Advances and Applications in Statistics
Volume 6, Issue 3, Pages 335 - 351
(December 2006)
|
|
MODELLING HOSPITAL ACTIVITY: ACCOUNTING FOR SMALL AREA AND PRIMARY CARE PRACTICE VARIATION
Peter Congdon (United Kingdom)
|
Abstract: Management of care in appropriate settings - for example, avoiding unnecessary attendances at hospital A&E units that could be handled in primary care - is an important part of health strategy. Hence methods are required to identify sources of excess attendances both geographically and in terms of primary care practices responsible for patients. This paper considers a Bayesian random effects approach to modelling small area and GP practice variation in hospital attendances or admissions with a view to detecting outlier areas or practices with unduly high rates. The model allows for the impact on small area attendance rates of deprivation and geographic access (to both primary care and to hospitals) and also for the interplay between small area health demand and the population distribution between GP practices. The case study involves a six month survey of A&E attendances at a North London Hospital by residents in a London Borough. It considers relativities in attendance rates between 149 small areas in this borough in relation to area deprivation scores and differential geographic access to GP surgeries, while also allowing for variation in attendance rates across the 53 GP practices in the case study area. |
Keywords and phrases: geographic access, primary care, random effects, spatial effects, practice variation. |
|
Number of Downloads: 362 | Number of Views: 1173 |
|