Original Article
Associations between patient and system characteristics and MET review within 48 h of admission to a teaching hospital: A retrospective cohort study

https://doi.org/10.1016/j.ejim.2019.05.021Get rights and content

Highlights

  • At admission, patient diagnosis, comorbidities are associated with later MET review.

  • Charlson Score 1–2 or ≥ 3; 3+ prior emergency admissions; are important risk factors.

  • MET risk double in admissions for colorectal, respiratory, upper GI or infection.

  • Health system factors including time or season of admission associated with MET.

  • Associations similar for MET within 48 or 72 h of admission.

Abstract

The Medical Emergency Team (MET) has enhanced the recognition and response to clinical deterioration in acute healthcare. However, patients reviewed by the MET are at increased risk of in-hospital death. Identifying patients at risk of deterioration may improve patient outcomes.

Aim

To identify patient demographic, medical characteristics and healthcare systems and processes at the time of admission (baseline), associated with Medical Emergency Team (MET) review within 48 h (MET-48 h) of admission.

Methods

Single-site, year-long, retrospective cohort comprising patients admitted for at least 24 h, using routinely collected hospital data. A three-stage modelling approach was used to identify baseline factors associated with MET-48 h

Results

The study included 15,695 patients with mean age 62.1 years (SD 19.6), male (53.5%), born in Australia or New Zealand (60.9%) and 51.6% held a low-income concession card. A total of 4.3% of patients received a MET review within 48 h of admission. Variables independently associated with MET-48 h in a fully adjusted logistic model included age of 80 years or more (OR = 1.37); ≥3 previous emergency admissions (OR = 1.59); Charlson Comorbidity Index 1 or 2 (OR = 1.47), or ≥ 3 (OR = 1.99); history of alcohol-related behaviour concerns (OR = 2.04), chronic heart failure (OR = 1.48); chronic obstructive pulmonary disease (OR = 1.35); admission for colorectal (OR = 2.66) or upper gastro-intestinal (OR = 1.94) surgery, respiratory or tracheostomy (OR = 2.24); immunology and infections (OR = 1.90); emergency admission (OR = 1.36); admission at night (OR = 1.74), or summer (OR = 1.41)

Conclusions

This is the first study to demonstrate the potential to predict clinical deterioration using data that is readily accessible at the time of admission to hospital.

Introduction

Rapid response systems (RRSs) were developed to support the recognition and response to clinical deterioration, in order to decrease the high mortality associated with adverse events such as in-hospital cardiac arrest and unplanned Intensive Care Unit (ICU) admissions [1]. RRSs are organisation-wide systems and consist of four parts [2]. First, the afferent limb which comprises the recognition of deterioration and activation of the response system; second, a Rapid Response Team (RRT) to respond to deteriorating patients; third, an administration limb consisting of the day-to-day resourcing and management of the RRS; and finally, a governance limb for monitoring and quality improvement of the RRS [2]. In Australia, the most common model of RRT is the Medical Emergency Team (MET) [3,4], an ICU-specialist led response team, summonsed in response to pre-defined single -parameter criteria for recognising clinical deterioration [2,4].

Although the effectiveness of RRSs has been questioned [5,6], in three meta-analyses RRSs were associated with a decrease in in-hospital cardiac arrests and, to a lesser extent, hospital mortality [[7], [8], [9]]. However, the MET is a reactive approach to recognising and responding to clinical deterioration, is resource intensive and often under-funded, and patients reviewed by the MET are at high risk of in-hospital death [[10], [11], [12]]. This reactive approach may also encourage ward-based clinicians to defer care decisions and abdicate responsibility for patient care [6,13].

Earlier recognition of deterioration may further improve patient outcomes [14,15], and experts have recommended that the focus of researchers and clinicians shifts to predicting and preventing clinical deterioration [11,16]. A better understanding of risk factors for clinical deterioration will assist clinicians to identify high-risk patients and enable proactive, informed clinical care decisions to be made before deterioration occurs. MET review records are currently the earliest, reliable, evidence-based indication of clinical deterioration events in acute care patients. Most MET reviews take place within the first 48 h of an admission [17]. Therefore identifying patient and admission characteristics that have important associations with MET review within 48 h and are readily accessible to clinicians at the time of admission, is a first step towards proactive, preventative care for patients at increased risk of clinical deterioration.

The aim of this study was to identify: i) patient demographic characteristics; ii) patient medical characteristics; and iii) factors related to the healthcare system and processes present at the time of admission to hospital that are associated with clinical deterioration in the first 48 h using routinely collected hospital data.

Section snippets

Study design

This was a retrospective cohort study, using routinely collected hospital data for the period January 1 to December 31, 2016.

Ethics approval

Approval was received from the Human Research Ethics Committee at the study site (LNR/16/Austin/562), and Deakin University (2017–041), prior to study commencement. The requirement to obtain consent from participants was waived according to local and national policy regulations.

Setting

The setting was the major acute care campus of a three-campus health network in the northern

Data sources and management

The demographic, clinical and systems data used in this study were routinely collected by Victorian health services as part of the Victorian Admitted Episodes Dataset (VAED) [27]. Healthcare services report these data to the state government to enable service evaluation and funding decisions. Information is initially entered into an individual's health record by clinicians and hospital clerks contemporaneously, then extracted and coded by professional coders following hospital separation.

Results

A total of 22,241 admissions in 15,695 patients meeting the eligibility criteria occurred during the study period. Of these, 12,209 had a single admission in the study period, while 3486 patients had multiple admissions, of which we selected the last admission in the study period.

The mean age at admission was 62.1 years (SD 19.6), and 53.5% were male. Most held a low-income concession card (51.6%) and only 29.6% were privately insured. The majority were born within Australia or New Zealand

Discussion

The aim of this study was to identify patient and admission characteristics that have important associations with clinical deterioration, and are readily accessible to clinicians at the time of admission. The study included 15,695 patients, representative of an acute inpatient population of a tertiary referral hospital. MET review within 48 h of admission was selected as a surrogate outcome indicating clinical deterioration. One in twelve patients received a MET review at any time during their

Conclusions

This was the first study to demonstrate the potential to use information that is readily accessible at the time of admission to predict clinical deterioration. In this single-site cohort study several factors including previous emergency hospital admissions; Charlson Comorbidity Index score; history of chronic cardiac or respiratory illness; and admission diagnosis, time of day, and season were each independently associated with MET review within 48 h. These results enable the development of

Conflicts of interest

There are no competing interests to declare.

Acknowledgements

We gratefully acknowledge the contribution of Ray Robbins in extracting and collating the data used in this study.

References (40)

  • M.A. Devita et al.

    Findings of the first consensus conference on medical emergency teams

    Crit Care Med

    (2006)
  • D. Jones et al.

    Introduction of medical emergency teams in Australia and New Zealand: a multi-centre study

    Criti Care (London, England)

    (2008)
  • D.A. Jones et al.

    Rapid-response teams

    New Engl J Med

    (2011)
  • G. Salvatierra et al.

    Rapid response team implementation and in-hospital mortality*

    Crit Care Med

    (2014)
  • T. Sakai et al.

    Rapid response system

    J Anesth

    (2009)
  • R. Maharaj et al.

    Rapid response systems: a systematic review and meta-analysis

    Crit Care

    (2015)
  • A. Jong et al.

    Effect of rapid response systems on hospital mortality: a systematic review and meta-analysis

    Intensive Care Med

    (2016)
  • R.S. Solomon et al.

    Effectiveness of rapid response teams on rates of in-hospital cardiopulmonary arrest and mortality: a systematic review and meta-analysis

    J Hosp Med

    (2016)
  • The Joint CICM and ANZICS SIG

    On RRTs and ANZICS-CORE. resource use, governance and case load of rapid response teams in Australia and New Zealand in 2014

    Crit Care Resusc

    (2016)
  • D.A. Jones et al.

    Clinical deterioration in hospital inpatients: the need for another paradigm shift

    Med J Aust

    (2012)
  • View full text