Health care in England faces increasing demands due to ageing populations and people with chronic conditions living longer. Missing waiting time standards in hospitals signals this growing pressure. The demand for secondary care is controlled by gatekeeper general practices. Higher quality and quantity of primary care facilities can help reduce the pressure on secondary and tertiary care services. There were several reforms to improve the quality of practices such as Quality and Outcome Framework in order to control chronic conditions at primary cares. In addition, increasing the supply of GPs was considered to reduce both elective and emergency hospital admissions, with the two main UK political parties in the 2015 general election both proposing an increase in the supply of GPs. However, the growth in GP supply doesn't cope with population growth. As a result, the GP workload is increasing over time. The GP work-life survey shows a constant decrease in job satisfaction in terms of working conditions and responsibilities.
The focus of this work is on the effect of this increasing workload of GPs working in London on the health outcome of patients and namely the choice of diagnostic inputs by GPs. Importantly, one of the main diagnostic inputs is referral to secondary care which has many crucial consequences. On one hand, accurate referral is necessary to diagnose medical problems; on the other, unnecessary referrals should be controlled by the general practitioner to alleviate the pressure on hospitals.
The aim is to use CPRD to identify a measure of workload and then identify the effect of workload on the health outcomes; namely, the probability of referral. We focus on London to minimise the confounding effect of geographical differences in the access to health care and underlying health conditions among patients. We will look at cross-sectional variation of workload across practices and within each practice.
Our research will consist of the following steps:
1. Constructing a measure for daily workload at practice level. This is the average patient consultation to GP ratio.
2. Identifying the fluctuations of the daily workload during the year within a practice.
3. We use the unexpected fluctuations in the workload pressure through unexpected absence of a GP to identify an exogeneous and random variation in the workload pressure. Absence is measured if the doctor was working in the practice in a 2-week window period but has zero consultation in a given day
4. Then we will examine whether there is any association between this exogeneous variation in the workload and the choice of diagnostic inputs. We will focus on a linear probability regression for the referral to specialist, but we will also include referral to test labs.
5. The analysis will be expanded to explain how the workload of other staffs, namely practice nurses varies and how it helps reducing the workload of GPs.
The analysis will be done using 2 stage least square analysis for two years.
GP daily workload, nurse daily workload, practice average workload, referral probability by the urgency of referral (2 week wait, urgent, soon, routine, dated), referral to test laboratories.
Toby Watt - Chief Investigator - The Health Foundation
Andrew Campbell - Corresponding Applicant - The Health Foundation
Hanifa Pilvar - Collaborator - Queen Mary University of London
Patient Level Index of Multiple Deprivation;Practice Level Index of Multiple Deprivation