Outcomes as a Management Tool

This issue of Journal of Cardiopulmonary Rehabilitation includes an article by Jungbauer and Fuller presenting the work of the Indiana Society of Cardiovascular and Pulmonary Rehabilitation in establishing a multi-state outcomes program. The article describes an enormous effort to design and implement an outcomes database for programs serving both cardiac and pulmonary patients. The need to assess outcomes has been questioned by some, particularly those working under conditions of minimal resources. The value of monitoring outcomes, however, is becoming increasingly apparent for both patient and program management as well as for the advancement of cardiopulmonary rehabilitation as a profession. Why Measure Outcomes? The primary reason to measure outcomes is to better manage our patients' health. Without outcomes measures how can we know if we are successful? Without feedback, how can we adjust our treatment plan for those who are not improving? In the past, we have assumed our programs were beneficial, particularly in terms of quality of life, social support, and functional capacity. We often based our assumptions on observations of the success of those who regularly participate and on positive comments from patients or their families. Relying on such highly selective data, however, can be deceptive. Without systematic data collection, we are left with little but clinical impressions to guide our therapeutic decisions. Outcome assessment serves a similar management function in our attempts to build better, more cost-effective programs. Summary feedback on how patients are progressing (i.e., percentage of smokers remaining abstinent at 6 months) helps us determine the type and mix of services that are most effective. Successful outcomes also demonstrate to hospital administrators, managed care reviewers, and other third party payers that our programs significantly improve the lives of our patients and are worthy of continued funding. Continuous enhancement of our services is essential to the long-range success of our profession. If cardiopulmonary rehabilitation services are to survive and prosper in the competitive environment of the years ahead, we must demonstrate that we offer the most cost-effective services available. This is particularly true as we compete with disease management programs that target the total population of patients with cardiac or pulmonary disease, rather than only those who decide to participate in a traditional on-site rehabilitation program. To document our effectiveness in providing these more comprehensive services, we will have to rely heavily on computerized medical records and data management systems. How Do You Measure Outcomes? Once convinced of the importance of monitoring outcomes, one of the first frustrations voiced by practitioners is "I'm overwhelmed-what am I supposed to measure? What do I use to measure it? How often should I measure it?" Practitioners may be unaware of options that are available, or they may be overwhelmed by too many choices! Additional questions often include: Who will organize and report the data, and to whom? How do I ensure the reliability and validity of the data entered into database (i.e., flagging obvious errors, such as blood pressure reported as 42/94)? What do I do with missing data and dropouts? When do I measure (i.e., at entrance, exit, or 3 months, 6 months, or 1 year)? The American Association of Cardiovascular and Pulmonary Rehabilitation (AACVPR) has been most helpful in providing resources to guide programs in developing outcome assessments. An article authored by the Outcomes Committee,1 Outcomes Tools Resource Guide (available from the AACVPR national office), and Outcomes Committee presentations at the annual meetings are excellent places to find answers to many of these questions. Be forewarned, however, that although these resources are helpful, they are not likely to provide one standard set of outcome measures and procedures useful for all programs. It will be necessary for each program to tailor their outcome assessment to the specific needs and resources of the program. What Does the Data Mean? Because programs rarely have the luxury of measuring outcomes on patients not receiving cardiopulmonary services (i.e., there is no control group), programs cannot directly assess the benefits their patients derive specifically from participating in cardiopulmonary rehabilitation. The general effectiveness of cardiopulmonary rehabilitation, however, already has been established and is well documented in the Agency for Health Care Policy and Research Clinical Guideline for Cardiac Rehabilitation (1995)2,3 and the recent evidence-based guidelines on pulmonary rehabilitation authored by the American College of Chest Physicians and AACVPR (1997).4 Hence, our challenge is to optimize the delivery of services already shown to be effective. Individual programs can assess their performance by comparing their outcomes to those of other programs offering similar services to similar patient populations. This comparison is called benchmarking. Benchmarking is the process of comparing outcomes from several programs in a standardized fashion (i.e., same programmatic outcomes and measurement instruments, similar time frame for taking measurements). Such comparisons provide feedback regarding program performance in terms of being above, at, or below average performance. With benchmarking, we can identify programs that excel in one or more areas, thus providing the opportunity to learn about "best practices." However, benchmarking across programs is not without problems. Patient populations differ on many factors, for example disease severity, education, income level, and culture; these are all factors that can affect outcomes. In addition, programs will differ in how aggressively they pursue information about individuals who are not compliant in completing outcome assessment instruments or who drop out of the program entirely. Programs that do not attempt to follow these individuals (who might reasonably be expected to have worse outcomes) might appear to be more "successful" than programs that try to follow the progress of all of their patients. Thus, there is need for caution when comparing results across programs. The benchmarking process could be greatly facilitated by compiling a minimum set of standardized outcomes, much like the National Committee for Quality Assurance (NCQA) has done with their test measures for accrediting managed care organizations. Once we as a profession agree on a minimum set of outcomes with standard instruments and a timetable, we could more easily apply the benchmarking process. Internet technology could be used to enable programs to input this standard set of outcomes into a databased website, allowing us to benchmark programs and examine the relative effectiveness of various therapeutic approaches across countries and cultures. With so much to gain, perhaps our discussion should focus not on whether or not to measure outcomes (and not on whether or not our programs demonstrate effectiveness) but on what we can agree on as a minimum set of outcome measures that best capture management efforts across the patient, program, and professional spectrum. It also would enable us to continue to build on our success in doing what we do best-helping patients live with cardiopulmonary disease.