Project Team
Dr Lisa Gibbons, Primary Care Lead and GP, NIHR Clinical Research Network: South West Peninsula
Dr Anna Turner, GP, SW Peninsula Clinical Research Fellow
Dr Daisy Robinson, GP, SW Peninsula Clinical Research Fellow, Education Lead, Devon Training Hub
Dr Hannah Oram, GP
Professor Richard Byng, Deputy Director of PenARC and Complex Care Theme Lead
Richard Blackwell, Associate Director of Insight, SW Peninsula Academic Health Science Network
Background
People with ‘complex lives’, are identified by clinicians as those whose lives are often characterised by multiple health and social issues, the prescription of multiple psychotropic and analgesic drugs and frequent attendence at healthcare settings. However, there can be challenges in easily identifying these patients from GP records.
A novel data driven approach has been developed using attendance and prescribing data as opposed to clinical codes to identify a manageable sized group in practices. The majority of the patients in this group could be seen to have ‘complex lives’ but are ‘hidden in plain sight’. Qualitative note review identified consistent themes such as social adversity, chronic pain and mental health problems, many of which were not coded. This cohort was found to have a higher use of urgent appointments and often consulted multiple GPs.
This data-driven approach allows meaningful comparison of similar sized cohorts across different practices by analysing equivalent but not the same data from each practice. No data extraction is required, rather search techniques and search files are shared. This would not be possible in the same way through analysis of amalgamated population health level data which requires data extraction and re-matching and would use arbitrary cut offs for values as opposed to practice centiles. This technique allows the identification and visualization of a previously known but hidden population of patients with ‘complex lives’.
Aim
The aim of this project is to use this replicable approach to identify a group of patients from practices within a primary care network (PCN) who may benefit from a multi-disciplinary approach, social and prescribing interventions and well-being support.
It is hoped that the approach can be replicated across GP surgeries (for example in a PCN) to identify a manageable sized group of patients who attend frequently, have multiple psychotropic prescriptions and characteristically have complex social factors, chronic pain and mental health problems; a group not classically identifiable by using a code search approach.
Our aim is to use this active identification approach to allow interventions to reduce psychotropic prescribing and appointment use and offer opportunities to fulfil unmet patient needs within this group.
Impact
The data extraction and analysis for this project required the support of a data analyst. To facilitate the dissemination and replication of this approach across healthcare settings we have developed a simplified technique using SystmOne searches which exports the outputs into Excel spreadsheets allowing detailed analysis of the data.
Files are available with the SystmOne search files for appointment data and prescribing data. Instructions are included to assist in analysing the data from these searches to allow the cohorts to be identified.
Resources
Read the project report: Hidden in plain sight: Identification of people with ‘complex lives’ in General Practice using a data-driven approach
Use the Hidden in Plain Sight Tool
You can use the detailed instructions below to use the data analysis tool developed as part of this project. Where the instructions have referenced particular files these have been included below for your ease.
Hidden in Plain Sight Instructions (EMIS Web)
Hidden in Plain Sight Instructions (SystmOne)
Contact us
The authors are keen to hear from anyone who would like to know more. Please contact Professor Richard Byng if you would like to use this method in your practice or PCN, or contact Richard Blackwell (South West AHSN) if you are based in Cornwall, Devon or Somerset and need assistance in implementing these data searches. The project team are happy to create an online discussion forum or group to share experiences and publish future evaluations.
*