Catalyzing the Reactor Grid: A Web Application for Constructing Administrative Closure Algorithms for Syphilis Lab Reports

Background: Many health departments use a “reactor grid” to determine which syphilis serological results require investigation. When done manually, the use of such grids is labor intensive.  We developed a web-based tool, the Syphilis Reactor Grid Evaluator (SRGE), to facilitate reactor grid evaluations and demonstrated its utility using data from Public Health-Seattle & King County (PHSKC). 

Methods: We developed SRGE using the R Shiny™ web application framework. When populated with syphilis lab data, SRGE constructs the grid and displays the percent of verified early syphilis cases (based on final disposition codes) in each cell by titer result and patient age, optionally stratifying by gender, test type, and previous titer (2+ titer decrease or otherwise). The impact of administrative closure is dynamically computed as the user selects cells by calculating the efficiency gained (proportion of labs closed) and case-finding lost (proportion of cases missed) and the ratio of percent of cases identified over percent of labs investigated (efficiency ratio). After defining algorithms, users can compare them as a table or graph, combine subgroup-specific algorithms, and export the final algorithm. We used SRGE to compare the current PHSKC reactor grid to five alternate algorithms using all RPR results reported 01/01/2006-12/31/2015. 

Results: Of 13,593 lab reports, 1,592 were linked to verified cases.  Using SRGE, we found that updating PHSKC’s current administrative closure algorithm could result in an efficiency gain of 4.7%-24.4% (640-3314 labs) and case-finding loss of 1%-8.4% (10-97 cases). One alternate algorithm that administratively closes RPR results of Non-reactive, 1:1, or 1:2 in men aged 20+ and 1:4 in men aged 60+ would result in investigating 16.7% (2,160) fewer labs, missing 4.6% (70) more cases, and increasing the efficiency ratio from 1.13 to 1.33. 

Conclusions: SRGE is a user-friendly, quick and flexible tool for optimizing the syphilis reactor grid in the context of local epidemiology resources.