An online frailty calculator developed by academics at The University of Nottingham will help GPs to identify their most vulnerable older patients for enhanced care. The QFrailty calculator combines complex information about a patient’s life expectancy and their risk of unplanned hospital admission over the coming 12 months.
The QFrailty calculator combines complex information about a patient’s life expectancy and their risk of unplanned hospital admission over the coming 12 months. These estimates are then used to classify patients within four frailty categories – severe, moderate, mild and fit – allowing doctors to target assessments and interventions at those most in need.
The tool was developed by Professors Julia Hippisley-Cox and Carol Coupland in the University’s School of Medicine using the QResearch database, a not-for-profit partnership with EMIS Health. It uses anonymised patient data from approximately 1,500 general practices across England that use EMIS Health’s clinical systems.
Research to test the accuracy of the new calculator, published in the British Medical Journal (BMJ), has shown that the tool can reliably estimate risk of dying within 12 months and risk of unplanned admissions among patients aged between 65 and 100 years old.
Professor Hippisley-Cox said: “Instead of creating a frailty index in the hope that it would predict unplanned admissions and life expectancy, we decided to work the other way round. Starting with principled estimators of unplanned admissions and life expectancy, we have developed a new classification system of frailty, known as QFrailty. This identifies people into four categories—severely frail, moderately frail, mildly frail, or fit—based on their absolute risks of an unplanned hospital admission or death within a year. QFrailty is therefore an outcomes-based classification designed to improve on current approaches”
Since July of this year, GPs have been contractually obliged by NHS England, the commissioning body for the English national health service, to identify those patients with moderate and severe frailty to ensure that services can be directed to those most in need. However, existing methods of calculating mortality risks are so inadequate that NICE (The National Institute for Health and Care Excellence) has been unable to endorse any of the 41 studies that had previously claimed to predict mortality, instead calling for more research to be undertaken. In response to this, the Nottingham researchers set out to overcome these poor performance issues by developing their own, more accurate model for prediction. They used anonymised information on 1.47 million patients aged 65 to 100 from 1,079 GP surgeries in England from 1 January 2012 to 30 September 2016 to identify factors which are important in determining the risk of imminent death in older patients.
In response to this, the Nottingham researchers set out to overcome these poor performance issues by developing their own, more accurate model for prediction. They used anonymised information on 1.47 million patients aged 65 to 100 from 1,079 GP surgeries in England from 1 January 2012 to 30 September 2016 to identify factors which are important in determining the risk of imminent death in older patients. This included their age, their body mass index and ethnic group, whether they smoke, how much alcohol they drink and how many emergency admissions to hospital they’d had in the previous 12 months. It also took into account their medical history and whether they suffered from a range of illnesses including cancer, asthma, epilepsy, heart disease, chronic liver disease, Parkinson’s disease and diabetes, as well as looking at medicines being used such as antipsychotics or anti-inflammatory medication.
They then used information from 500,000 different patients from a separate set of 357 practices to test the reliability of the new prediction calculator. They found that the new tool was very effective in predicting the short-term risk of death in men and, women.The researchers then went on to use this calculator in tandem with another prediction tool, the QAdmissions tool, which predicts a person’s risk of emergency hospital admission, to divide the patient groups into four broad frailty groups – severe frailty; moderate frailty, mild frailty and fit.
