Hospital overnight and evaluation of systems and timelines study: A point prevalence study of practice in Australia and New Zealand

Background: Diurnal variation in the performance of rapid response systems has not been fully elucidated. Afferent limb failure (ALF) is a significant problem and is an important measure of performance of rapid response systems. Objective: To determine the diurnal variation in the detection and response to acute patient deterioration as measured by ALF, completeness of patient observations (Respiratory rate (RR); Pulse rate (PR) and Systolic blood pressure (SBP), and to explore the diurnal variation in the consequences of ALF in unanticipated admissions to the Intensive care unit (ICU) from the ward. Design, setting and participants: Point Prevalence study conducted on two days in 2012 in 41 ICUs in Australia and New Zealand, examining emergency (unanticipated) admissions to the ICU from the ward. Results: 51 patients from the ward were admitted as an emergency to the ICU following a rapid response team call, of whom 48 patients had complete datasets and were enrolled; 32 (67%) were men. The prevalence of ALF was 37.5% (18/48). Median age was 62.5 (IQR 51.5-74.0), Median APACHE II score was 21.0 (IQR 17-26). There was no diurnal variation in the prevalence of ALF (day 28% versus night 28%; p = 0.92), patient observations documented over time (p = 0.78 for RR, p = 0.95 for PR and p = 0.74 for SBP) or 28-day mortality (p = 0.24). There was a significant diurnal variation between the least recorded observation (SBP) and the most recorded observation (PR) (p < 0.01). ALF was more likely (day and night) if a complete set of observations had been taken (p < 0.01). Conclusion: The prevalence of ALF amongst patients admitted to the ICU from the ward is high. SBP is the least recorded patient observation. This study was unable to identify a diurnal variation in the prevalence of ALF, its consequences (i.e. mortality) and the completeness of patient observations. Observational studies with a larger sample are required to explore this important problem. (C) 2015 Elsevier Ireland Ltd. All rights reserved.