In 2011, the Medicare Electronic Health Record (EHR) Incentive Program began paying physicians and hospitals for the measurement and reporting of performance metrics through the utilization of certified electronic medical record (EMR) software. This epic program began in earnest the irrevocable shift in the provision of healthcare from an individual craft-based art to a structured, evidence-based system. The EHR Incentive Program has three specific phases that progressively build from data capture and reporting, through clinical process refinement, to the final goal of achieving measurably improved healthcare outcomes.

 Currently, we are in the first phase of this transition, and in oncology, trends are emerging about what metrics are important to three key groups—patients, payers, and providers. Each seeks improvements to quality and lower&mdash;or at least stable&mdash;costs for cancer care. The broad utilization of EMR technology has, for the first time in history, enabled systematic measures of care delivery on a broad scale.

While management of clinical care using modern information technology is important to providers and payers of oncology care, it is critical to those of us who have a cancer diagnosis in our family or in our future— and that is just about everyone. Providers must participate in the move to active measurement of clinical variables or become obsolete. They must also integrate reporting and measurement of important process measures into routine clinical management. Payers have a critical role in promoting this technology adoption. They are positioned to promote payment methodologies that support and reward clinical measurement. Without such support, innovative practices suffer economically as they push toward rational care. We all should expect basic quality process measures that are indicative of good clinical management of cancer care for our families and ourselves. Successful payers and providers will encourage patients to seek care where these quality indicators are available and exemplary.

Three visions of performance metrics in oncology were presented at the Cancer Center Business Summit held October 11-12, 2012, in Fort Worth, Texas. The theme of the conference was transitioning to value-based oncology, and most of the program content, including the full program that is summarized in this article, is available online.

 While the three perspectives are different, there were consistent themes throughout. All speakers endorsed the use of available data, keeping metrics simple, delivering measurements to the care providers who can affect the numerator, and showing these providers how they compare with peers. The speakers all understand that current payment methodology does not support process optimization, yet view that as a transitional barrier that will not persist.

		Figure 1. Survey Data Showing Who Is Entering Information Into the EMR

Within your practice, who is primarily responsible for entering information for the following data fields? 

I presented the three criteria that are generally accepted for choosing clinical performance measures.

 To be successful, measures must be important, scientifically sound, and feasible. Importance is measured by the relevance of the measure to patients and providers and the promise that being measured offers improvement. To be scientifically sound, there must be substantial, explicit evidence with validity, reliability, and sufficient specificity to patient factors to be clinically useful. Feasibility requires an explicit definition of the numerator and denominator needed for the measure from data that are available at low cost and with low administrative burden. Today, the fact is that payment drives process measurement, and most systematic data are about billing, not about medical care. The EHR Incentive Program, coupled with many providers and payers seeking a departure from fee for service, are factors driving oncology data feasibility toward important, scientifically sound measurements.

The presence of specific, essential clinical information as structured data in the clinical database is necessary for oncology care management. These data support pathway and guideline adherence, reduce treatment variability, take advantage of emerging companion treatment diagnostics, and promote cost savings through the avoidance of valueless care. In oncology, six basic elements are needed to enable such measurement. For every patient, we need to have structured data in the electronic medical record for stage, intent of therapy, toxicity, disease status, patient status, and line of therapy. When there is substantial data density for these six process measures, the foundation is in place to begin to measure clinical outcomes across large populations of patients, and then in real time, see what is working.

		Figure 2. Sample Process Metric Presented to Physicians

So how do we get from low density, as presently observed in oncology EMR data, to high density on these process measures? 

 shows data from an Oncology Metrics survey performed in 2009 and reveals that physicians directly enter most of these data as part of their clinical management activity.

There is ample evidence that peer review does alter physician behavior. Oncologists who received feedback about the amount of chemotherapy that they were giving at the end of life decreased from 50% of patients to 20% of patients in only 6 months.

 Structured clinical data offer the opportunity to show individuals how they compare with named physicians in their own practice. Additionally, this lets the physician know that others see his score too (

		Figure 3. Service Line Dashboard From Catholic Health Initiatives

Deb Hood, vice president of the national oncology service line of Catholic Health Initiatives (CHI), spoke about the experience of her institution in beginning to establish what she called 

. While there are five dimensions to their hospital perspective, she emphasized improvement in quality and reduction of cost. Although there are many quality metrics in oncology, CHI has decided to focus on data that are readily available and easy to audit. The best data source for them is registry data, as they are extracted in real time and encompass more that 30,000 new cases each year. From these data, they have created “dashboards” for breast cancer, exception reporting for chemotherapy and radiation therapy, readmission tracking for all oncology patients, and pharmaceutical expense tracking. To leverage these reports to create measurable improvement in care, they are presenting comparative measures at the institutional level with color coding to promote easy understanding. Green signifies compliance, yellow is nearly in range, and red is out of compliance standard. A sample of such a report card is shown in 

Wes Chapman, president and CEO of PCD Partners, provided an overview of the application of ISO 9001 and Lean Six Sigma in two large-scale projects.

 Noting that outcome metrics are only viable in the context of a uniform process, and understanding that healthcare provision is not at all a uniform process, useful metrics are limited to process metrics (

		Figure 4. Hierarchy of Metrics and the Organizational Sponsors

Transitioning to Value-Based Oncology: Strategies to Survive and Thrive 

To adequately measure processes in large systems, the integration of documents and data into a single system is necessary. Chapman said that this novel functionality could be created in a cloud-based relational database but noted that these two types of information are not standardized. This shortcoming limits the ability of process measurement to provide compelling information about problems addressed and determine if improvements are realized. To achieve process control and improvement, it is necessary to define the process and train process operators, collect process performance data, and then audit process for compliance. None of this functionality is present in large-scale systems today and this is a limiting reality to the application of tested process management technology that is used in the manufacturing environment. To no small degree, this fractioning of healthcare is part of the payment methodology. Fee for service provides incentives for more process steps and leads to variable endpoints, uncoordinated care, and undocumented outcomes. When healthcare delivery and payment more closely resemble typical manufacturing environments, characterized by a drive to reduce process steps and maximize value, then the tools from that sector will become applicable

Articles

In 2011, the Medicare Electronic Health Record  Incentive Program began paying physicians and hospitals for the measurement and reporting of performance metrics through the utilization of certified electronic medical record software. 

Thomas R. Barr, MBA

Articles

Performance Metrics in Oncology