Exploring the feasibility of Conjoint Analysis as a tool for prioritizing innovations for implementation

BACKGROUND In an era of scarce and competing priorities for implementation, choosing what to implement is a key decision point for many behavioural change projects. The values and attitudes of the professionals and managers involved inevitably impact the priority attached to decision options. Reliably capturing such values is challenging. METHODS This paper presents an approach for capturing and incorporating professional values into the prioritization of healthcare innovations being considered for adoption. Conjoint Analysis (CA) was used in a single UK Primary Care Trust to measure the priorities of healthcare professionals working with women with postnatal depression. Rating-based CA data was gathered using a questionnaire and then mapped onto 12 interventions being considered as a means of improving the management of postnatal depression. RESULTS The 'impact on patient care' and the 'quality of supporting evidence' associated with the potential innovations were the most influential in shaping priorities. Professionals were least influenced by whether an innovation was an existing national or local priority, or whether current practice in the Trust was meeting minimum standards. Ranking the 12 innovations by the preferences of potential adopters revealed 'guided self help' was the top priority for implementation and 'screening questions for post natal depression' the least. When other factors were considered (such as the presence of routine data or planned implementation activity elsewhere in the Trust), the project team chose to combine the eight related treatments and implement these as a single innovation referred to as 'psychological therapies'. CONCLUSIONS Using Conjoint Analysis to prioritise potential innovation implementation options is a feasible means of capturing the utility of stakeholders and thus increasing the chances of an innovation being adopted. There are some practical barriers to overcome such as increasing response rates to conjoint surveys before routine and unevaluated use of this technique should be considered.