Capturing judgement strategies in risk assessments with improved quality of clinical information: How nurses’ strategies differ from the ecological model

Huiqin Yang, Carl Thompson

Nurses' risk assessments of patients at risk of deterioration are sometimes suboptimal. Advances in clinical simulation mean higher quality information can be used as an alternative to traditional paper-based approaches as a means of improving judgement. This paper tests the hypothesis that nurses' judgement strategies and policies change as the quality of information used by nurses in simulation changes. Sixty-three student nurses and 34 experienced viewed 25 paper-case based and 25 clinically simulated scenarios, derived from real cases, and judged whether the (simulated) patient was at 'risk' of acute deterioration. Criteria of judgement "correctness" came from the same real cases. Information relative weights were calculated to examine judgement policies of individual nurses. Group comparisons of nurses and students under both paper and clinical simulation conditions were undertaken using non parametric statistical tests. Judgment policies were also compared to the ecological statistical model. Cumulative relative weights were calculated to assess how much information nurses used when making judgements. Receiver operating characteristic (ROC) curves were generated to examine predictive accuracy amongst the nurses. There were significant variations between nurses' judgement policies and those optimal policies determined by the ecological model. Nurses significantly underused the cues of consciousness level, respiration rate, and systolic blood pressure than the ecological model requires. However, in clinical simulations, they tended to make appropriate use of heart rate, with non-significant difference in the relative weights of heart rate between clinical simulations and the ecological model. Experienced nurses paid substantially more attention to respiration rate in the simulated setting compared to paper cases, while students maintained a similar attentive level to this cue. This led to a non-significant difference in relative weights of respiration rate between experienced nurses and students. Improving the quality of information by clinical simulations significantly impacted on nurses' judgement policies of risk assessments. Nurses' judgement strategies also varied with the increased years of experience. Such variations in processing clinical information may contribute to nurses' suboptimal judgements in clinical practice. Constructing predictive models of common judgement situations, and increasing nurses' awareness of information weightings in such models may help improve judgements made by nurses.