Observation charts with overlapping blood pressure and heart rate graphs do not yield the performance advantage that health professionals assume: an experimental study

Authors

  • Melany J. Christofidis BSc,

    PhD Candidate, Corresponding author
    1. School of Psychology, The University of Queensland, St Lucia, Brisbane, Australia
    Search for more papers by this author
  • Andrew Hill LLB BA PhD,

    Senior Research Fellow
    1. School of Psychology, The University of Queensland, St Lucia, Brisbane, Australia
    2. Clinical Skills Development Service, Queensland Health, Herston, Brisbane, Australia
    Search for more papers by this author
  • Mark S. Horswill BSc PhD CertFPS,

    Associate Professor
    1. School of Psychology, The University of Queensland, St Lucia, Brisbane, Australia
    Search for more papers by this author
  • Marcus O. Watson BSc MSc PhD

    Associate Professor and Executive Director
    1. School of Psychology, The University of Queensland, St Lucia, Brisbane, Australia
    2. Clinical Skills Development Service, Queensland Health, Herston, Brisbane, Australia
    3. School of Medicine, The University of Queensland, Herston, Brisbane, Australia
    Search for more papers by this author

Abstract

Aim

To investigate whether overlapping blood pressure and heart rate graphs improve chart-users' ability to recognize derangements in these vital signs on hospital observation charts.

Background

Many health professionals prefer blood pressure and heart rate graphs to overlap. One justification is the use of a visual cue called the ‘Seagull Sign’ to detect physiological abnormalities.

Design

A 3 × 2 × 2 mixed-design experiment, with three independent variables: participant group, graph format (separate vs. overlapping) and alerting system (integrated colour-based track-and-trigger system present vs. absent).

Methods

Over 64 experimental trials, ‘Seagull-trained’ nurses and novices randomly assigned to receive ‘Seagull training’ or remain untrained, viewed sequences of blood pressure and heart rate observations recorded on four different chart design extracts. The designs represented a crossing of the graph format and alerting system variables. For each design, eight cases contained normal data and eight contained an abnormal systolic blood pressure or heart rate observation (half of which yielded a Seagull Sign on overlapping plots). Participants (tested between January–May 2011) judged whether observations were physiologically normal or abnormal.

Results

Across all cases, participants from all groups responded faster and made fewer errors when blood pressure and heart rate observations were graphed separately, especially when a track-and-trigger system was present. Even for ‘Seagull-trained’ participants viewing ‘Seagull Sign available’ cases, no advantage of overlapping graphs was found.

Conclusions

These findings suggest that overlapping graphs do not yield the performance advantage that many health professionals assume, either for novices or experienced nurses, even when the Seagull Sign is used.

Ancillary