December 2003 Bulletin


The importance of outcomes research By Michael G. Vitale MD, MPH and
Michael W. Keith, MD

Outcomes research differs from more traditionally performed clinical research in the rigor and variety of methodologies that are used and in the nature of the endpoints that are chosen to define the patient’s outcome. Randomized clinical trials, controlled prospective studies, economic evaluations, analysis of large administrative data structures and studies that attempt to affect public policy are now being performed with increasing frequency. Evidence-based medicine is now part of the orthopaedic lexicon and some of our orthopaedic publications now rate the methodological quality of the work.

Measures of patient outcome
There has been a flurry of attention aimed at identifying standardized, appropriate, valid and reliable measures of patient outcome including health-related quality of life. Traditional quantitative measures of clinical results, such as radiographs and range of motion, still provide important information and there has been significant progress in developing and validating disease-specific scores, which summarize this information.

The AAOS Musculoskeletal Outcomes Data Evaluation and Management System (MODEMS) project involved the development of disease- specific outcome questionnaires for each of the major areas of orthopaedic surgery. Though the online mechanism of the MODEMS project is no longer in operation, the tools — including the Disabilities of the Arm, Shoulder and Hand Instrument (DASH) and the Pediatric Parent-Child/Adolescent Instruments — continue to be commonly used in clinical research.

The importance of broader, generic, patient-oriented measures is now widely accepted. Once viewed with some skepticism, generic measures of health status, such as the SF-36 and its pediatric equivalent, the Child Health Questionnaire, are now routinely seen in the orthopaedic literature. Most orthopaedic surgery is performed to decrease pain and improve function and it is not surprising that these ortho- paedic interventions have a dramatic impact on quality of life. In fact, numerous orthopaedic treatments rate at the very top of scales that compare the improvement of quality of life with various types of medical intervention.

There has also been a stronger interest in examining the full range of effects that an orthopaedic disability has on the family and on society as a whole. Work is underway to better define issues related to the “burden of care,” that is, the broader impact that disease and treatment have on the family unit as a whole. Any family with a child with cerebral palsy or a relative that is no longer independently ambulatory will appreciate the social consequences of a family member’s disease.

Outcomes assessment
The science of outcomes assessment is still very much in the “steep part of the learning curve” and we are witnessing rapid improvements in this field. Disease-specific and generic measures continue to evolve and shorten. Yet, routine use of these measures does place some time and effort burdens on both the physician and the patient. Recent advances in psychometrics (the science of design, administration and interpretation of questionnaires) have created an exciting opportunity to measure patient outcomes much more efficiently.

Computer Adaptive Testing (CAT), or “dynamic assessment,” is an exciting methodology that has emerged to simplify the measurement of scholastic performance on standardized tests. Using the power of the computer and newly developed statistical principles, it is now possible to tailor the test (or questionnaire) to the student (or patient) so that a precise valid score can be obtained using a much smaller set of questions. These methods have been applied to the science of outcomes assessment. For example, an “adaptive” SF-36 is currently in use and available online at

As the quality of administrative databases has continually improved, we are witnessing an expansion in the use of these data to inform discussion regarding health policy. For example, numerous studies have now convincingly demonstrated relationships between procedure volume and improved outcomes. Work is underway to develop a state-of-the-art national trauma registry for children and a national total joint registry. Such efforts promise to accelerate the process of clinical research and subsequent improvements in practice patterns.

While the use of the computer and the Internet can greatly facilitate multicenter clinical research, it is not clear what the effect of the Health Informa- tion Portability and Accountability Act will be, given new restrictions on sharing data.

Our objectives must be to continue to make the patient’s perception of outcome relevant to our clinical decision-making process and to help guide the selection of orthopaedic procedures, devices and pharmaceuticals. Although there continue to be challenges to the design and conduct of rigorous patient-oriented outcomes research, improvements in the methodology and infrastructure promise to improve our ability to perform clinical research and to better understand how we can best serve our patients.

Michael G Vitale, MD, MPH, and Michael W. Keith, MD, are both members of the AAOS Evidence-Based Practice Committee.

Dr. Vitale is the Herbert Irving Assistant Professor of Orthopaedic Surgery, Children’s Hospital of New York, New York-Presbyter- ian Hospital, and Director of the Division of Pediatric Outcomes Assessment at the International Center for Health Outcomes and Innovation Research, Columbia University. He can be reached at (212) 305-5465.

Dr. Keith is Professor of Orthopaedics and Biomedical Engineering at the MetroHealth Medical Center campus of Case Western Reserve University in Cleveland, Ohio. He can be reached by phone at (216) 778-4399.

Home Previous Page