For more than 20 years, the Dartmouth Atlas Project (www.dartmouthatlas.org) has focused on regional variations in the quality of health care services delivered across the United States. This research has emphasized that too often supply drives demand. Patients receive surgeries, tests, and procedures they may not need and that sometimes cause harm, which drives costs up and quality down. While the results are eye opening, few who provide and pay for health care appear to have acted on this information for younger, commercial populations. Questioning treatment choices or the use of specific services is tricky.
The purchasers, payers and providers who are part of the Puget Sound Health Alliance, an Aligning Forces for Quality (AF4Q) community, embraced the challenge, recognizing that variation can be symptomatic of inconsistent, wasteful, and potentially inappropriate care.
After several years of issuing “Community Checkup” reports that compared provisions
of generally recommended “effective care” across the Puget Sound region, the Alliance set its sights on developing measures of value in health care services. The Alliance highlighted two important steps on the path to assessing value: (1) gauging the appropriateness of treatment choices and (2) comparing the composition of services providers deliver for similar treatments. The Alliance conducted the following two separate analyses.
First, the Alliance examined geographic variation in treatment choice for “preference sensitive” conditions—conditions with a variety of treatment options, each with different risk benefit tradeoffs, from which patient and doctor choose. The Alliance examined how frequently the more aggressive surgical options (see sidebar) occurred in the population, asking the question: “Does where patients live influence what treatments they receive?” Significant variation in the rates of back surgery, cardiac procedures, and hysterectomies emerged for distinct residential areas. The results potentially could improve the appropriateness of treatment selection and target deployment of shared decision-making aids.
The Alliance and other AF4Q communities are encouraging providers and patients to engage in shared decision making to choose treatments that align with patient values and treatment goals. Research shows that when patients are fully informed of all options available for their conditions, they tend to choose less aggressive options. These options tend to be less expensive, involve fewer clinical risks, and may be associated with faster recovery.
The second analysis measured service intensity as reflected by differences in the composition of similar treatments (e.g., the number of tests, the length of hospitalization, the types of physician services, etc.). It captured how the quantity and type of services for a particular treatment vary among delivery systems. The Alliance chose to focus on the most common hospitalizations but included the activities of all health professionals who were active during the patient’s hospital stay—a perspective suggestive of accountable care organizations.
As writer Atul Gawande, MD, surgeon at Brigham and Women’s Hospital in Boston, associate professor, Harvard School of Public Health, and associate director, Center for Surgery and Public Health, pointed out, performance reporting and changes in payment are radically altering the relationship between hospitals and doctors. “They offered us space and facilities, but what we tenants did behind closed doors was our business. Now it’s their business, too,” wrote Dr. Gawande in the August 16 New Yorker article “Big Med.”
Service intensity is an important lens through which to view health care variation—research suggests local practice culture and medical discretion may account for this variation among providers. Estimates place as much as 30 percent of all health care spending in the potentially unnecessary category.
The Alliance drew results from its large multi-payer claims database supplied by health plans, self-insured employers, and union trusts. The database contains all hospital encounters, office visits, procedures, tests, and other services for which insurance pays. While there are no transaction prices in the database, the service intensity data provide a directional signal about potential drivers for the cost of care.
Project consultant James Andrianos of Calculated Risk, Inc., used an analogy to distinguish between price and service intensity: “Think about an entrée at a restaurant instead of a hospitalization. The menu price would be analogous to the fee-for-service transaction price paid by the health plan. Service intensity, on the other hand, is more akin to the calories in the entrée. If one restaurant offers this entrée with significantly more or less caloric content than we typically see, we may wish to look at the ingredients they are using to understand what is driving the difference. Therefore, service intensity is not about who is most or least expensive. It is concerned with variation in the ‘recipe’ of care.”
The Alliance’s service intensity results have practical implications for the region: All things being equal, delivery systems with more consistent, lower-intensity service patterns will appeal both to employers and to hospitals and medical groups.
Employers could use the results to identify treatments and delivery systems where potentially unnecessary “discretionary services” are less likely to occur. Providers could use the data to identify sources of variation that could threaten their ability to succeed with payment reforms, such as bundled payment.
Hospital CFOs have been quick to express interest in what the comparisons say about their production activities, particularly within multi-hospital systems. For their part, employers recognize service intensity variation as a key component in beginning to identify higher-performing delivery systems.
These two analyses are an important step forward for the Alliance community. They join the process quality measures that also are emphasized in the Alliance’s Community Checkup report. As the Community Checkup expands to encompass patient experience, cost, and outcomes, it will begin assembling a more complete picture of health care value in the region. With each step, the Alliance moves closer to its overarching goal of physicians, other providers, and hospitals in the region achieving the top 10 percent in performance nationally in delivery of quality, evidence based care, and reduction of unwarranted variation, resulting in a significant reduction in medical cost trend.
Surgeries associated with conditions having multiple treatment options:
Heart valve replacement
Surgery to clear neck artery
Leg artery restoration
Abdominal artery repair
Hip and knee replacement
Gall bladder removal