How can practices fulfil the new frailty requirements?
Pulse answers your questions on the new contractual frailty work starting from 1 July
From 1 July, GP practices in England have a new contractual requirement to identify elderly patients with moderate and severe frailty in a systematic way, and carry out regular clinical reviews in those who are severely frail. The work is funded with £157m previously available to practices for carrying out the Avoiding Unplanned Admissions enhanced service, terminated at the end of the 2016/17 contract.
NHS England says getting GPs to identify their frail elderly in a systematic way will help them to ‘stratify populations by risk of future health and care use’ and ‘target and tailor appropriate care and support for older people with the greatest need’.
Here we summarise what practices need to do to make sure they fulfil the new obligations.
What are the key requirements?
GP practices need to identify patients aged 65 and over who are living with moderate or severe frailty and record the diagnosis in their patient record. For those with severe frailty, they must also record that they have completed regular clinical reviews. This includes an annual medication review and – when clinically appropriate – asking if the patient has fallen in the past 12 months, as well as providing any clinically relevant interventions as needed.
Practices will also need to show they have explained the benefits of the enriched Summary Care Record (SCR) to patients living with severe frailty and seek their consent to activate it.
How will the work be monitored?
NHS England will collect data on the number of patients:
- with a diagnosis of moderate frailty
- with a diagnosis of severe frailty
- with severe frailty who have a record of annual medication review
- with severe frailty who are recorded as having had a fall in the preceding 12 months
- with severe frailty who have provided consent to activate their enriched SCR.
GP practices should use read codes provided by NHS Employers in technical guidance to record moderate frailty, severe frailty, medication reviews and consent to the enriched SCR. The guidance also lists over 50 available falls read codes, but the GPC has said practices can record falls with their existing codes.
NHS England says the data collected will not be used for performance management or benchmarking practices against one another. It says the data will be used to understand what interventions are made – such as referrals to falls clinics – and the prevalence of frailty by degree among practice populations and nationally.
However, NHS Employers has warned that failure to fulfil the requirements could result in a breach of contract notice.
How should we identify patients with ‘moderate’ and ‘severe’ frailty?
Practices should use the Electronic Frailty Index (eFI) to identify the two target groups of patients. The eFI uses data routinely collected in the Electronic Health Record and is the most accurate available tool for distinguishing between degrees of frailty and predicting nursing home admission, hospitalisation and mortality.
The eFI covers 36 potential clinical ‘deficits’ – variables including dizziness, hearing or visual impairment, cognitive impairment and mobility problems – and generates a cumulative frailty score.
The index was developed by researchers at Leeds University; their paper published in 2016 in the journal Age and Ageing describes how the eFI was developed and validated using data from over 900,000 UK primary care patients.
Can we perform automated searches with the eFI?
Yes. The tool searches existing data in the patient record for over 2,000 read codes to automatically identify patients with moderate and severe frailty. The tool is currently available to the estimated 99% of GP practices that are using SystmOne, EMISweb or Vision software.
Should we automatically record the diagnosis based on the eFI?
No. NHS England says practices cannot rely on the eFI tool alone because no tool is 100% sensitive or specific and therefore GPs need to review each patient’s eFI results to ‘prevent false negatives and false positives’.
According to GPC, this can be done opportunistically. Practices should set up their computer system to send a ‘pop-up’ alert during a consultation to flag up that the patient has been identified by the eFI as potentially frail – it is for the GP to then use their clinical judgement as to whether the diagnosis is appropriate and requires coding, and if so whether to do any necessary review as part of that clinical contact.
What if my practice’s computer system does not support the eFI?
There are no other automated tools available currently. GPs can use another approach such as the clinical frailty scale and clinical judgment to identify patients with frailty opportunistically.
How many patients do we need to identify?
GPC says there is no target number and practices are not expected to have recorded a specific proportion of their eligible population.However, as a guide, the eFI tool will categorise roughly 12% of your elderly population as ‘moderately’ frail and 3% as ‘severely’ frail. An average practice will have 117-156 patients with moderate frailty, or around 39 per FTE GP, and 21-36 with severe frailty, or nine per FTE GP.
How often should we re-run the eFI to update the records?
Both GPC and NHS England have advised that it is up to GP practices how often they run the frailty assessment tool. However, NHS England has said that the ‘minimum requirement’ will be for GP practices to do ‘a yearly frailty assessment in all patients aged 65 years and over’.
Practices and CCGs will receive quarterly CQRS reports on which of their eligible patients have had their assessment and, where appropriate, a frailty score recorded. NHS England says this will ‘help practices identify the patients who are yet to receive their assessment and help practices manage their workflow, in order to deliver the contractual requirement’.
3. Clegg A, et al. Development and validation of an electronic frailty index using routine primary care electronic health record data. Age Ageing 2016; 45 (3): 353-360