Temporal patterns in vital sign recording within and across general hospital wards

Abstract

Introduction: The use of mobile devices on hospital wards to record patient vital signs and Early Warning Scores provides opportunity for secondary analysis of the data collected. This research investigated how such analysis can contribute to the understanding of the complexities of managing clinical care in hospital environments.

Methods: The influence of ward type and the distribution of patient observation intervals was evaluated in relation to the timing of vital signs observation patterns in data collected from eight adult in-patient wards over a 12-month period. Actual and projected observation times were compared across patients with higher and lower National Early Warning Scores (NEWS).

Results: Both ward type and the distribution of patient observation intervals were significant predictors of temporal observation patterns. Observation patterns showed evidence of grouping of observation recordings. This was, however, not found for observations of patients with higher NEWS scores (3 or more).

Conclusions: Secondary analysis of vital signs observation data can reveal insights into how ward operate. The patterns of observation recordings within a ward are a reflection of ward type and the distribution of patient observation intervals. The grouping of observation recordings of patients with low NEWS (<3) result in late or early observations to fit activity peaks characteristic of the ward culture.

Keywords: Care strategy; CareFlow; Early warning score; Observations; Vital signs.

Fig. 1

Proportion of actual and projected observations for an example ward, considering only (a) on-time observations and (b) all observations.

Fig. 2

Proportion of observations per TTNO and per hour of day for all wards.