Design

This is a multi-site, contemporaneous study of the centralisation of specialist surgical pathways for four cancers in two large conurbations in England. To understand stakeholder preferences for the organisation of cancer specialist surgical services, a discrete choice experiment will be conducted; to analyse the centralisations, we will combine quantitative analysis of the impact of centralisation on clinical processes, clinical outcomes, cost-effectiveness and patient experience (‘what works and at what cost?’), with qualitative analysis of their development, implementation and sustainability (‘how and why?’); a similar approach was used previously in a study of major system change in acute stroke services [30].

The study will focus on four surgical pathways (for prostate, bladder, renal and OG cancers). These pathways have been selected because they are being centralised in both areas, facilitating analysis of how changes occur in different contexts. The study will also analyse different degrees of centralisation, as the reduction in number of specialist centres varies significantly (see Table 1).

We will conduct a discrete choice experiment (DCE) [39–41] to examine stakeholder (patients, the public, healthcare professionals) preferences for centralisation, in relation to such attributes as travel time to hospital to undergo surgery, and risk of serious complications.

We will use documentary analysis, stakeholder interviews and non-participant observations to identify drivers for change; how the centralisations were planned and implemented and factors influencing this; the extent to which the proposals were implemented; and factors influencing sustainability of the changes.

This component of the evaluation will use a controlled before and after design. The centralisations will be analysed in terms of the extent to which changes were implemented and the impact of centralisation on care provision, clinical outcomes, patient experience and cost-effectiveness.

To develop lessons that might support centralisation in other contexts, we will host a workshop for stakeholders involved in planning centralisations of specialist cancer services elsewhere and other types of specialist service. Stakeholders will include providers, commissioners and patients and patient groups from across the country. Working with attendees, we will identify factors influencing generalisability of our findings and develop lessons that will be of use in different settings.