Technology-driven outcomes: information technology will play a significant role in identifying those interventions that generate best outcomes - Disease Management

Health Management Technology, Jan, 2004 by Thad Perry

The disease management (DM) industry has come a long way in a short time. Early on, the major focus of DM was on the general practice guidelines for chronic conditions of interest, with the majority of effort placed on patient compliance with standards of care and adherence to pharmaceutical therapies. As this area of supplemental care evolved, support efforts shifted from a patient focus to a provider focus.

Proponents of evidence-based medicine (EBM) have made the healthcare community aware that a disturbingly high number of providers were not practicing or even aware of the most current EBM guidelines for their patients suffering from chronic conditions. To make matters worse, professional organizations and health plans experienced little success in disseminating this information to the provider community. The problem was not with the information itself; the problem was with information delivery.

There are too many patients and too few providers, too few hours to develop and foster strong patient-provider relationships, and too many demands on providers' time. Based on this reality, DM has become the hope and solution of choice. Health plans and employer groups have begun to embrace the potential that DM brings to the healthcare delivery system, and the Centers for Medicare and Medicaid Services has been evaluating a series of pilot programs designed to assess the efficacy of DM in a traditional Medicare population. Recently, the Medicare Prescription Drug and Modernization Act of 2003 included a provision for additional DM pilot projects.

In general, large-scale, population-based DM programs have only been in existence since the mid-1990s. The industry is several years away from having enough information to conclusively determine how DM works within and across populations. Consequently, DM continues to evolve as outcomes assessment and evaluation procedures are validated. Technology-driven processes and trends will be essential to DM during the next few years.

Integration of Data Sources

Over the last decade, there have been countless data warehousing efforts by health plans and DM vendors attempting to consolidate multiple sources of information into a system that will allow for easier data mining and analytic activities--some successful, others not. Regardless of the success of failure of a data warehousing initiative, the end result inevitably comes at great cost and effort to the organization.

Why have companies been so eager to attempt data warehousing initiatives when the costs remain in the tens of millions of dollars and the results are mixed at best? The reason is simple: Those organizations with the most complete information have the highest probability of uncovering what has become the "holy grail" for DM. That holy grail is the identification of the actual DM interventions that drive outcomes.

To date, no one can conclusively say what drives DM outcomes, such as printed educational material, outbound nurse telephone calls, inbound telephonic advice, scheduling assistance, electronic medical records of Web-based information services for providers. However, through data management and data mining techniques, the answer to this question can be grasped. The group that is able to identify and take advantage of the identification of those interventions will enjoy a significant advantage over competitors.

Management of Multiple Data Sources

The role that data management plays in DM cannot be overemphasized. To effectively search for and detect meaningful patterns, the source data must be organized into an analyzable format. Corrective adjustments such as missing or invalid data and coding errors must be addressed as data are extracted from source systems. Assuming that data were reliably and validly collected from their original source, data preparation techniques are used prior to loading these data into a database, data mart of data warehouse, depending on the system architecture.

Currently, to ensure data quality and integrity, most large database or data warehousing projects have adopted warehousing solutions and ETL (extract, transform, load) processes. There are many different commercial solutions available. For example, SAS Data Warehousing is designed for the integration of data sources across multiple platforms. Using a combination of scalable hardware and software, a system architecture is created to deliver accurate information across data sources and types.

Since the complexities involved with the management of DM data are numerous, this type of system would enable a health plan or DM vendor to gain a more complete view of the success of failure of their DM programs. It is only through this type of open and scalable system architecture that the integration of multiple data sources and types is possible.

Most DM activities center on the analysis of administrative claims data such as medical, pharmacy and laboratory claims. Additionally, data associated with health risk appraisals, general health assessments, provider and patient satisfaction, and claims groupers have become increasingly available. When these data are integrated, a patient's risk, treatment and health status can be determined.