Syphilis is caused by the bacterium Treponema pallidum and is highly infectious to sexual partners during the primary and secondary stages. The diagnosis is usually based on treponemal and nontreponemal antibody tests that become positive (or reactive) at some point during primary syphilis.1 Treponemal tests may sometimes become positive before the nontreponemal test, and usually remain positive for life.1–3 Nontreponemal antibody titers increase during early syphilis and often remain positive for years after successful treatment. False-positive low-titer nontreponemal tests have been associated with a variety of conditions, such human immunodeficiency virus infection, autoimmune disease, pregnancy, and injection-drug use. A 4-fold increase (eg, from 1:8 to 1:32) in nontreponemal antibody titer compared with past titers using the same serologic test is evidence of treatment failure or reinfection.
In the early 1960s, many states made positive serologic tests for syphilis reportable as part of their syphilis control efforts.4 Files of test results were established to allow programs to manually identify new cases and interpret future results. Because of the overwhelming volume of incoming tests, programs developed administrative procedures to prioritize tests for further investigations, with the goal of identifying the reactive serologic results most likely to result in new cases.5 By the 1970s, 1.4 million positive serologic tests for syphilis were reported each year in the United States6 and half of these were not investigated.7 Prioritization methods evolved slightly over time into current syphilis reactor grids.
A reactor grid prioritizes syphilis serologic tests based on the patients’ age, gender, test type, and test titer.8 The reactor grid does not apply to cases reported by providers, which are investigated regardless of the test titer. However, as automated electronic laboratory reporting increased,9,10 case reporting by health providers has become less common.4,11 Reactive tests are assigned for follow-up or are administratively closed without investigation based primarily on the likelihood of yielding a new infection. In theory, this likelihood is determined by periodically investigating all reported tests to determine how many new infections would have been missed because investigations were closed by the grid.8,12 However, very few published studies have evaluated reactor grids and those studies used data from select populations (eg, correctional facilities) during a time when syphilis rates were at historic lows.13,14 We wondered how syphilis reactor grids were currently being used, and if the focus on high-titer reactors might prevent investigations of primary syphilis cases for which intervention is warranted.
Study aims were to: (1) describe reactor grids and their role in syphilis surveillance and intervention activities, (2) assess trends of reactive nontreponemal tests reported to health departments, (3) examine how many reactive nontreponemal tests are reported for every case of syphilis, and (4) evaluate the potential for missed primary syphilis cases in the United States.
MATERIALS AND METHODS
A convenience sample of syphilis reactor grids from diverse health departments were reviewed (Arizona, California, District of Columbia, Florida, Kansas, Louisiana, New York City, North Carolina, Oklahoma, Pennsylvania, Philadelphia, Texas, and Virginia). Staff at various jurisdictions were interviewed for information about the use of reactor grids, data flow between state and local health departments, and programmatic activities and time required to process reported reactive nontreponemal tests.
Surveillance data provided trends in the number of reactive nontreponemal tests (eg, rapid plasma regain, venereal disease research laboratory) reported to Florida, Louisiana, New York City, Virginia, and Washington D.C. by gender and year from 2006 through 2015. Multiple reports of the same test result for the same person on the same day were deduplicated. To calculate ratios of the number of tests per case of syphilis (TPC), numbers of all reactive nontreponemal tests with and without titers were compared to the number of syphilis cases of all stages (ie, primary syphilis, secondary syphilis, early latent syphilis, late latent syphilis, unknown latent syphilis, late syphilis with clinical manifestations, and congenital syphilis) reported in the 5 jurisdictions.
Florida surveillance data from 2006 to 2015 were used as an example of the use of a reactor grid in a large public health program. An analysis of the number of reactive tests reported per syphilis case diagnosed (TPC ratio) was conducted using the number of deduplicated reactive nontreponemal tests with titers as the numerator and number of reported syphilis cases of all stages (with a titer at the time of diagnosis) as the denominator. Reactive nontreponemal tests with titers were deduplicated using specimen collection date, type of test, and nontreponemal test titer. The TPC ratios were calculated for subgroups by age, gender, and test titer to determine the likelihood that an investigation of a test would lead to a reported case. Reactive tests with unknown gender or unknown titers were not included.
Syphilis cases reported to the Centers for Disease Control and Prevention in 2013 to 2015 were analyzed to see if reactor grids might lead to missing primary cases with low titers. First, we calculated the percentage of reported primary syphilis cases that had nonreactive or low titers (≤1:4) that might be excluded by reactor grids. Second, because primary syphilis cases have lower test titers than secondary cases, and most grids exclude older persons with low test titers, we computed primary-to-secondary case ratios by age groups and gender to see if the ratios suggested older age groups were missing primary cases when compared to younger age groups. Potentially missed primary syphilis cases for each age group were estimated as:
where P1, number of primary syphilis cases (≤30 years old); S1, number of secondary syphilis cases (≤30 years old); P2, number of primary syphilis cases (older age group, eg, 31–40 years); X, number of potential missed primary cases (older age group, eg, 31–40 years); S2, number of secondary syphilis cases (older age group, eg, 31–40 years).
Processing of Reported Syphilis Serologic Results
The approach to investigating reported syphilis serologic results was highly variable across jurisdictions. In general, reactive tests are submitted to health departments by phone, mail, fax, and (most often) electronic laboratory reporting. Sometimes the same positive test is reported multiple times. For example, in 1 jurisdiction, 13.7% of all reactive nontreponemal tests received by the health department in 2006 to 2015 were duplicates that had the same specimen collection date. Reports that are not electronic must be manually entered into data systems. For large areas, after health departments receive all reactive serologic tests for syphilis, these are routed to appropriate regional or district offices based on patient address or ordering provider location for further processing (Supplemental Figure 1, http://links.lww.com/OLQ/A260).
Typically, when a local program receives a reactive test to investigate, it is compared with the health department database to see if the person has previous records in the surveillance and case management system (“profile search”). The report is also assessed for adequacy of information provided by the laboratory or provider and differences in information (eg, name, address, date of birth, test result) for the same person. One large state health department reported that it took 10 to 12 person-hours per week to reconcile duplicate syphilis serology records. Deduplication efforts can occur at multiple points during an investigation and are difficult to quantify.
Next, a reactor grid is used to close investigations of laboratory-reported reactive serologies unlikely to lead to a new case. Grids varied in structure and design but all prioritized incoming tests based on patient age, gender, and test titer (Fig. 1). In general, priority was given to persons from younger age groups (≤30 years) and with high titers (≥1:8). Reactive tests that do not meet the criteria for follow-up, based on low probability of syphilis diagnosis or transmission potential (eg, 65 years old with a low titer), are usually administratively closed, not investigated, and not reported as cases unless other circumstances apply. Other circumstances that may lead to further investigations include provider case reports, laboratory reports from high priority populations (eg, symptomatic, sexually transmitted disease (STD) clinic patients, pregnant women, human immunodeficiency virus-infected), or lab reports linked to an investigation of another case. The reactor grid would be overruled in these situations. Some reactor grids are applied electronically by surveillance data systems, while others are applied by a staff member reviewing records. Combinations of age and titers selected for administrative closures from sampled reactor grids varied with age cut-offs that ranged from 20 to 71 years, and titer cut-offs that ranged from 1:4 to 1:32. For example, 1 jurisdiction administratively closed 25% of all incoming reactive tests while another jurisdiction administratively closed less than 2%.
Reactive tests that are not administratively closed are compared to old syphilis records to help determine if they represent new infections or old, previously reported and treated infections. The amount of time required to check records can vary depending on the availability of historical laboratory data, number of staff, access to medical records, and changes to patient demographic information. Process data are not readily available at the local level, and the method of estimation is not uniform for sites, but in 1 jurisdiction, it took an estimated 4 to 8 minutes to review each record, while another jurisdiction took approximately 13 to 19 minutes. Field staff in some areas have to contact their state health department for assistance with record searching. Many investigations end after the record search determines that the test titer does not represent a new case (eg, the patient was previously treated and the current titer is not 4-fold higher than the last titer).
For reactive tests being investigated, providers are interviewed by surveillance staff or disease intervention specialists (DIS) to elicit information about clinical assessment, prescribed medications, treatment status, history of infection, social and medical risk factors, and pregnancy status. In some cases, laboratories are contacted about test results that were ordered by providers but were not reported to the health department. The time required for a provider interview in 1 jurisdiction took 45 minutes to 1 hour. However, the amount of time to contact and speak with providers varied by site. Health department staff stage cases after a provider interview or later in the investigation based on additional information.
For investigations of tests prioritized for DIS follow-up and partner services, patients are often interviewed for risk factors and for information about partners who may be infected. Partners are sought out for notification, testing, and treatment. Index case-patients may participate in memory-jogging exercises (eg, field tours) and reinterviews for additional sexual partners if the investigation is unable to identify the source of the infection. In-person or phone interviews last from 30 minutes to 3 hours. Frequency of contact attempts and waiting time for callback varies by patient.
Trends in Tests and Tests-per-Case Ratio
Trends in the number of reactive nontreponemal tests were obtained for 5 areas. These health departments received a total of 48,573 to 496,503 reactive serologies (median, 140,865) over a 10-year period. The number of tests varied by jurisdiction and year, ranging from 3044 to 57,242. From 2006 to 2015, the number of tests increased in all areas, with increases ranging from 36.6% in Virginia to 169.2% in Louisiana. The increase was greatest for men (69.5–220.3%) (Fig. 2). The number of tests for females decreased in 2 jurisdictions (−21.0% to 121.4%). The ratio of reported reactive nontreponemal tests per new case of syphilis (TPC) ranged from 6.2 to 11.1 at various jurisdictions. Despite large increases in incoming tests for men, the TPC for men remained relatively stable over time (Fig. 3). The TPC ratio was smaller for men than women in all jurisdictions.
Tests Linked to Syphilis Cases by Age, Gender, and Titer
There were 341,363 reactive nontreponemal tests with titers (216,813 men; 124,550 women) reported to Florida Department of Health in 2006–2015 (Table 1). Among men, more than half of tests were for men older than 40 years (59.8%), and men with test titers of 1:4 or less (55.1%). For women, 44.1% of reported tests were among women older than 40 years, and 76.1% had titers of 1:4 or less. Overall, there were 6.3 tests per reported case of syphilis in men, and 13.2 tests per reported syphilis case in women (Table 2). Among persons with titers 1:8 or less, TPC ratios for males ranged from 8.2 to 34.4; and for females, from 11.2 to 38.8. Reactors were more likely to represent a new case (smaller TPC ratio) as test titer increased and age decreased. Among women, a direct relationship between increasing age and TPC ratio was observed for those with titers of 1:32 or higher.
In Florida, the number of cases was likely underascertained among persons with low titers and older age groups because their investigations would have been administratively closed by the reactor grid, with the exception of provider-reported cases and other priority individuals (eg, pregnant women). The TPC ratio was high in shaded areas of the table that would have been administratively closed by the reactor grid without investigation (Table 2).
Potential for Missing Primary Syphilis
Our first estimate of potentially missed primary syphilis cases used the number of primary syphilis cases with test titers 1:4 or less reported in the United States. In 2013 to 2015, there were 18,988 primary syphilis cases reported (17,705 men; 1283 women) and 15,098 had nontreponemal test titers reported (14,137 men; 961 women) (Table 3). Of these, 23.6% of men and 22.9% of women had results that were reactive, weakly reactive, nonreactive tests, or low titers (≤1:4).
To see if older age groups were more likely than younger age groups to have missed primary syphilis cases, ratios of primary-to-secondary syphilis cases by age were assessed (Table 4). Compared with men 30 years or younger, the primary-to-secondary syphilis case ratio in 41- to 54-year-old men was 11.4% smaller. This suggests that 488 primary syphilis cases may have been missed in men aged 41 to 54 years. Ratios for men aged 31 to 40 years, and 55 years or older were 0.1% to 0.4% smaller than the referent group, suggesting that 8 primary syphilis cases might have been missed. Among women, the difference in primary-to-secondary syphilis case ratios for older age groups compared with those aged 30 years or younger grew larger with increased age. The numbers of potential primary syphilis cases missed were: 54 (31–40 year olds), 37 (41–54 year olds), and 12 (≥55 year olds) for a total of 103 potentially missed female primary syphilis cases.
The number of reactive nontreponemal tests reported in select jurisdictions has nearly doubled over the past 10 years, increasing the workload for STD programs already struggling with fiscal challenges.15,16 Prioritization of investigations is necessary given the large number of false positive tests or previously treated old infections.17,18 Reactor grids help prioritize serologic syphilis tests for investigation based on age, gender, and nontreponemal test titer. Although 2% to 25% of incoming tests were closed by current grids in select jurisdictions, administrative closures could increase if the number of reported tests continues to increase. For instance, more than half of all reactive tests reported in the United States were administratively closed in the 1970s amidst high syphilis rates.7 Furthermore, reactor grids may lead to potential missed cases of primary syphilis that are high priority for treatment and prevention.1 Trends in the number of reported syphilis cases reflect not only the epidemiology of disease, but are also influenced by the prioritization of reactive tests for follow-up.
Recent increases in reactive tests reported to health departments, particularly among men, may be in part due to frequent screenings of high-risk men who have sex with men (MSM).1,19 However, it is notable that the number of tests and TPC ratio were likely higher decades ago than today due to widespread screening of low-risk individuals (eg, premarital screening laws).5,20 In 1975, the national TPC ratio was approximately 17.4 based on 80,356 reported cases21 and 1.4 million positive tests.6 In the current study, the overall 10-year TPC ratios ranged from 6.2 to 11.1 for select jurisdictions. Reasons for site differences are not clear but variable processing algorithms and reactor grid criteria, screening policies (eg, increased reverse sequence screening in which specimens are first tested with an automated treponemal assay followed by testing reactive sera with a nontreponemal test),22,23 staff size, available resources, population, and other contextual factors unique to each program could affect the number of reactors prioritized for investigations and reported cases. Stable TPC ratios in males (despite increases in tests) could reflect true increases in disease prevalence and/or more follow-up testing for high-risk individuals.1,19 Compared with men, higher TPC ratios among women suggest lower disease prevalence and/or threshold for screening (low-risk pregnant women).24,25
Our findings using Florida data show that the likelihood of finding a syphilis case increases with test titer and younger age. However, the number of reported cases is influenced by the reactor grid, which likely results in under-ascertainment of cases with low titers of older age. The number of truly missed cases for Florida could not be determined; however, based on observed TPC ratios, there was a precipitous drop in the number of cases in areas of the grid that would have been administratively closed without investigation (ie, older age, low titers). This suggests that the low case yield for these areas could partly be an artifact of the grid. Grids that administratively close low titers could lead to missed cases. Future studies could compare findings from provider-reported cases or surveillance data using less restrictive grids.
One in 3 primary syphilis cases reported in the United States had low titers, but this is likely an underestimate of the true proportion because investigations of low-titer tests are often administratively closed. The ratios of primary-to-secondary cases reported suggest that almost 600 primary syphilis cases were missed in 2013 to 2015, mostly among men aged 41 to 54 years. However, some of these missed primary syphilis cases among older males could be due to MSM who are older than heterosexual men with syphilis26 and less likely to be diagnosed with primary infections.27 Missed cases would not likely receive partner services which has important public health implications for groups at high risk for syphilis infection and transmission (eg, MSM).21 Among women, more than half of potentially missed primary cases were within reproductive age (31–40 years). Missed syphilis in women of reproductive age is concerning since untreated syphilis during pregnancy can lead to devastating consequences for the mother and infant.1,28 Primary syphilitic chancres (skin lesions characteristic of primary syphilis) may go undiagnosed in women (or MSM) because the chancre is painless and not easily visible (eg, in the vagina, rectum).29 Thus, all reactive tests for women of reproductive age, regardless of titer, should be investigated to prevent congenital syphilis.
Reactor grids varied in design (eg, gender, age, test type, and titer), thresholds for investigation or administrative closure, and application during the processing of tests. Thus, the percentage of true missed cases depends on syphilis prevalence among persons with reactive tests and grid design.14 The true extent to which local reactor grids affected the number of reported syphilis cases in the current study is uncertain. Reactor grids are not frequently evaluated due to the amount of work involved to investigate all reported serologic tests,8 and grids may change simply because of financial constraints. Previous studies that reviewed data sources not using reactor grids in Chicago in 1998 (ie, STD clinics, county jail) reported 17% of men, 1% of women, 17% of persons with primary syphilis, and 11% of persons diagnosed with early syphilis would have been excluded by the reactor grid.13
The findings in this study are subject to several limitations. Most site-specific data were from high-morbidity areas and may not be representative of low-morbidity jurisdictions. Similarly, the summary approach to investigating reported reactive tests and the role of reactor grids may not be representative to other jurisdictions. In addition, there was no way to determine exactly how jurisdictions were reporting test titers for reported cases that had multiple test titers; for this analysis, the assumption reported was that the test was the “initial laboratory specimen used for diagnosis.” Furthermore, confounding due to contextual and individual-level factors (eg, biologic false positive results, sexual orientation) may have affected the trends in reactive tests reported to jurisdictions or number of potential missed primary cases. Lastly, the number of truly missed cases could not be determined due to lack of information. Similarly, it was unclear how many reactors that were not prioritized for investigation were still treated. Future studies may want to consider assessing missed cases using data from populations universally investigated.
Prioritization of reactive tests might be improved with computerized record searches that automatically compare all reported serologies to old records to see if they are likely to be new cases. Although serologic tests with high titers are likely to yield true syphilis cases, reactor grids could be leading to missed low-titer primary syphilis cases. Reactor grids were introduced at a time when record searches were done by hand looking through paper files. Now, records are electronic but most searches still involve an individual looking at previous serologic test results on a computer. Automating record searches of case registries with at least several years worth of prior serologic reports could help save time by eliminating previously treated old infections and biologic false positives or duplicate records from unnecessary investigation by surveillance staff or DIS. Already, an automated system for assigning reactor dispositions has been developed and implemented in San Francisco; 44% of reactors would have been automatically closed as an old case or biologic false-positive test based on an algorithm applied to retrospective data.17 Universal record searching would increase sensitivity in detecting new syphilis infections with fewer missed cases because all reactive tests would be record searched. Registries or databases of nonreactive and reactive results could improve detection of persons with evidence of seroconversion (ie, nonreactive then reactive tests). Specificity would increase because persons with false-positive tests would not be investigated, thus, supporting the allocation of program resources towards other activities. Although an automated system is not perfect, the potential to save program staff from unnecessary investigations while improving sensitivity with fewer missed cases should be explored. Future research could consider the efficiency gained and cost savings from implementing automatic record searches.
1. Workowski KA, Bolan GA. Sexually transmitted diseases treatment guidelines, 2015. MMWR MMWR Recomm Rep 2015; 64:1–137.
2. Larsen SA, Steiner BM, Rudolph AH. Laboratory diagnosis and interpretation of tests for syphilis. Clin Microbiol Rev 1995; 8:1–21.
3. Peterman TA, Schillinger J, Blank S, et al. Syphilis testing algorithms using treponemal tests for initial screening—four laboratories, New York City, 2005–2006. MMWR Morb Mortal Wkly Rep 2008; 57:872–875.
4. Lossick JG, Kraus SJ. Syphilis. In: Evans AS, Brachman PS, eds. Bacterial infections of humans: epidemiology and control. Second edition ed. New York and London: Plenum Medical Book Company; 1991:675–695.
5. Dandoy S, McKenna EM. Laboratory reporting of syphilis reactors in the Los Angeles program. Public Health Rep 1964; 79:1015–1020.
6. U.S. Department of Health E, and Welfare. Venereal Disease Fact Sheet. 1975.
7. U.S. Department of Health E, and Welfare. Venereal Disease Statistical Letter. 1975.
9. Lamb E, Satre J, Hurd-Kundeti G, et al. Update on progress in electronic reporting of laboratory results to public health agencies—United States, 2014. MMWR Morb Mortal Wkly Rep 2015; 64:328–330.
10. Turner K, Hamilton J, Hall CJ, et al. Progress in increasing electronic reporting of laboratory results to public health agencies—United States, 2013. MMWR Morb Mortal Wkly Rep 2013; 62:797–799.
11. St Lawrence JS, Montano DE, Kasprzyk D, et al. STD screening, testing, case reporting, and clinical and partner notification practices: A national survey of US physicians. Am J Public Health 2002; 92:1784–1788.
13. McLean CA, Kohl K, Baker MA, et al. The syphilis reactor grid: Help or hindrance for syphilis surveillance? Sex Transm Dis 2003; 30:650–653.
14. Schaffzin JK, Koumans EH, Kahn RH, et al. Evaluation of syphilis reactor grids: Optimizing impact. Sex Transm Dis 2003; 30:700–706.
15. Leichliter JS, Heyer K, Peterman TA, et al. US public sexually transmitted disease clinical services in an era of declining public health funding: 2013–14. Sex Transm Dis 2017; 44:505–509.
16. McCullough A, Ranney LM, Simons DJ, et al. “The job has become advocating for the job”: Threats to funding dramatically influence program outcomes. Am J Health Promot 2016:1–4.
17. Kohn RP, Calero G, Engelman J. Developing an automated system for assigning reactor dispositions. Atlanta, Georgia: STD Prevention Conference; September 22, 2016, 2016.
18. Gunn RA, Lee M, Oh C, et al. Syphilis serologic prevalence monitoring among STD clinic clients: Correlation with reported syphilis incidence, San Diego, CA, 1985-2004. Sex Transm Dis 2007; 34:749–753.
19. Cantor AG, Pappas M, Daeges M, et al. Screening for syphilis: Updated evidence report and systematic review for the US Preventive Services Task Force. JAMA 2016; 315:2328–2337.
20. Kingon RJ, Wiesner PJ. Premarital syphilis screening: Weighing the benefits. Am J Public Health 1981; 71:160–162.
22. Radolf JD, Bolan G, Park IU, et al. Discordant results from reverse sequence syphilis screening—five laboratories, United States, 2006–2010. MMWR Morb Mortal Wkly Rep 2011; 60:133–137.
23. Rhoads DD, Genzen JR, Bashleben CP, et al. Prevalence of traditional and reverse-algorithm syphilis screening in laboratory practice: a survey of participants in the College of American Pathologists Syphilis Serology Proficiency Testing Program. Arch Pathol Lab Med 2017; 141:93–97.
24. Lee KC, Ngo-Metzger Q, Wolff T, et al. Sexually transmitted infections: Recommendations from the U.S. Preventive Services Task Force. Am Fam Physician 2016; 94:907–915.
26. Pathela P, Braunstein SL, Schillinger JA, et al. Men who have sex with men have a 140-fold higher risk for newly diagnosed HIV and syphilis compared with heterosexual men in New York City. J Acquir Immune Defic Syndr 2011; 58:408–416.
27. Brewer TH, Peterman TA, Newman DR, et al. Reinfections during the Florida syphilis epidemic, 2000–2008. Sex Transm Dis 2011; 38:12–17.
28. Qin J, Yang T, Xiao S, et al. Reported estimates of adverse pregnancy outcomes among women with and without syphilis: A systematic review and meta-analysis. PLoS One 2014; 9:e102203.
29. Sparling PF, Swartz MN, Musher DM, et al. Clinical Manifestations of Syphilis. In: Holmes KK, Sparling PF, Stamm WE, et al, eds. Sexually Transmitted Diseases. Fourth edition ed: McGraw-Hill Medical; 2008:661–684.
Supplemental Digital Content
© Copyright 2018 American Sexually Transmitted Diseases Association