Using the electronic frailty index (eFI) to redefine deprescribing in Scotland

Background to the project

Frailty is a critical concept and a key contributor to overall global ageing and impacts a large global market with high public health significance: Frailty is a multifactorial syndrome defined by an increased vulnerability to stressors that is driven by poor resolution of homeostasis. This leads to an elevated risk of developing a range of adverse events, including a fall, hospitalization or rapid changes in healthcare needs that result in increased healthcare costs. The NHS spends £5.6Bn extra per annum on the admission of UK frail individuals in comparison to age matched non-frail patients1. Within Scotland, in 2019 alone £1.7Bn was spent on unplanned bed days for the frail population. In the context of an ageing population and finite health and social care resources, the early identification of individuals at a high-risk of frailty is crucial to maintaining their wellbeing, dignity, and fulfilment in later life. This is also vital with respect to patient care, as understanding an individual’s ability to respond and recover from clinical treatments is critical to inform the selection of appropriate healthcare interventions and maximize post-operative recovery and healthspan.

The electronic Frailty Index (eFI) has been developed to identify frailty using 35 items of routine data held on primary care databases – such data is available within general practice settings of NHS Scotland. Pilot studies on 900,000 individuals in England have demonstrated that the eFI is simple and quick to use, acceptable to practice staff, and is able to discriminate older patients referred for comprehensive geriatric assessment from the total practice population. This evidences that the eFI is a valuable tool in primary care, can identify patients living with frailty early on and contribute to a reduction in the unplanned admissions register.

But in Scotland the eFI has not been adopted and implemented in the same way as in England due to several reasons. Firstly, platform visibility, secondly General Practice constraints worsening post Covid 19 and concerns about validity in a population out with academic clinical practice. Whilst there are well described capacity constraints within General Practice there is an emerging Pharmacist workforce within primary care which could support more usage of eFI to support optimal prescribing. We know this population often has significant polypharmacy and associated harms. 

Research Aims

The student will be embedded within a multidisciplinary team of scientists, consultants, and general practitioners to develop a thorough understanding of how the electronic frailty index is implemented within the NHS Scotland.

The project will cover the following areas:

1)          Developing a DELPHI study on eFI use in Scotland

Working with NHS Lanarkshire colleagues to determine GP practices that use the eFI

Planning a multilayer questionnaire on eFI where questions for each round are based in part of the findings of the previous one, allowing the study to evolve over time in response to earlier findings. 

2)          Develop extraction, handling, and harmonization approaches to eFI data.

Identify and connect to a small number of GP surgeries using the eFI 

Compare data access protocols and data formats for eFI and prescription data

Assess the differences and determine a more streamlined method of data annotation

3)          Map eFI data to prescribing data develop a deprescribing model

Model eFI data across practices using regression-based tools with the aim of identify potential routes to standardization. Use literature and multivariate and regression based tools to determine the primary contributors to polypharmacy and deduce the drugs most likely to be utilized in deprescribing. Development of a joint LASSO-regression/Deep learning model that defines the relationship axis between eFI score, polypharmacy burden and benefit of deprescribing.

Training outcomes: Questionnaire Generation, create DELPHI study protocol, training in the use of NHS Scotland data entry systems, understanding in eFI modelling, understanding in deprescribing patterns, use of regression based tools, use of consensus deep learning based tools, coding in R and Python.

Details of Supervision

The successful applicant will work across active research teams within the Strathclyde Institute of Pharmacy and Biomedical Sciences (SIPBS) and clinicians within NHS Lanarkshire.

Lead Supervisor: Dr Nicholas JW Rattray, Senior Lecturer in Clinical Metabolism. He holds a Royal Society Industry Fellowship in surgical frailty biomarkers and is an experienced Masters (n~20), doctoral (n=10) and post-doctoral supervisor.

Co-supervisors: Dr Zahra Rattray, Senior Lecturer in Translational Pharmaceutics and qualified pharmacist with experience in applied health research projects and PPIE involving NHS Trusts and the pahrmaceutical industry.

Ms Ana Talbot is a consultant in older peoples medicine within NHS Lanarkshire and is a Innovation Fellow West of Scotland focusing on developing models and technology that can reduce harm to frail individuals.

The successful candidates will benefit from a Postgraduate Certificate in Academic Practice from the University of Strathclyde and will work and interact directly with the staff and students in three highly productive and multidisciplinary teams across SIPBS and NHS Lanarkshire. The University of Strathclyde is currently ranked second for Pharmacy and Pharmacology in the UK Complete University Guide.