Outcomes research has achieved considerable stature over the last two decades. An increased emphasis on outcomes and assessment of effectiveness has come about in response to a number of factors. Among these factors is the observation of unexplained variation in medical care across geographic areas (1). Variation in practice patterns can, in part, be explained by patient preferences and/or differences in demographics (2). However, variation may also reflect suboptimal care in the low-use areas or needless cost in the high-use areas (3). Considerable variations in practice patterns are likely influenced by physicians’ approaches to clinical decision making. Physicians may approach similar medical problems with different theoretical assumptions. Eckman described it in Bayesian terms: Physicians bring differing prior probabilities to the decision making process, which leads to variations in their medical decisions (4). Difficulty incorporating an understanding of disease probabilities into decisions is potentially magnified further by variability in the perceptions of outcomes (2,4). The economics of health care have further driven the increased emphasis on outcomes. The need for cost containment in response to escalating health-care costs (2,3,5), competition within the healthcare sector (3), limitations of randomized control trials (6), findings of inappropriate use of medical services (1), and the need for greater accountability have all driven the increased interest in outcomes.

The field of outcomes research distinguishes itself from traditional research in that it can embrace a broader range of endpoints. It may focus on a number of patientcentered outcomes such as measurements of patient satisfaction and quality of life, daily functioning activities, and economics (3,7,8). Another distinguishing feature of outcomes research is its emphasis on effectiveness rather than efficacy. Efficacy denotes “usefulness of a medical intervention tested under optimal conditions” (9). Effectiveness can be thought of as the “utility” of a health-care intervention in routine clinical practice (9). An additional factor determining the usefulness of a medical intervention is its efficiency, that is, the value of the intervention to the individual patient (9). Johanson summarized these in a practical way: “efficacy addresses the question of whether an intervention can work, effectiveness answers the question of whether it works in a routine practice setting and efficiency determines whether it is worth doing” (9). These concepts figure prominently into the development of practice guidelines from the results of outcomes research. Similar to other research methods, outcomes research focuses on two things: choosing the right variable to study and selecting the data source to make inferences about that specific variable and its relationship to other independent variables (10). Perrin and Mitchell state that “outcomes research gets its name from the choice of the dependent variable.” It is the choice of, the definition of, and the development of measures for the dependent variable that has been the focus of research for the last decade. He stressed that current emphasis should be on the inference process: “defining the independent variables and thinking about study design and data sets that allow inferences to be made to set the tone for outcomes research for the future” (10).

A number of goals can be achieved by focusing on outcomes: increased understanding of the effectiveness of various treatment strategies; identifying the most efficient and effective use of limited resources and integrating these into the development of medical standards
and practice guidelines; and finally, optimizing the use of resources by third-party payers (3,9). The creation of medical practice guidelines by consensus panels assists physicians in the clinical application of specific diagnostic tests. The American College of Cardiology (ACC) and the American Heart Association (AHA) have developed guidelines for the clinical application of echocardiography. A panel of experts in the field of echocardiography developed the guidelines based on expert opinion and extensive review of the literature. The guidelines serve as a summary of current knowledge on the effectiveness of the imaging test, specifically classifying the clinical utility of echocardiography for specific cardiovascular diseases and common cardiovascular symptoms. Class I reflects evidence or general agreement that the given procedure is effective. Class II reflects conditions where there is conflicting evidence or opinion with regard to the usefulness of echocardiography. Class III reflects conditions for which there is evidence or general agreement that the procedure is not effective or useful (11,12). Recommendations for the use of transesophageal echocardiography (TEE) follow a similar guideline structure, provided by the American Society of Anesthesiologists and the Society of Cardiovascular Anesthesiologists Task Force on Transesophageal Echocardiography (13).

Outcomes research may incorporate a variety of methodologies and approaches to collecting, examining, and analyzing data. Data sources include results from randomized clinical trials, quasiexperimental designs, and effectiveness trials (9), cohort or case-control designs, or other retrospective analyses from databases and routine observational data from hospital discharge summaries (3,7,8,14). Data synthesis can be performed in multiple ways, such as metaanalysis, decision analysis, and crossdesign synthesis. It is useful to consider the strength of the evidence produced by a study as it relates specifically to a study design; data sources can be listed in descending order of the strength of evidence they produce: randomized controlled trial, nonrandomized trial with contemporaneous controls, nonrandomized trial with historical controls, cohort study, case-control study, cross-sectional study, surveillance (database), descriptive study, and case report (15,16). All of these methods have strengths and weaknesses and can provide valuable information for outcomes research.