Background Two North East Syphilis outbreaks underlined the need for complete reporting of cases. Completeness was assessed by comparing GUMCAD with regional Enhanced Syphilis Surveillance. Capture–recapture studies were undertaken at yearly intervals to estimate the true number of cases and put forward recommendations to improve reporting completeness.

Objectives The objectives were to assess completeness of reporting of Syphilis cases in two surveillance systems, estimate the true number of cases, and provide recommendations for improving Syphilis reporting.

Methods Completeness of reporting was assessed for the period January 2010–June 2010 using capture–recapture methodology to estimate the true number of cases. Intervention A (clinic visits discussing findings) was performed and the effect measured by repeating the process. Follow-up studies were undertaken covering January 2010 to Jun 2011 to measure the effect and a different intervention B (linelists sent to clinics) was performed.

Results In the initial audit neither system included more than 73% of the estimated true burden of cases, although the aggregate figures suggested they were similar. After intervention A this increased to 94% in GUMCAD. The second audit showed a decreasing trend of reporting completeness after June 2010 to an average of 84% in January 2010–June 2011. After intervention B this improved to 96%. Causes of incompleteness in GUMCAD included incorrect coding and failure to record cases identified during screening leading to under-matching and overestimation of the total burden.

Discussion Completeness of reporting before intervention was 84% in GUMCAD suggesting the true number of syphilis cases was 19% higher than reported by official statistics. Completeness increased after intervention A but subsequently declined and a further intervention was required to maintain data quality. Regular feedback of case lists to clinics to check for missing cases can enhance completeness of Syphilis surveillance.