While significant progress has been made in improving cancer survival rates—as evidenced by a drop in cancer deaths in the U.S.—the number of new cases continues to increase. Experts say far too many patients receive inferior care. Mistakes in care can be fatal, and yet some people do not receive enough treatment, while others receive too much or the wrong kind.
“It’s quite surprising, but the quality of cancer care in America varies dramatically,” Stephen B. Edge, MD, chairman of surgery at the Roswell Park Cancer Institute in Buffalo told The New York Times in a 2007 article. “It’s scary how much variation there is.”
Government and medical groups acknowledge that the quality of care is uneven. In 1999, a report by the Institute of Medicine said, “For many Americans with cancer, there is a wide gulf between what could be construed as the ideal and the reality of their experience with cancer care.” The institute noted that there was no national system to provide consistent quality. Rates of concordance with clinical guidelines vary across different types of cancer care, patients and institutions. The New York Times, in that same article from 2007 said, “Cancer patients lose in a maze of uneven care.” Decisions for consumers can be agonizing, in part, because the quality of cancer care varies among doctors and hospitals.
The foundation for this journey is a platform of combined administrative claims and clinical data, allowing for sophisticated analytics that measure cost, outcomes, and guideline adherence.
It is impossible to improve what cannot be measured or to measure what hasn’t been defined. Take, for example, the topic of healthcare quality. Everyone wants quality, but everyone’s keeping score differently.
A challenge for payers
Payers are challenged with both the financial burden of cancer, and ensuring that the members in their network are receiving the highest quality care. Cancer quality measures and guidelines, such as those endorsed through the National Comprehensive Cancer Network (NCCN) are more complex, and require the addition of clinical data for accurate reporting. Combined clinical and administrative data analytics will facilitate improvements in the care received by cancer patients and its cost efficiency at a level of granularity and precision not currently available with administrative data alone.
Payers are looking for answers to the following questions:
- Are members in my network receiving the highest quality care?
- Which providers in my network are adhering to clinical treatment guidelines and to what degree?
- How is my plan performing compared to other plans when it comes to providing quality care?
- How can I identify areas of wide variation and prioritize messages to my members and providers?
- How can I adjust my benefit design or payment strategy to encourage quality care?
Benchmarking drives impact
According to the American Productivity & Quality Center, “. . . benchmarking (is) . . . ‘the process of identifying, understanding, and adapting outstanding practices and processes and processes from organizations anywhere in the world to help your organization improve its performance.”
Benchmarking can be used for many purposes, including healthcare coordination and delivery, medical management programs, information sharing programs, and quality improvement. Of course, the main purpose is comparison: application of data for the delivery of care for better outcomes.
Benchmarking drives impact in the areas of quality and cost of care. Quality is defined as enhanced clinical data, such as disease state for comparison of actual treatment to the NCCN guidelines (the recognized standard in oncology care). Cost of care is evaluated using episode of care units, risk assessment technologies, and evidence-based rules regarding cost-effectiveness and benchmarks from a national database. The benefit is education, consumer and provider, guideline concordance and cost of care analysis including drug therapies and treatment gaps.
Benchmarking typically begins with administrative (claims) data. Administrative data alone provides a limited view of quality of care but has become the standard. This data is used as a proxy for performance measures and provides only a partial view. This data, however, cannot be categorized to what matters in oncology, such as stage, tumor status, node status, and metastasis status which include the initial and treatment status.
Claims data analysis issues:
- Data reasonableness
- Comprehensive
- Data quality
- Incurred VS. paid
- Incomplete data – physician group
- Prospective VS retrospective
There is agreement that the combination of administrative and clinical data is a better solution, but the optimal solution also includes comparison to the clinical (NCCN) guidelines. This concept, administrative data, plus clinical data compared to guidelines can be utilized in care management, disease management, and pharmaceutical use. Clinical data allows a more robust view and comparison to treatment protocols.
For example, with administrative data only, we know that a breast cancer patient has had a physician visit and an annual mammogram. With clinical data, hormone therapy can be identified. The fact that this treatment is not recommended can only be identified by comparison to the NCCN clinical guidelines. This is an example of care management.
This example also applies as compared to drug use that is or is not part of the NCCN drug compendium. Several payers are only reimbursing physicians for those drugs on the drug compendium.
The United Health Group (UHG) has implemented a new program to collect oncology data for colon, breast, rectal and lung cancer. The clinical data collected is allowing UHG to accurately measure and assess the type of care the patient is receiving. Only information necessary to classify patients into clinically similar groups for assessing the care provided against evidence-based quality and efficiency is being collected. One result of this program has already addressed Herceptin cost, and appropriate usage. The data illustrated that 12% of Herceptin users were not been tested for HER2 status or had an under-expressed HER2 status. Utilizing this information, a new program was implemented requiring over-expressed HER2 status test result submission prior to initial Herceptin claim payment. This program has generated a reduction of in-appropriate Herceptin claims, with an annual savings of $10 million realized. For one case, the cost of treatment can be as much as $80,000 annually. This is for the fully insured HMO membership of UHC.
UHG also expects $10 million in annual savings in oncology spend for drugs ordered and not on the NCCN drug compendium. All oncology related pharmacy claims are evaluated against the NCCN drug compendium which is utilized as the standard.
Another example is the use of Erythropoietin (EPO) which was being prescribed for patients whose blood cell level did not warrant use of the drug. The program was implemented whereby the hematocrit level results are submitted with each claim prior to the claim payment. This resulted in 35% reduction in the EPO spend.
Data leads to improved value
This effort – the collaboration of payer, employer and provider data creates a comprehensive source to support performance measurement resulting in improved value.
It is not enough to compare to the NCCN guidelines, but the need is to create rules to utilize for the comparison and application. Rules can be created in the areas of disease management, care patterns, and medication adherence.
Collaborative oncology management
A solid feedback/reporting/evaluation loop can be created by a phased process of selected oncology reporting measures aligned through the use of data. One natural progression is benchmarking against regional or national norms.
Given these challenges, initiatives for benchmarking can follow several steps:
1. Evaluate administrative data claims set to provide analytics and interpretation on the top gaps in care for Oncology
2. Compare payers data (member experience) and (claims data) against regional and/or national data.
3. Create a gap analysis and strategic plan with recommendations to target areas for focus.
4. Define the business case for expansion of data collection capabilities to include clinical data beyond the administrative data set. Determine the analytic tool to compare this data.
5. Define the data collection options to engage physicians in the data collection process
6. Marry both the administrative and clinical data for analysis and comparison.
7. Identify options to include clinical guidelines, such as the NCCN guidelines to ensure patients are receiving appropriate treatment.
Conclusion
In today’s world, with an eye toward value-driven healthcare, all stakeholders (patient, providers and insurers) are increasingly demanding that healthcare be based on documented clinical knowledge. The application of evidence-based guidelines holds the promise of increasing consistency, reduction in medical errors leading to faster, more predictable, higher quality and less costly patient recovery.
Marybeth Regan is an expert in disease and care management. She has written numerous articles on strategies for care and disease management. She may be reached at mb@yleen.com. Robin Randall-Lewis is an expert in consumerism. She may be reached at robin.randall-lewis@Optumhealth.com.