Abstract
OBJECTIVE: To identify and classify all clinical decisions that emerged in a sample of patient-physician encounters and compare different categories of decisions across clinical settings and personal characteristics. DESIGN: Cross-sectional descriptive evaluation of hospital encounters videotaped in 2007-2008 using a novel taxonomy to identify and classify clinically relevant decisions (both actions and judgements). PARTICIPANTS AND SETTING: 372 patients and 58 physicians from 17 clinical specialties in ward round (WR), emergency room (ER) and outpatient (OP) encounters in a Norwegian university hospital. RESULTS: The 372 encounters contained 4976 clinically relevant decisions. The average number of decisions per encounter was 13.4 (min-max 2-40, SD 6.8). The overall distribution of the 10 topical categories in all encounters was: defining problem: 30%, evaluating test result: 17%, drug related: 13%, gathering additional information: 10%, contact related: 10%, advice and precaution: 8%, therapeutic procedure related: 5%, deferment: 4%, legal and insurance related: 2% and treatment goal: 1%. Across three temporal categories, the distribution of decisions was 71% here-and-now, 16% preformed and 13% conditional. On average, there were 15.7 decisions per encounter in internal medicine specialties, 7.1 in ear-nose-throat encounters and 11.0-13.6 in the remaining specialties. WR encounters contained significantly more drug-related decisions than OP encounters (P=0.031) and preformed decisions than ER and OP encounters (P<0.001). ER encounters contained significantly more gathering additional information decisions than OP and WR encounters (P<0.001) and fewer problem defining decisions than WR encounters (P=0.028). There was no significant difference in the average number of decisions related to the physician's and patient's age or gender. CONCLUSIONS: Patient-physician encounters contain a larger number of clinically relevant decisions than described in previous studies. Comprehensive descriptions of how decisions, both as judgements and actions, are communicated in medical encounters may serve as a first step in assessing clinical practice with respect to efficiency and quality on a provider or system level.