The authors declare that they have no competing interests.
The paper was conceived by CA, NA, JC, AB, SJ and CM at an event organised and co-ordinated by SJ. All authors were present at at least one of the stakeholder events. CA led in writing the first draft, with substantial input from NA, and all authors contributed to the writing of the manuscript. All authors read and approved the final manuscript.
Decision Analytic Models (DAMs) are established means of evidence-synthesis to differentiate between health interventions. They have mainly been used to inform clinical decisions and health technology assessment at the national level, yet could also inform local health service planning. For this, a DAM must take into account the needs of the local population, but also the needs of those planning its services.
Drawing on our experiences from stakeholder consultations, where we presented the potential utility of a DAM for planning local health services for sexually transmitted infections (STIs) in the UK, and the evidence it could use to inform decisions regarding different combinations of service provision, in terms of their costs, cost-effectiveness, and public health outcomes, we discuss the barriers perceived by stakeholders to the use of DAMs to inform service planning for local populations, including (1) a tension between individual and population perspectives; (2) reductionism; and (3) a lack of transparency regarding models, their assumptions, and the motivations of those generating models.
Technological advances, including improvements in computing capability, are facilitating the development and use of models such as DAMs for health service planning. However, given the current scepticism among many stakeholders, encouraging informed critique and promoting trust in models to aid health service planning is vital, for example by making available and explicit the methods and assumptions underlying each model, associated limitations, and the process of validation. This can be achieved by consultation and training with the intended users, and by allowing access to the workings of the models, and their underlying assumptions (e.g. via the internet), to show how they actually work.
Constructive discussion and education will help build a consensus on the purposes of STI services, the need for service planning to be evidence-based, and the potential for mathematical tools like DAMs to facilitate this.