The Maximising STI Control (MSTIC) webtool: a new approach to facilitate the planning of services for sexually transmitted infections to maximise public health benefit

Abstract

BackgroundIn the UK, national guidelines have sought to improve access to sexual health care, in part by expanding community-based provision of sexually transmitted infection (STI) testing, diagnosis, and care. Although useful data sources exist, there is little evidence for or guidance on the effect of different combinations of sexual health services on STI control locally. With finite resources, service planners need to identify the most effective combinations of STI services for their population. We sought to develop an evidence-based webtool aimed at supporting local service planning by synthesising the key data to show the relative consequences of different service configurations for different types of population. MethodsWe used a decision analytic model (DAM) to identify and model the effects of the characteristics of STI service provision, demography, and sexual behaviour that are likely to have the greatest effect on the incidence of common STIs. The DAM synthesised data from a range of sources, including census, surveillance systems, and probability sample surveys. We validated the DAM using data obtained from a standardised survey of people attending health services in four geographically and sociodemographically contrasting populations across England. The results of the model were presented as a webtool, requiring users to enter routinely available demographic data for their population along with characteristics and outcomes data for existing, proposed, or both health services providing care for STIs locally. The results are processed to estimate the effect on incidence of chlamydia and gonorrhoea locally. We consulted intended users attending a national sexual health conference to gain feedback on our webtool by giving them the opportunity to try the instrument for themselves, as well as ask questions of the designer and the mathematical modeller, for example, regarding the assumptions underlying the webtool. FindingsThe DAM identified the following demographic and service-level characteristics as key for estimating likely STI incidence in a locality: population type (urban, suburban, or semi-rural, taking account of assumptions about the area's ethnic mix and the proportion of men who have sex with men); size of the population aged 16–44 years, stratified by sex; type or types and capacity of services offering STI testing; level of partner notification achieved by these services; and mean delay between STI test and result. Consultation and usability testing revealed that commissioners and service planners were keen to explore the relative effect on disease control of varying assumptions about the type or types and capacity of services. However, clinicians were concerned about its reductionism and absence of individual patient perspective, although interested in the potential for exploring population outcomes for their services using an interactive decision-making instrument. We addressed these concerns by revising the accompanying user guide to ensure that we clearly explained the assumptions and limitations of our webtool, as well as including a technical appendix that provided details of the mathematical model. InterpretationOur evidence-based webtool allows service planning stakeholders to exploit routine data to inform rational STI service planning for their local population, showing how a one-size-fits-all approach does not apply to STI service planning. However, a willingness of users to engage with new approaches to service planning cannot be assumed. To achieve the buy-in necessary for the deployment of such instruments, user involvement has proven vital, resulting in improved communication and clarification of its assumptions and limitations. Evidence-based service planning needs to be promoted to ensure that STI services benefit both individual and public health. FundingMRC/DH Sexual Health and HIV Research Strategy Committee (grant G0601685).