Controls in clinical research

Abstract

The problems of controlled experiments in the human subiect are by no means solved by the balanced group method and double‐blind procedures‐a fact to be co!1JSidered in attempts to reduce error and raise confidence levels in clinical results. Among the sources of experimental error, particularly in psychiatry, may be the personality uf the investigator himself, affecting physiologic rapport ("concordance") with the patient, not to mention the premises he selects for establishing criteria. The vagaries uf the placebo reactor, the position of the placebo in the series of drugs tested, what the patient is told concerning the drug, the attitude uf the staff, the types of communication received from other patients in the group, the subiect's “set," his interest and ability in responding to his environment are among the factors which determine reliability of the results and the validity of the conclusions. “Love of truth" and “plain horse sense" remain the mainstays in sound clinical evaluation.