Covariates of individual patient vital sign observation timeliness in hospital wards

Ironside-Smith, Rupert, Allen, Stuart ORCID: https://orcid.org/0000-0003-1776-7489, Noe, Beryl and Turner, Liam ORCID: https://orcid.org/0000-0003-4877-5289 2025. Covariates of individual patient vital sign observation timeliness in hospital wards. Presented at: IEEE International Conference on Healthcare Informatics (ICHI) 2025, Cosenza, Italy, 18 - 25 June 2025.

Preview

PDF - Presentation Download (1MB) | Preview

Abstract

Vital sign observations are typically carried out by healthcare staff at regular intervals, known as ward rounds, to monitor the overall health status of individual patients. Patients who present symptoms of deterioration, or are deemed 'at risk' by clinical staff, will have their vital signs observed more frequently (e.g., hourly instead of on a 12 hour interval), decided by the nature of their condition. The frequency and documentation accuracy of vital signs observations has been well studied, but less consideration has been given to how clinical staff manage specific observation intervals amongst those that have been routinely scheduled. The primary aims of this study are to assess the current adherence to prescribed observation intervals and identify any primary factors that affect timeliness. This study uses empirical survival estimation methods to determine the empirical likelihood that, for a patient with a specified observation interval, the subsequent observation would be recorded within the planned interval. We discuss the management of various patient observation intervals across 20 study wards in south Wales and how these formulate routine ward rounds, or coexist with them. A semi-parametric proportional hazards model is then used to determine the extent that individual patient covariates, such as Early Warning Score (EWS), time of day, and sepsis, mediate a significant change to the baseline of vital sign observation timeliness. At the ward level, our findings suggest regular batching of vital sign observations irrespective of the planned schedules and a moderately positive linear relationship between observation interval length and the likelihood of timely recording. On an individual patient basis, elevated EWS was the strongest indicator that a patient's subsequent observation would be taken earlier than baseline, however, it is clear that baseline vital sign observation management is largely governed by existing ward round policies.

Item Type:	Conference or Workshop Item (Paper)
Status:	Unpublished
Schools:	Schools > Computer Science & Informatics
Subjects:	H Social Sciences > HA Statistics L Education > LF Individual institutions (Europe)
Uncontrolled Keywords:	Vital Sign Observations, Electronic Health Record, Secondary Data Study, Survival Analysis, Proportional Hazards
Date of Acceptance:	13 March 2025
Last Modified:	20 May 2025 10:00
URI:	https://orca.cardiff.ac.uk/id/eprint/178210

Actions (repository staff only)

Edit Item