The third step is to determine how best to use the resultant ‘impact’ information to identify the priority that should be given to one intervention rather than another. This analysis usually requires additional information (e.g. about the cost of interventions) to be combined with impact information. It involves finding a ‘solution’ (sometimes called a decision rule) that uses impact (outcome) and other information to specify how much each intervention should be resourced. For example, NICE uses impact information estimated using RCTs that measure patients’ health-related quality of life (e.g. EQ-5D) and applies a decision rule that requires that (incremental) cost-effectiveness of the intervention exceeds a value-for-money threshold.
The application theme is concerned with these second and third analytical steps. In particular, it focuses on how to use existing datasets of relevant outcome information to assess impact. These datasets include survey data (e.g. the Adult Social Care Survey or the GP patients’ survey) and administrative datasets. We do not intend to carry out primary research using bespoke methods such as randomised controlled trials.
Aims and methods
This project considers methods that could be used to measure the impact and then apply these methods to estimate the impact of services and activities using existing data. The general approach is be to draw on the range of survey datasets that collect outcomes-relevant information using standardised health- and social care-related quality of life measures. Our method is to link these survey data with service utilisation records, remuneration and regulation data. Not only do these datasets provide relevant real-world data for assessing impact and other performance relevant indicators, but they also are datasets that are being used by decision-makers and practitioners in the field.
Strand 1: Conceptual development
We will develop and describe the approach of using outcomes information can be used to guide decision-makers. In particular, we will describe the three analytical steps as outlined above and their conceptual underpinning. The aim is contrast the outcomes approach with other methods to allocate social and health care resources.
Strand 2: Applying an outcomes approach
This work involves a number of applications.
Strand 2a: the role of the provider and its workforce on variations in the outcomes impact of home care from the service user’s perspective
The first is to assess the effect of differences in the characteristics of home care providers and their workforce on service users’ outcomes.
In this project we aim to assess the impact of different home care provider types as differentiated by: their organisational type; their assessed CQC quality ratings; and, crucially, the characteristics of their workforce. The methods are as follows. First, we will link data from the home care User Experience Survey (UES) (2008/9), the national minimum dataset for social care workforce (NMDS-SC), and CQC quality ratings data. Second, we will use multiple regression techniques to control for service user need characteristics and in this way attribute remaining changes in SCRQoL to service use according to the intensity of home care packages used. Moreover, we will use these models to assess how provider and workforce characteristics affect this service impact.
Strand 2b: impact of informal care on health status and mortality
A second application is in assessing the whether the provision of informal care has an impact on that person’s health status and mortality. Using secondary data, we aim to conduct analysis to attribute observed changes in health status and mortality (in Census data) to the provision of informal care.
Other applications can also be considered, for example looking at the impact of quality frameworks on quality of life outcomes. Examples of relevant quality frameworks include CQC compliance ratings of social care providers and the GP quality and outcomes framework (QOF). Both individual-level and small-area datasets could be used for these analyses.
Strand 3: Making better use of survey and administrative data sets to improve outcomes
As a longer-term project we will investigate the value of using alternative datasets with outcomes-relevant information. The aim is to: (a) improve the methods for attributing impact and effect (in outcome terms) to the actions and activities of health and social care; and (b) using these results to develop the evidence base about which of these actions and activities best (e.g. most cost-effectively) produce desired outcomes.
We will explore the possibility of developing links with local authorities with the aim of obtaining and analysing user record data, including indicators of need, service receipt and outcome. We will contact English authorities – potentially via the Information Centre for health and social care – about providing outcomes information in a form to best support attribution analyses. For example, we intend to investigate the possibility of getting anonymous outcomes data at a finer level of aggregation (e.g. by ‘small’ area). We will also link this project with our related project concerning the impact of adult social care (IIASC).
Timing and outputs
The outputs for policy are:
- Providing tools to decision-makers (e.g. commissioners) to improve the outcomes-consequences of their activities
- Adding to the evidence base about the effectiveness and cost-effectiveness of services and support
Footnote 1: We can contrast person-centred goals and approaches with other approaches, such as prioritisation on the basis of professional opinion (rather than patient outcomes directly) or political preferences. The latter may serve person-centred outcomes, but they might also serve other interests such as prestige or profit, or professional sustainability.