Connections : Join the Conversation - The blog of Biomet CEO Jeffrey R. Binder

President Obama and Congress have consistently expressed three key goals for the U.S. healthcare reform legislation recently signed into law: expand access, reduce cost per beneficiary, and improve quality.¹ Of these, improving quality is the most problematic, simply because a universally accepted definition of healthcare quality is an elusive ideal.

At best, policymakers (and even groups of doctors) can attempt to define quality standards for the average patient within a defined subgroup, and for patients with average expectations and needs. However, patients are real people, not defined by some theoretical norm.

The notion that the government can define quality healthcare for any given individual is at odds with the concept of personalized care. Quality is defined by each patient's condition and expectations. For example, a frail patient may simply expect to live independently, while an active patient may expect to return to a physically demanding job. Given the wide variety of patient expectations and conditions, it is inconceivable that the government can define quality with sufficient precision for all, or even most, patients.

Current Quality Programs

Nonetheless, the government has initiated two key voluntary quality reporting programs. The hospital program, Reporting Hospital Quality Data for Payment Updates (RHQDAPU) is a voluntary program. The program provides bonuses for following and reporting the processes specified by the government. CMS financed these bonuses by reducing hospital reimbursement at the program's outset.²

The Physicians Quality Reporting Initiative (PQRI) is also voluntary, does not reduce pay directly (although proposed legislation will do so), and provides a 2% bonus for following and reporting on certain practice guidelines.^3,4

Here's the problem with using processes and guidelines as proxies for quality. If a process or guideline is widely accepted, it is unlikely that any but negligent providers are not following it.⁵ Thus, programs that link payment to adherence to such processes do not represent a significant opportunity for quality improvement.

Indeed, hospital adherence to the RHQDAPU program is so high that, by current definitions, only minimal improvement is possible. In 2009, 99% of hospitals participated in the hospital quality reporting program, and 97% received compliance bonuses.⁶

PQRI has the opposite problem. Of the 1 million eligible Medicare practitoners⁷ only 15% participated in PQRI.⁸ Of those who participated, slightly more than half qualified for bonuses.⁸

A survey by the Medical Group Management Association sheds some light on the low level of participation. In 2008, fewer than half of participating physicians were able to access their feedback reports from the PQRI program. Of those who were able to access feedback, 60% were either dissatisfied or very dissatisfied with the information. Only 11% were satisfied with the program's effectiveness in providing quality guidance.⁹

As one practice manager stated, "PQRI reporting has no impact on the quality of care provided. It merely adds an administrative requirement to document clinical best practices already part of our routine standard of care. As a consequence, it is a hindrance rather than something that helps."⁹ (AMed News, 3/1/10).

Process-based Quality Measures: Any correlation to improved outcomes?

This is not surprising. It has been shown that payer-defined quality measures do not necessarily correspond with improved outcomes.^10-12 Worse, research has also demonstrated that adherence to practice guidelines can be detrimental to outcomes in some patients.¹³

Both RHQDAPU and PQRI offer limited opportunities for quality improvement; RHQDAPU, because participation is so high that little improvement is possible. PQRI, because compliance is low and the data provided is not perceived as helpful. Yet the reform legislation offers more of the same, ramping down provider payments and requiring reporting on scorecards in order to recoup the lost revenue.

Focus on data, not treatment mandates: the potential value of comparative effectiveness research

In order to encourage quality without restraining access, government should eliminate quality programs that tie reimbursement to centrally-defined processes. Guidelines cannot keep up with changing practices and technologies; nor can they determine the best course of treatment for each individual patient.

Instead, government should focus on encouraging the generation of better information on outcomes, and allow doctors to use it to improve care. Toward this end, comparative effectiveness research may be a promising tool.

The healthcare legislation creates a comparative effectiveness research council.⁴ The idea is to evaluate the relative effectiveness of various treatments, providing physicians with data to support clinical decision-making. Despite language that prohibits the council from making coverage recommendations, nothing prevents CMS from using the data for that purpose.⁴ And while we do not support the use of comparative effectiveness of data to measure cost-effectiveness or to make coverage decisions, we must accept that there is a certain inevitability to this result in today's environment.

Well-designed research on the effectiveness of various treatments is a good idea, and basing reimbursement on effectiveness only makes sense. Medicare certainly should pay more for the most effective treatments. However, all payors, including the government, must recognize the limits of comparative effectiveness research and not stretch the data beyond its utility.

Again, any research that is not limited to very discrete subgroups has limited applicability to any given patient. Physicians and patients should always retain the freedom to choose a course of treatment, even if the research indicates that it is not, on average, as effective as an alternative. There is no such thing as an average patient. Providers should not be expected to incur a loss under these circumstances.

For example, let's say research indicates that, in 70 year old osteoarthritic patients, ceramic-on-ceramic hip replacements are not cost-effective compared to conventional metal-on-polyethylene hip replacements. As a result, Medicare adjusts its payment schedule to cover the cost of the less expensive polyethylene option. In some cases, the patient and doctor may determine that ceramic is the best option. The hospital should have the ability to bill the patient for the difference between the cost of the more expensive procedure and Medicare reimbursement.

For certain technologies and techniques, where results may vary by doctor, or do not yet have a long clinical history, Medicare (and other payers) may very well decide to not cover the treatment in full, paying instead for a less advanced, but more predictable treatment. This is a reasonable position; however, if the doctor and patient decide that the more controversial treatment is the right choice, Medicare should not block access, and allow the patient to pay the difference in reimbursement.

In this fashion, comparative effectiveness research can work hand-in-hand with reforms that allow doctors and hospitals to bill patients beyond Medicare allowable amounts. Such an arrangement will allow for technological advancements and avoid centralized rationing, while preserving personalized patient care and controlling Medicare's costs.

Conclusion: Quality is measured one patient at a time

Payers cannot centrally define quality and value.. Varying patient expectations and conditions demand the ongoing availability of personalized care. Effective policies will preserve the ability of patients and physicians to collaborate on treatment decisions, and provide access to data on the effectiveness of treatments. Payers are certainly within their rights to determine payment based on average effectiveness; however, payment policies should not prevent patients and physicians from choosing the optimal solution every time.