causal inference in the health sciences a conceptual introduction精品.pdfVIP

Health Services Outcomes Research Methodology 2:189–220, 2001 # 2002 Kluwer Academic Publishers. Manufactured in The Netherlands . Causal Inference in the Health Sciences: A Conceptual Introduction JUDEA PEARL Cognitive Systems Laboratory, Computer Science Department, University of California, Los Angeles, CA 90024 E-mail: judea@cs.ucla.edu Received February 15, 2001; revised December 18, 2001; accepted January 3, 2002 Abstract. This paper provides a conceptual introduction to causal inference, aimed to assist health services researchers benefit from recent advances in this area. The paper stresses the paradigmatic shifts that must be undertaken in moving from traditional statistical analysis to causal analysis of multivariate data. Special emphasis is placed on the assumptions that underlie all causal inferences, the languages used in formulating those assumptions, and the conditional nature of causal claims inferred from nonexperimental studies. These emphases are illustrated through a brief survey of recent results, including the control of confounding, corrections for noncompliance, and a symbiosis between counterfactual and graphical methods of analysis. Keywords: structural equation models, confounding, noncompliance, graphical methods, counterfactuals 1. Introduction The research questions that motivate most studies in the health sciences are causal in nature. For example, what is the efficacy of a given drug in a given population? What fraction of deaths from given disease could have been avoided by a given treatment or policy? What was the cause of death of a given individual, in a specific incident? Not surprisingly, the central target of such studies is the elucidation of cause-effect relation- ships among variables of interests, for examp