Creation of a model of GP workload in terms of patient, practice and geographical variables to inform the GMS Contract Formula

Application Number
15_158
Lay Summary

The purpose of this research is to inform a review of the General Medical Services (GMS) contract formula in England. This resource is required to maintain essential services such as management of patients who are ill, believe themselves to be ill with conditions from which recovery is generally expected, and general management of patients who are terminally ill and management of chronic disease. The current contract includes a formula (the “Carr-Hill” formula) designed to reflect workload at practice level. This formula is now recognised to be out of date as the General Practice (GP) workload has changed in nature since it was created. This research will propose a new approach for assessing GP workload in terms of patient, practice and geographical characteristics. In a separate non-CPRD study, an assessment will be made of how the costs of delivering this workload vary around England. The combined workload and cost formula will inform both Clinical Commissioning Group (CCG) level primary care resource allocation and the resource allocation for each GMS practice.
The expectation is that the new formula will better allocate resources to CCGs/NHS England Regions to better meet need across primary care, benefitting practices and patients across the whole system.

Technical Summary

The purpose of this research is to inform a review of the General Medical Services (GMS) contract formula in England. The current contract includes a formula (the “Carr-Hill” formula) designed to reflect workload at practice level. This formula is now recognised to be out of date as the nature of General Practice (GP) workload has changed since this was developed. The proposed research will produce a new model measured either by total ‘case opening’ time or ‘number of file openings’ and will include age, sex, new registrations and deprivation, alongside practice characteristics such as QOF disease prevalence as explanatory variables in the new workload formula. Generalised linear models will be developed to model the relationship between case opening times and model parameters in a training dataset. The final model will be evaluated in a test and independent dataset. The GP workload formula will be combined with a formula for cost (developed in a separate non-CPRD study) to inform both CCG level primary care resource allocation and resource allocation for each GMS practice.
The expectation is that the new formula will better allocate resources to CCGs and practices to better meet need across primary care, benefitting patients across the whole system.

Collaborators

Lindsay Gardiner - Chief Investigator - NHS
Tarita Murray-Thomas - Corresponding Applicant - CPRD
David Vincent - Collaborator - Deloitte LLP
Hugh Gravelle - Collaborator - University of York
Katherine Everard - Collaborator - NHS
Samuel Beatson - Collaborator - NHS

Linkages

HES Admitted Patient Care;Patient Level Index of Multiple Deprivation;Practice Level Index of Multiple Deprivation;Other