Financial Modelling and service design
Research team: Simon Dixon, Alan Brennan Praveen Thokala, Pete Dodd, Kinga Lowrie from Health Economics and Decision Science, School of Health and Related Research, University of Sheffield
The aim of this workstream was to develop a financial model for telehealth in two clinical populations – COPD and CHF – to allow current and potential telehealth services to be tested for cost-effectiveness at different scales and under difference service configurations.
Qualitative interviews with key telehealth stakeholders (operational and strategic managers, commissioners, industry) were completed across the four health service sites to feed into the design of the financial model. This was used to inform a model using a variety of clinical and outcome data - based on analysis of the Hospital Episodes Statistics, a meta analysis of the clinical effectiveness of telehealth and survey data on telehealth user quality of life. The model was subsequently tested and refined with input from two research sites and a series of other stakeholders.
- Telehealth can be effective in reducing mortality, hospitalisations and improving health, but the existing evidence of effectiveness is variable.
- Telehealth needs system change in order to be delivered effectively and this needs to be supported by an implementation plan, not just a final service model. Telehealth is about more than boxes in houses; the way in which staff, organisations and patients behave and interact all need to change. Changes in staff mix, roles or referral pathways are also needed; and what changes, when they happen, and how much it will cost to change, are all important.
- Whether telehealth is cost-effective depends on implementation, as well as the final service model. Implementation costs are part of the decision about whether to adopt telehealth, and implementation also impacts on the effectiveness of the final service model.
- Understanding the changes over time is essential, not just the steady state costs of the final service model. In addition, implementing service change is not immediate, nor are the health changes.
- Knowing the purpose of your service is essential as this drives key parameters. These include, which patients are eligible and how long will they use telehealth for; is this a temporary deployment for teaching self-management, or long-term support for the highly vulnerable; is this for all patients, or ‘frequent fliers’ (people frequently admitted to hospital); and is this ‘one size fits all’ or will multiple systems be utilised in the service?
- Mapping out stakeholders, together with their financial and operational relationships to one another is key to documenting and understanding the complexity of service models. It is useful to consider who does what, with whom, for how much and depending on what within the service model?
- Data is important – gathering appropriate data to monitor and evaluate service delivery and performance takes away a lot of guesswork. Service delivery includes number of patients, drop-outs, duration of monitoring and redeployment time. Performance includes hospitalisations, nurse visits and alerts.
- Informed contracting is only possible with sufficient data, and a good understanding of implementation, the system and the final service model. Knowing whether it is worth it is essential, as too, is knowing why it’s not worth it. The ‘whys’ may be solved by an alternative contracting arrangement that can be discussed with other stakeholders.