Research - CriSTAL

Research Project Overview

CriSTAL: Criteria for Screening and Triaging to Appropriate aLternative care.

Synopsis

This project has validated a screening tool that identifies older terminal/dying patient on admission, to estimate their risk of death in the short term and facilitate discussions on care preferences including hospitalisations or community care. The project activities consist of collecting items from the medical record using a standard one page instrument (the CriSTAL tool). Development manuscript published in BMJ Supportive and Palliative Care journal and available here.

Objectives

The objectives of the study in collaboration with ECI:

1. To establish the ability of individual and combined parameters in the CriSTAL tool to predict imminent death [imminent death defined as occurring within 90 days of assessment/admission]

2. To determine the minimum number of variables that are sufficient to predict in-hospital death

3. To develop a weighted score and set the thresholds of certainty for clinicians diagnosing the dying elderly

Project Overview

During Phase 1 the CriSTAL tool was validated retrospectively in a sample of hospitalised patients aged 60 years and above who had been the subject of a Medical Emergency Team (MET) call. The single centre study found that a patient profile of old age, frailty, cognitive impairment and chronic disease was a good statistical predictor of in-hospital death and should be used as a trigger to hold end-of-life conversations rather than calling the MET. 

Manuscript published in Resuscitation and available here

A retrospective validation was also conducted in a US teaching hospital to examine whether using CriSTAL on admission would have identified patients with palliative care needs earlier. The manuscript has been submitted to the International Journal of Nursing Studies.

Three Dutch hospitals have also collected CriSTAL parameters on 300 patients subject of a MET call. Analysis is underway.

During Phase 2 the CriSTAL tool was administered prospectively in emergency departments to over 2,870 patients aged 65 years and older in 10 hospital in Australia, Denmark and Ireland. The Sydney cohort study data has been analysed and found the predictive accuracy of CriSTAL to be high and the total number of items in the tool can be reduced. The manuscript will be submitted to a peer reviewed journal in February 2017.

The European component of the validation completed data collection in November 2016 and data analysis is underway, with a manuscript planned for submission in June 2017.

© Agency for Clinical Innovation 2017

Send us feedback on the site