Introduction The elimination target for congenital syphilis (CS) is <50 cases per 1 00 000 live births. We describe trends of laboratory data for possible early CS across South Africa (SA) and estimate its incidence in infants over a six year period.

Methods We conducted a cross-sectional analysis of laboratory data collected through the National Health Laboratory Service (NHLS) Corporate Data Warehouse (CDW) from 2010 to 2015. The NHLS CDW is a central repository of all laboratory tests performed in all public sector clinical laboratories in South Africa. We defined possible early CS as a positive serum Rapid Plasma Reagen (RPR) result in a child <2 years. Syphilis seropositivity rate was calculated as the number <2 years with a positive RPR test result divided by the total number <2 years tested during this time. Children with multiple RPR tests were counted once during this period. Incidence rate was calculated as the number of infants<1 year who had positive RPR results in a given year divided by the total number of livebirths for that year.

Results Of 84 341<2 years tested during 2010–2015 (92.3%<1 year, 40.7% male), 6357 (7.5%) tested positive. The number of children tested per year ranged from 23 555 in Gauteng Province to 1555 in the Northern Cape. RPR seropositivity in <2 years increased from 6% in 2010 to 9% in 2015 (p-value<0.001). Estimated national annual CS incidence rates in infants increased from 99/100,000 livebirths in 2010 to 119/100,000 livebirths in 2015. Largest increases in incidence rates were KwaZulu-Natal (44-115/100,000 livebirths), Limpopo (10-25/100,000 livebirths), and Mpumalanga (26-67/100,000 livebirths), while declines were observed in Eastern Cape (278-178/100,000 livebirths), Northern Cape (168-66/100,000 livebirths) and Free State (84-51/100,000 livebirths).

Conclusion Though limited to laboratory results, the analysis shows high incidence possible CS. SA may need to intensify elimination efforts to reach the WHO target. Further clinical and maternal treatment information is needed in order to confirm incidence rates of early CS.