Measuring and modelling the 'spillover' effects of healthcare interventions
Lead Research Organisation:
University of Birmingham
Department Name: Health and Population Sciences
Abstract
Decisions about the funding of new healthcare treatments depend on the health benefits of the treatment. In decision-making it is common to focus only on benefits to the patients, however this neglects any knock on impact that treatments may have on the lives of those close to patients. Improvements in a patient‘s health may, for example, relieve family members of stressful caring tasks and reduce their anxiety about their loved one‘s health.
The proposed research will study whether poor patient health has a knock-on impact on unpaid carers (normally family members). Data from studies of carers of patients with Parkinson‘s disease and carers of children who have survived meningitis will be used investigate any link between patients‘ health and carers‘ health and wellbeing. Research will be carried out to investigate whether it is caring per se or whether it is the poor health of a loved one that affects carers‘ health the most. Ultimately the research findings will be used to develop a way of predicting the likely knock-on health impacts on carers in different situations so that these can be considered in healthcare funding decisions.
The proposed research will study whether poor patient health has a knock-on impact on unpaid carers (normally family members). Data from studies of carers of patients with Parkinson‘s disease and carers of children who have survived meningitis will be used investigate any link between patients‘ health and carers‘ health and wellbeing. Research will be carried out to investigate whether it is caring per se or whether it is the poor health of a loved one that affects carers‘ health the most. Ultimately the research findings will be used to develop a way of predicting the likely knock-on health impacts on carers in different situations so that these can be considered in healthcare funding decisions.
Technical Summary
Aim
To estimate the ‘spillover‘ impact of the patient‘s health condition on unpaid carers and to develop econometric models to forecast these impacts for use in economic evaluation.
Objectives
Phase 1: Validation of outcome instruments to measure the spillover impact in carers.
Phase 2: Measurement of the extent of spillover impacts on carers using trial and survey data.
Phase 3: Disentangling the spillover impact to distinguish between the ‘caring‘ and ‘family‘ effect.
Phase 4: Econometric modelling of spillover impacts for future use in economic evaluation.
Design/Methodology
Phase 1: Tests of the validity and sensitivity of a new measure of ‘care-related‘ quality of life (the Carer Experience Scale) will be conducted using existing survey data from the UK, Australia and the Netherlands and data from a trial of drugs for Parkinson‘s disease (PD trial).
Phase 2: Longitudinal data on outcomes for carers in the PD trial and carers of meningitis survivors (MEN study) will be collected using standardised instruments. The data will be used to explore how spillover impacts differ by illness, patient health state and hours of care provided. The analysis will be extended, where possible, by incorporating carer outcome data from other studies and outcome data for non-caring family members.
Phase 3: Data from the MEN study will be used to identify the degree to which carer health is affected by the caring tasks undertaken as opposed to the patient‘s health condition. These ‘caring‘ and ‘family‘ effects will be estimated using regression models with the carer‘s health (and wellbeing in the second model) as the dependent variable and patient health, the number of caring tasks and other contextual characteristics of the caring situation as independent variables.
Phase 4: Econometric modelling will be used to develop predictive models of the spillover effects of poor patient health. Data will be synthesised from different sources, where feasible, to estimate regression models with the carer‘s quality-adjusted life year (QALY) change as the dependent variable and a range of variables such as patient QALY change, hours of care and nature of patient disease as independent variables.
Scientific and medical opportunities
By identifying spillover effects and developing regression models to forecast these, this research will promote: (i) improved adherence to economic evaluation guidelines and (ii) decisions that come closer to maximising health benefits from available resources. This work will further explore the potential for using econometrics to estimate spillover impacts for economic evaluation.
To estimate the ‘spillover‘ impact of the patient‘s health condition on unpaid carers and to develop econometric models to forecast these impacts for use in economic evaluation.
Objectives
Phase 1: Validation of outcome instruments to measure the spillover impact in carers.
Phase 2: Measurement of the extent of spillover impacts on carers using trial and survey data.
Phase 3: Disentangling the spillover impact to distinguish between the ‘caring‘ and ‘family‘ effect.
Phase 4: Econometric modelling of spillover impacts for future use in economic evaluation.
Design/Methodology
Phase 1: Tests of the validity and sensitivity of a new measure of ‘care-related‘ quality of life (the Carer Experience Scale) will be conducted using existing survey data from the UK, Australia and the Netherlands and data from a trial of drugs for Parkinson‘s disease (PD trial).
Phase 2: Longitudinal data on outcomes for carers in the PD trial and carers of meningitis survivors (MEN study) will be collected using standardised instruments. The data will be used to explore how spillover impacts differ by illness, patient health state and hours of care provided. The analysis will be extended, where possible, by incorporating carer outcome data from other studies and outcome data for non-caring family members.
Phase 3: Data from the MEN study will be used to identify the degree to which carer health is affected by the caring tasks undertaken as opposed to the patient‘s health condition. These ‘caring‘ and ‘family‘ effects will be estimated using regression models with the carer‘s health (and wellbeing in the second model) as the dependent variable and patient health, the number of caring tasks and other contextual characteristics of the caring situation as independent variables.
Phase 4: Econometric modelling will be used to develop predictive models of the spillover effects of poor patient health. Data will be synthesised from different sources, where feasible, to estimate regression models with the carer‘s quality-adjusted life year (QALY) change as the dependent variable and a range of variables such as patient QALY change, hours of care and nature of patient disease as independent variables.
Scientific and medical opportunities
By identifying spillover effects and developing regression models to forecast these, this research will promote: (i) improved adherence to economic evaluation guidelines and (ii) decisions that come closer to maximising health benefits from available resources. This work will further explore the potential for using econometrics to estimate spillover impacts for economic evaluation.