The phrase "outcomes studies" is usually used to describe those studies where outcome is assessed in large cohorts of patients, often using data from administrative databases. They are used to determine the role and impact of variations in the structure and process of health care delivery play in routine practice. Optimally, outcome should be assessed in terms of measures that are important to patients (e.g., quality of life). More often, clinically relevant outcome measures are lacking, and outcome is measured in terms of mortality, morbidity, and length of hospital stay. Although the outcomes movement continues to expand and much valuable information can be learned from this type of study, there are limitations. These studies are essentially observational and most often are performed using databases set up for other purposes. Thus data may be incomplete and incorrect. Moreover, they cannot assess the impact of patient preferences on outcome. The term outcomes studies has also been applied to small area variation and volume outcome studies.