The University of Southampton
University of Southampton Institutional Repository

Do higher primary care practice performance scores predict lower rates of emergency admissions for persons with serious mental illness? An analysis of secondary panel data

Do higher primary care practice performance scores predict lower rates of emergency admissions for persons with serious mental illness? An analysis of secondary panel data
Do higher primary care practice performance scores predict lower rates of emergency admissions for persons with serious mental illness? An analysis of secondary panel data
Design and theoretical/conceptual framework: Effective primary care can have an important preventive role, and should therefore be associated with lower emergency admission rates. Quality indicators for mental health have been routinely measured in English primary care over a number of years as part of the Quality and Outcomes Framework (QOF). Our null hypothesis is that there is no association between QOF performance and emergency hospital admissions for people with mental illness either for mental or physical conditions. We will test for an association between changes in practices' mental health QOF indicators and changes in their rates of emergency admission using data on all practices in England over the period 2004/05 to 2009/10. Our analysis will also estimate the impact of potential improvements in the QOF on subsequent mental health expenditure on secondary care. Sampling: We will construct a national dataset covering around 8000 English GP practices by drawing together routinely available secondary data. To examine the impact of the QOF on hospital expenditure, we will merge this national dataset with the costs of all 1.2 million patients who use specialist psychiatric hospital or community care in a given year. Our analysis will therefore be representative and produce generalisable results. Setting/context: We will use Hospital Episode Statistics (HES) data on emergency admissions for mental health patients from general practices for both physical and mental conditions over the period 2001/02 to 2009/10 in England. QOF data is available from 2004/05 to 2009/10. We will merge a number of data sources at practice level to create a panel which will provide statistical power and precision to the econometric analysis. To examine the relationship between the QOF and subsequent mental health expenditure we will use individual level cost data in the Mental Health Minimum Dataset (MHMDS) which has been derived from Reference Costs and is available for 2007/08 and 2008/09 covering care by specialist psychiatric teams in hospital or in the community. Data collection: We will use the following data: QOF Indicators: The Quality Management Analysis System (QMAS) provides QOF achievement and prevalence data at practice level. Some QOF indicators have remained constant, others have been modified, dropped or introduced. This represents a 'natural experiment' that allows variations in admissions between practices to be investigated. Covariates: Local population characteristics: Neighbourhood Statistics (ONS) socio-economic and demographic data will be attributed to GP practices using the Attribution Data Set which contains information on the number of patients in each practice resident in each Lower Super Output Area. The socio-economic data are very rich and include measures of deprivation, education, morbidity, ethnicity, rurality and small area characteristics. GP Practice variables: General Medical Statistics (GMS) data on GP and practice characteristics, and MHMDS data which will be aggregated to practice level. Mental Health Services Mapping Data will be used to construct supply variables. Hospital variables: hospital characteristics and quality indicators. Data analysis: We will estimate both cross-sectional and panel data models for the period 2004/05 to 2009/10 to examine the association between QOF and admissions over time. Examining the within practice temporal variation will remove the risk of unobserved factors which might affect both practice emergency admissions and quality. We will estimate both random and fixed effects multiple regression and count data models. All models will include year indicators to allow for temporal trends, and a rich set of relevant local population and practice covariates [1]. Random effects panel data models will include the average admissions for a practice over the period 2001/02 to 2003/04. This pre-sample 'baseline' will pick up unobserved practice and patient confounding characteristics which are time-invariant [2]. We will also include lags of QOF to allow for delayed effects of quality. We will carry out a variety of robustness checks. For ease of presentation the scores for all variables will be reported as incidence rate ratios (IRR). Our models for examining the association between QOF and subsequent hospital costs will include both OLS cross-sectional and random and fixed effects panel data models where we control for practice fixed effects and year dummies. We will regress costs in 2007/08 and 2008/09 on QOF scores from 2004/05 up to 2008/09. Given the positive skewness of costs we will also estimate transformed OLS and generalised linear models (GLM). We will use two-part models to test if better quality scores have an effect on average patients by reducing the probability of admission and/or by reducing costs once admitted. Possible lagged quality effects on costs will be captured by modelling 'baseline' quality from 2004/05 to 2006/07 on patient expenditure in 2007/08 and 2008/09. [1] Dusheiko M, Doran T, Gravelle H, Fullwood C, Roland M. Does higher quality of diabetes management in family practice reduce unplanned hospital admissions? Health Services Research 2011; 46:27-46. [2] Blundell R, Griffith R, Windmeijer F. Individual Effects and Dynamics in Count Data Models. Journal of Econometrics 2002; 108:113-31.
Health Services and Delivery Research
Jacobs, R.
6011817b-2d0a-4dbf-94f3-ac523a51b8e9
Gutacker, N.
483d818b-e6cf-4c46-8d54-b7ed795903fe
Mason, A.
0cfc2ff4-70c2-4d1e-ad12-6b90ffe60b3f
Goddard, M.
bd52d741-1b9a-4b68-969b-a0c84526063a
Gravelle, H.
4ec9e07d-24f5-43ad-818e-db1df8c6fe2c
Kendrick, A.
c697a72c-c698-469d-8ac2-f00df40583e5
Gilbody, S.
9de3029e-dc51-46c8-9938-868a501aa3d9
Aylott, L.
9d7b85e4-ff16-4ce8-962a-253e0b48a0f2
Wainwright, J.
b1469599-d09e-483d-af82-805e680c83a5
Jacobs, R.
6011817b-2d0a-4dbf-94f3-ac523a51b8e9
Gutacker, N.
483d818b-e6cf-4c46-8d54-b7ed795903fe
Mason, A.
0cfc2ff4-70c2-4d1e-ad12-6b90ffe60b3f
Goddard, M.
bd52d741-1b9a-4b68-969b-a0c84526063a
Gravelle, H.
4ec9e07d-24f5-43ad-818e-db1df8c6fe2c
Kendrick, A.
c697a72c-c698-469d-8ac2-f00df40583e5
Gilbody, S.
9de3029e-dc51-46c8-9938-868a501aa3d9
Aylott, L.
9d7b85e4-ff16-4ce8-962a-253e0b48a0f2
Wainwright, J.
b1469599-d09e-483d-af82-805e680c83a5

Jacobs, R., Gutacker, N., Mason, A., Goddard, M., Gravelle, H., Kendrick, A., Gilbody, S., Aylott, L. and Wainwright, J. (2014) Do higher primary care practice performance scores predict lower rates of emergency admissions for persons with serious mental illness? An analysis of secondary panel data.

Record type: Other

Abstract

Design and theoretical/conceptual framework: Effective primary care can have an important preventive role, and should therefore be associated with lower emergency admission rates. Quality indicators for mental health have been routinely measured in English primary care over a number of years as part of the Quality and Outcomes Framework (QOF). Our null hypothesis is that there is no association between QOF performance and emergency hospital admissions for people with mental illness either for mental or physical conditions. We will test for an association between changes in practices' mental health QOF indicators and changes in their rates of emergency admission using data on all practices in England over the period 2004/05 to 2009/10. Our analysis will also estimate the impact of potential improvements in the QOF on subsequent mental health expenditure on secondary care. Sampling: We will construct a national dataset covering around 8000 English GP practices by drawing together routinely available secondary data. To examine the impact of the QOF on hospital expenditure, we will merge this national dataset with the costs of all 1.2 million patients who use specialist psychiatric hospital or community care in a given year. Our analysis will therefore be representative and produce generalisable results. Setting/context: We will use Hospital Episode Statistics (HES) data on emergency admissions for mental health patients from general practices for both physical and mental conditions over the period 2001/02 to 2009/10 in England. QOF data is available from 2004/05 to 2009/10. We will merge a number of data sources at practice level to create a panel which will provide statistical power and precision to the econometric analysis. To examine the relationship between the QOF and subsequent mental health expenditure we will use individual level cost data in the Mental Health Minimum Dataset (MHMDS) which has been derived from Reference Costs and is available for 2007/08 and 2008/09 covering care by specialist psychiatric teams in hospital or in the community. Data collection: We will use the following data: QOF Indicators: The Quality Management Analysis System (QMAS) provides QOF achievement and prevalence data at practice level. Some QOF indicators have remained constant, others have been modified, dropped or introduced. This represents a 'natural experiment' that allows variations in admissions between practices to be investigated. Covariates: Local population characteristics: Neighbourhood Statistics (ONS) socio-economic and demographic data will be attributed to GP practices using the Attribution Data Set which contains information on the number of patients in each practice resident in each Lower Super Output Area. The socio-economic data are very rich and include measures of deprivation, education, morbidity, ethnicity, rurality and small area characteristics. GP Practice variables: General Medical Statistics (GMS) data on GP and practice characteristics, and MHMDS data which will be aggregated to practice level. Mental Health Services Mapping Data will be used to construct supply variables. Hospital variables: hospital characteristics and quality indicators. Data analysis: We will estimate both cross-sectional and panel data models for the period 2004/05 to 2009/10 to examine the association between QOF and admissions over time. Examining the within practice temporal variation will remove the risk of unobserved factors which might affect both practice emergency admissions and quality. We will estimate both random and fixed effects multiple regression and count data models. All models will include year indicators to allow for temporal trends, and a rich set of relevant local population and practice covariates [1]. Random effects panel data models will include the average admissions for a practice over the period 2001/02 to 2003/04. This pre-sample 'baseline' will pick up unobserved practice and patient confounding characteristics which are time-invariant [2]. We will also include lags of QOF to allow for delayed effects of quality. We will carry out a variety of robustness checks. For ease of presentation the scores for all variables will be reported as incidence rate ratios (IRR). Our models for examining the association between QOF and subsequent hospital costs will include both OLS cross-sectional and random and fixed effects panel data models where we control for practice fixed effects and year dummies. We will regress costs in 2007/08 and 2008/09 on QOF scores from 2004/05 up to 2008/09. Given the positive skewness of costs we will also estimate transformed OLS and generalised linear models (GLM). We will use two-part models to test if better quality scores have an effect on average patients by reducing the probability of admission and/or by reducing costs once admitted. Possible lagged quality effects on costs will be captured by modelling 'baseline' quality from 2004/05 to 2006/07 on patient expenditure in 2007/08 and 2008/09. [1] Dusheiko M, Doran T, Gravelle H, Fullwood C, Roland M. Does higher quality of diabetes management in family practice reduce unplanned hospital admissions? Health Services Research 2011; 46:27-46. [2] Blundell R, Griffith R, Windmeijer F. Individual Effects and Dynamics in Count Data Models. Journal of Econometrics 2002; 108:113-31.

This record has no associated files available for download.

More information

Published date: 2014
Additional Information: Links to the project specification document, protocols and final report.
Organisations: Primary Care & Population Sciences

Identifiers

Local EPrints ID: 364977
URI: http://eprints.soton.ac.uk/id/eprint/364977
PURE UUID: d7422f41-a8e9-4371-b44b-83a10de919eb
ORCID for A. Kendrick: ORCID iD orcid.org/0000-0003-1618-9381

Catalogue record

Date deposited: 19 May 2014 13:43
Last modified: 16 Mar 2024 03:00

Export record

Contributors

Author: R. Jacobs
Author: N. Gutacker
Author: A. Mason
Author: M. Goddard
Author: H. Gravelle
Author: A. Kendrick ORCID iD
Author: S. Gilbody
Author: L. Aylott
Author: J. Wainwright

Download statistics

Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.

View more statistics

Atom RSS 1.0 RSS 2.0

Contact ePrints Soton: eprints@soton.ac.uk

ePrints Soton supports OAI 2.0 with a base URL of http://eprints.soton.ac.uk/cgi/oai2

This repository has been built using EPrints software, developed at the University of Southampton, but available to everyone to use.

We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we will assume that you are happy to receive cookies on the University of Southampton website.

×