BY EMILY BAZAR California HealthCare Foundation Center for Health Reporting
Residents in a wide swath of the Central Valley, from Fresno to Bakersfield, undergo two common heart procedures at rates substantially higher than in other regions of the state, according to a new study.
Corcoran residents, for example, are five times more likely than the average Californian to have an elective angiogram, which helps doctors find clogged arteries. Residents of Lindsay are four times more likely to have a procedure called elective PCI, used to open blocked arteries supplying blood to the heart.
ABOUT THIS PROJECT
This project is a partnership between The Californian and the California HealthCare Foundation Center for Health Reporting, a news organization that reports on health care issues that concern Californians. Emily Bazar is a senior writer at the center, which is based at the USC Annenberg School for Communication and Journalism. The center is funded by the California HealthCare Foundation, which aims to improve health care quality, increase efficiency and lower costs.
The study of medical procedures and geographical variation was conducted by Dr. Laurence Baker of Stanford University and was also funded by the California HealthCare Foundation. The work was commissioned by the Campaign for Effective Patient Care, an interest group that promoted the involvement of patients in making medical decisions. Formed during the health reform debate, the group recently disbanded.
The center is editorially independent of the foundation and had no role in the study.
Stanford professor Laurence Baker analyzed five years of statewide hospital (and some outpatient) discharge data from the Office of Statewide Health Planning and Development. He focused on three elective heart procedures: angioplasty (also called PCI), angiography and Coronary Artery Bypass Graft surgery.
The data, representing 2005 through 2009, are tied to patient zip codes and reflect where the patients lived, not where they received care. Data from federal facilities, such as VA hospitals, are not included.
Baker broke the data into 24 established medical geographies known as Hospital Referral Regions. HRRs are made up of smaller regions called Hospital Service Areas, which also were analyzed.
To control for demographic and health differences among the regions, Baker risk-adjusted his results for age, sex, race, education, income and health insurance status. In the case of the three heart procedures, he also adjusted for rates of hospitalizations for heart attack and when the patient had a diabetes diagnosis.
To define procedures as elective, he relied on the Dartmouth Atlas Project, which specializes in research on variations in Medicare usage, and in the case of the heart procedures, he used a published model. Take angioplasty, for instance. Baker excluded procedures that were done when someone had a concurrent coronary emergency.
He organized his findings into three categories: Rates for people under 65, over 65 and all.
The state rate is the baseline, and represents the average rate at which procedures are performed statewide. The state rate, however, is not necessarily the correct rate. It just provides a benchmark.
Baker's methodology and results were reviewed by an advisory board made up of 26 physicians and researchers, including Jack Lewin, CEO of the American College of Cardiology, and Rita Redberg, a UCSF cardiologist and editor of the Archives of Internal Medicine.
The art of applying risk adjustments is an inexact science, and there is debate among scholars on the best formula to use.
John Spertus, a cardiologist and professor at the University of Missouri, Kansas City, who conducts similar research, faults Baker for not accounting for patients' level of angina, the chest pain or pressure that is one of the main reasons patients seek PCI.
"What he's missing that's really critical is how bad these patients' angina was," he said. "Some patients have angina despite taking medications, and for those patients, angioplasty can improve their quality of life. The more severe the symptoms, the more legitimate the procedures."
In Bakersfield, many healthcare workers pointed to the area's air pollution and the potential adverse effects on heart health. They also questioned why obesity and smoking weren't taken into consideration.
Baker countered that he accounted for the health of the population by adjusting for heart attacks and diabetes diagnoses. The demographic factors he used for his adjustments, such as income and education, also correlate with health-related factors such as smoking and obesity, he said.
He cautioned against using the health of a population as an explanation, because there are no places in California where the health of the population is four or five times worse than any other place, he said.
John Wennberg, a professor at Dartmouth Medical School and founder of The Dartmouth Atlas of Health Care, said the argument that the differences could be explained "by some environmental or other factor just runs counter to all the history of these massive studies done on procedure variations. It's just not possible."
Spertus, too, acknowledged that even though Baker didn't account for angina, "the variation is never that large" as it is in some areas of the state. "Wennberg is right."
-- Emily Bazar
In fact, seven of the 10 regions where California residents post the highest rates for PCI are between Dinuba and Lake Isabella. The region has eight of the 10 highest rates for angiograms.
These results are part of a new analysis of statewide hospital discharge data by Stanford health research and policy professor Laurence Baker, which shows that the use of certain elective medical procedures varies wildly with geography.
Some health experts say these differences demand action.
"We have to work on reducing variation. We're in a country that's accumulating debt so fast, and most of it relates to rising health care costs," said Jack Lewin, CEO of the American College of Cardiology. "We have the opportunity to save money, reduce the risk of complications and actually achieve better results."
The new study also carries an important message for health consumers: Where you live in the Golden State can affect how likely you are to go under the knife.
"Depending on where people with chronic illnesses live, and which hospital or doctor they are loyal to, they receive very different levels of care," said John Wennberg, a professor at Dartmouth Medical School and founder of The Dartmouth Atlas of Health Care, which also examines patterns of medical use in the U.S.
There is heated debate about why these differences occur.
Health providers usually pin variation on the condition of their local population, saying their patients are unique because they're sicker, more obese, poorer or otherwise more vulnerable than people in other areas.
But longstanding research disputes this explanation and suggests that other powerful factors are at play, in particular differences in how doctors treat diseases.
"The more fundamental thing that drives variation is the nature and culture of the practice," said Eric Hammelman, a vice president at Avalere Health, a health care consulting firm in Washington, D.C. "You just have a group of physicians that tend to order more angiograms or PCIs. That's how they think and do things. They've never been told not to. They've never been told they're the outliers."
The financial and health implications of extreme variation are enormous, raising the prospect that billions of dollars are wasted each year on unnecessary and potentially dangerous treatments. About 600,000 PCIs alone are performed nationwide annually at a price tag of more than $12 billion, according to a recent study in the Journal of the American Medical Association.
Medicare and private insurance companies are beginning to act. They are using data to search for disparities in treatment patterns and in some cases, pressuring doctors and hospitals to change.
Baker's study, based on an analysis of hospital and some outpatient discharge data from the Office of Statewide Health Planning and Development (OSHPD), showed high and low use of several elective surgeries around the state between 2005 and 2009.
To get a more incisive look at how California regions compared, Baker adjusted the rates to try to eliminate differences in the health of the populations and their access to health care. He adjusted each region's results for age, sex, race, education, income and health insurance status.
Baker and other researchers say that these demographic factors are good predictors of population health. For the heart procedures, he also used two health indicators: rates of hospitalizations for heart attack and diabetes diagnoses during hospital stays.
The greatest disparities on the high side occurred with PCI and angiography, which are heart catheterization procedures. Angiography is a diagnostic test to find blockages in arteries. PCI is used to open the arteries, usually with stents.
PCI generally is considered elective when someone is not having a heart attack or other cardiac emergency. That means patients with stable coronary artery disease often have a choice: They can have PCI, or treat the disease with medication and lifestyle changes, which studies show yields comparable health results.
The valley isn't the only California region where rates for elective heart procedures are high. Residents in the poverty-laced city of Clearlake in Northern California have by far the state's highest rates of elective angiograms and PCI. Other high rates appear in counties such as San Diego, Stanislaus and Monterey.
But in no part of the state are the high rates as widespread as in the Central Valley.
In his study, Baker examined 208 small areas of California and then grouped them into 24 larger regions. Focusing on these larger areas, Baker found that the Bakersfield and Fresno areas had the highest rates of angiograms, at 2.5 times the state rate, and that Bakersfield had the highest rate of PCI, at more than two times the state rate. The Fresno region was tied for second on the frequency of PCI procedures, at 1.5 times the norm.
Over the years, health care experts have offered many hypotheses for the causes of variation, including:
* Financial incentives. When health care providers get paid by procedure, they earn more as they do more.
"In general, we have a fee-for-service system that reimburses procedures generously and it ... reimburses inappropriate procedures just as generously as appropriate ones," said Rita Redberg, a UCSF cardiologist and editor of the Archives of Internal Medicine.
* Patient demand. "There are places where people actually want PCI if they have coronary artery disease, and they think they're being denied effective care if they don't get it," Lewin said.
"The sad truth is that 50 percent of people with high coronary risk don't take their life-saving medications, and some doctors believe the use of PCI is a safer option in many patients," Lewin said.
But researchers who specialize in variation keep coming back to another factor. They point to differences in how physicians practice in different communities, especially when they have discretion in how to treat illnesses.
Doctors' training and styles often lead them to choose their favorite treatments. That, over time, can create differences in the number of procedures performed across regions, said Harlan Krumholz, cardiology professor at the Yale School of Medicine and director of the Yale-New Haven Hospital Center for Outcomes Research and Evaluation.
"We know that risk factors for heart disease vary across the country, but with five- and six-fold variation, it's just implausible that the differences in the population explain the differences in rates," he said. "The research indicates that the most likely reason for variation is the local clinical culture. In high-use regions, they are just more likely to intervene."
Government agencies, health insurers and the medical community are ramping up efforts to target and limit variation, both to reduce costs and the incidence of unnecessary procedures.
This summer, Medicare debuted a new program to detect fraud through technology and data mining. It will look for patterns that might indicate fraud, including extreme variation.
Health insurers, too, increasingly are seeking innovative ways to address outliers. Blue Shield has teamed up with some Sacramento and San Francisco hospitals and physicians to coordinate care for some of their HMO patients to reduce costs and improve care. One of their strategies will be to identify and reduce variation.
At the ground level, there's a growing push for what's called shared decision making. In this process, patients take a more active role in determining their own treatment. Doctors lay out options, provide balanced information about various techniques -- including potential complications -- and both sides collaborate to make a decision.
"Well-informed patients may not be as eager to undergo procedures. They might choose to do less rather than more," Krumholz said. "People often just go along without truly being informed, even though they bear the consequences of the choice."