Perioperative temperature monitoring for patient safety: A period prevalence study of five hospitals

Abstract

Background: Monitoring body temperature is essential for safe perioperative care. Without patient monitoring during each surgical phase, alterations in core body temperature will not be recognised, prevented, or treated. Safe use of warming interventions also depends on monitoring. Yet there has been limited evaluation of temperature monitoring practices as the primary endpoint.

Objective: To investigate temperature monitoring practices during all stages of perioperative care. We examined what patient characteristics are associated with the rate of temperature monitoring, along with clinical variables such as warming intervention or exposure to hypothermia.

Design: An observational period-prevalence study over seven days across five Australian hospitals.

Settings: Four metropolitan, tertiary hospitals and one regional hospital.

Participants: We selected all adult patients (N = 1690) undergoing any surgical procedure and any mode of anaesthesia during the study period.

Methods: Patient characteristics, perioperative temperature data, warming interventions and exposure to hypothermia were retrospectively collected from patient charts. We describe the frequencies and distribution of temperature data at each perioperative stage, including adherence to minimum temperature monitoring based on clinical guidelines. To examine associations with clinical variables, we also modelled the rate of temperature monitoring using each patient's count of recorded temperature measurements within their calculated time interval from anaesthetic induction to postanaesthetic care unit discharge. All analyses adjusted 95% confidence intervals (CI) for patient clustering by hospital.

Results: There were low levels of temperature monitoring, with most temperature data clustered around admission to postanaesthetic care. Over half of patients (51.8%) had two or less temperatures recorded during perioperative care and one-third (32.7%) had no temperature data at all prior to admission to postanaesthetic care. Of all patients that received active warming intervention during surgery, over two-thirds (68.5%) had no temperature monitoring recorded. In our adjusted model, associations between clinical variables and the rate of temperature monitoring often did not reflect clinical risk or need: rates were decreased for those with greatest operative risk (American Society of Anesthesiologists Classification IV: rate ratio (RR) 0.78, 95% CI 0.68-0.89; emergency surgery: RR 0.89, 0.80-0.98), and neither warming interventions (intraoperative warming: RR 1.01, 0.93-1.10; postanaesthetic care unit warming: RR 1.02, 0.98-1.07) nor hypothermia at postanaesthetic care unit admission (RR 1.12, 0.98-1.28) were associated with monitoring rate.

Conclusions: Our findings point to the need for systems-level change to enable proactive temperature monitoring over all phases of perioperative care to enhance patient safety outcomes.

Registration: Not a clinical trial.

Keywords: Monitoring; Perioperative care; Physiologic; Prevalence; Temperature; Vital signs.