Obstacles to effective developmental surveillance: errors in clinical reasoning

Abstract

Recent research and legislation support the importance of early identification and intervention for children with developmental and behavioral or emotional problems. Detecting these children often depends on medical professionals, especially pediatricians. However, few pediatricians use developmental screening tests to help them identify children. Rather, physicians usually rely on their clinical impressions to discriminate children with and without difficulties. Research on the accuracy of clinical impressions, although sparse, suggests that only half the children in need are identified. The most obvious reasons, such as severity of the problem or the type of clinical information physicians select (e.g., parents' concerns, observations of the child, history, etc.), do not fully explain why some children are identified and others are not. More complete explanations are found in research on clinical impression formation that suggests physician's selection from the array of clinical data is mediated by their unique experiences, beliefs, and attitudes. These qualities provide a set of judgment heuristics for sorting seemingly relevant from irrelevant information. Judgment heuristics, depending on their content, may lead to accurate or inaccurate impressions. This article suggests a model of ideal impression formation that may help physicians learn to more accurately identify children with developmental and behavioral or emotional problems.