To the Editor.
—A long-term goal and practice of mine has been to avoid wasting time in data gathering. This means choosing the most efficient, accurate, and least expensive approaches available for obtaining accurate information. The article by Drs Holleman and Simel1 shows that it is possible for three of four physicians to detect airflow obstruction when a cigarette smoking history of at least 70 pack-years is combined with a skillful chest examination.Their approach is time-consuming and results in an unacceptable high error rate, an important issue recently illustrated by Leape.2 The time-consuming aspect and inaccuracy were demonstrated by Schneider and Anderson.3 In their study, nine physicians were given 2 hours to examine 13 patients for evidence of chronic obstructive pulmonary disease. Some did not have enough time to complete the examination, and half the time their diagnosis was incorrect. Three peak flow measurements could have