Placing sexually transmitted infection surveillance data in perspective by using national probability sample surveys.

Abstract

BACKGROUND\nMany European countries do not have comprehensive sexually transmitted infection (STI) surveillance in place. The objective was to investigate whether national probability sample surveys are useful in placing STI surveillance into perspective.\n\n\nMETHODS\nWe used data from the Dutch national cross-sectional probability sample survey on sexual health 2016 (18-34-year-old sexually active individuals). Descriptive analyses were performed regarding STI testing (last year). Test numbers were extrapolated from the survey and compared with surveillance data from sexual health centres (SHCs) (complete) and general practitioners (GPs) (representative estimates from 7% of all GPs). Statistical differences in characteristics between SHC attendees and general population (according to weighted survey participants) were determined using χ2 statistic. Predictors of recent testing at GPs or SHCs were determined using multinomial multivariable logistic regression.\n\n\nRESULTS\nOf the 17 222 survey invitees, 3217 (19%) were eligible for analyses. Testing uptake was higher in women (17.2%, 14.8-20.0%) than men (11.5%, 9.1-14.3%). The majority of tests were conducted by GPs followed by SHCs and hospitals. Number of tests extrapolated from the survey was similar to SHC surveillance data, but higher than GP surveillance data (women only). Testing at SHCs was associated with high-risk behaviour and with living in highly urbanized areas. Low education level and older age were, next to high-risk behaviour factors, determinants of testing at GPs.\n\n\nCONCLUSIONS\nNational probability sample surveys are useful for placing STI surveillance data into perspective by providing insights in testing patterns in the general population and identifying strengths and weaknesses of national surveillance systems.