high-spending areas tend to be high-litigation areas, and there is a tendency to practice defensive medicine in areas that are more litigious."
But Fields also criticized the researchers' use of old data.
"It wouldn't surprise me if things haven't changed," he said. "But (another study) needs to be done with new data to see if the trends are different. You can't take 6- or 7-year-old data and draw conclusions about it today -- no matter what you're measuring."
Sirovich said the 1998-99 survey was the last time the Center for Studying Health System Change used the vignettes in its tracking survey, so more current data is not yet available. Still, she expressed confidence that newer data wouldn't yield significantly different results. CSHSC spokeswoman Alwyn Cassil agreed.
"It's unlikely that it has changed dramatically," Cassil said, adding that the vignettes were dropped from the survey because although they yielded interesting data, it was felt they were taking too much time to complete and they didn't want this to discourage doctors from participating.
She said physicians have multiple motivations for ordering more interventions which are often difficult to sort out.
"It's very difficult to unsnarl because physicians have a financial benefit -- doing more stuff results in physicians being paid more -- but there is also defensive medicine," Cassil said. "There is also regional culture and the standard of care in a particular community drives some of this, but it's a chicken-and-egg thing."
She also agreed with Sirovich that local supply may be a driver of this behavior.
"Folks in Miami get more because they can," Cassil said. "If you build it, it will get used."
View the study abstract.