Abstract
Background
Congenital syphilis rates in the United States have increased significantly over the past decade. Syphilis is a curable infection with the potential for lifelong sequelae in the absence of timely diagnosis and treatment. Routine serologic syphilis screening is universally recommended during prenatal care with the traditional or the reverse diagnostic algorithm. False positive syphilis serologic testing in pregnancy can occur and comparative performance data for recommended algorithms in pregnancy are limited.
Primary objective
To compare the performance of the traditional algorithm and the reverse algorithm for the diagnosis of syphilis in pregnancy.
Study design
This retrospective analysis included pregnant women who delivered at our tertiary care center in the Southeastern United States during a period of increasing syphilis rates with testing performed between November 1, 2012 and December 31, 2019. We evaluated results according to the diagnostic algorithm used by facility laboratories at the time of syphilis screening (traditional 2012–2014 and reverse 2015–2019). Screen positivity, false positive test results, confirmed infection, and pregnancy outcomes were compared between the two periods. For secondary outcomes, multivariable logistic regression models were conducted to identify factors associated with false positive screening results and confirmed infection including maternal age, race, insurance status, test timing and location, and human immunodeficiency virus/sexually transmitted infection coinfection.
Results
Of 26,519 pregnant women tested for syphilis during the study period, 8781 were evaluated using the traditional algorithm and 17,738 were evaluated using the reverse algorithm. Mean age was 27.9 years, 85.3% of women were initially screened in the first trimester, and the mean number of syphilis testing episodes in pregnancy was 2.3. Screen positivity was 0.6% among women screened using the traditional algorithm compared to 1.6% for those tested with the reverse algorithm (p < 0.001). The proportion diagnosed with confirmed infection was similar in both algorithms: 0.2% traditional algorithm versus 0.3% reverse algorithm. Among those who screened positive with follow‐up testing performed, 52.4% and 60.4% were classified as falsely positive with negative confirmatory testing with the traditional algorithm and the reverse algorithms, respectively. In an adjusted model, delayed testing in pregnancy (odds ratio [OR], 22.8; 95% confidence interval [CI], 14.1–36.8 for ≥28 weeks compared to <14 weeks), inpatient or ER screening location (OR, 22.7; 95% CI, 13.4–38.4 vs. clinic), Black race (OR, 5.4; 95% CI, 3.2–9.2) compared to White, other sexually transmitted infection in pregnancy (OR, 2.2; 95% CI, 1.2–4.1]), and lack of private insurance (OR, 1.8; 95% CI, 1.2–2.9) were associated with false positive syphilis screening. The same factors were associated with confirmed syphilis in pregnancy except for STI coinfection. The reverse algorithm was only associated with false positive screening in the crude model (OR, 2.2; 95% CI, 1.4–3.5).
Conclusion
Syphilis screen positivity rates in pregnancy were nearly twice as high with the reverse algorithm compared to the traditional algorithm. Since false positive screening tests were common, improved diagnostic testing for active infection in pregnancy is needed.
Keywords: congenital syphilis, diagnostics, rapid plasma reagin (RPR), reverse syphilis algorithm, syphilis in pregnancy, traditional syphilis algorithm, Treponema pallidum (T. pallidum)
1. INTRODUCTION
Treponema pallidum (T. pallidum) is readily transmitted to the fetus in pregnancy. Adverse outcomes of congenital syphilis (CS) infection include intrauterine fetal demise (IUFD), stillbirth, neonatal death, and infant CS [1, 2]. In the United States, syphilis cases in pregnancy increased 400% over the past decade with 3882 cases of CS reported in 2023 [3]. Timely diagnosis and penicillin therapy in pregnancy prevent adverse birth outcomes [4].
Universal serologic screening for syphilis during pregnancy has been a long‐standing recommendation from the American College of Obstetricians and Gynecologists (ACOG), Centers for Disease Control and Prevention (CDC), and the US Preventive Services Task Force (USPSTF) [5, 6, 7, 8, 9]. Syphilis screening is based on the initial serologic detection of specific or nonspecific antibodies and follow‐up confirmatory testing is required [10]. The traditional algorithm begins with a non‐treponemal screening test (i.e., rapid plasma reagin [RPR] or venereal disease research laboratory [VRDL] test) and the reverse algorithm begins with treponemal antibody screening (i.e., treponemal enzyme immunoassay [EIA], T. pallidum particle agglutination [TPPA], or other treponemal test). False positive treponemal and non‐treponemal testing in pregnancy can occur—this complicates the ability to have an accurate and timely diagnosis.
Many hospital laboratories have switched to the reverse algorithm due to the ease, availability, and affordability of automated treponemal testing platforms with more rapid turnaround time [11]. Test algorithm performance depends partly on population prevalence and pretest probability [12]. Since pregnant women have lower syphilis prevalence compared to other groups recommended for routine screening by USPSTF and the impact of a missed case can be serious, dedicated performance assessments in pregnancy are needed [8]. Interpretation of diagnostic testing in this key population can be complex [13, 14, 15]. Our goal in this study was to compare performance characteristics of the traditional and reverse syphilis algorithms in pregnancy and to assess potential factors associated with false positive screening.
2. METHODS
2.1. Study design and population
We performed a retrospective analysis of pregnant women who delivered at our tertiary care center in the Southeastern United States with at least one antenatal syphilis test performed between November 2012 and December 2019. The population was divided into two cohorts based on the syphilis testing algorithm used at the time of screening: the traditional algorithm was utilized between November 1, 2012 and October 31, 2014, and the reverse algorithm between November 1, 2014 and December 31, 2019. Population‐level state data during the study period showed lower primary and secondary syphilis rates in 2012–2014 (3.9 cases per 100,000) versus 2015–2019 (8.6 cases per 100,000 population).
Sociodemographic information (age, self‐reported race, insurance), testing location (outpatient, emergency room [ER], inpatient), other antenatal laboratory testing [Chlamydia trachomatis (CT), Neisseria gonorrhoeae (NG), Trichomonas vaginalis (TV) by nucleic acid amplification testing (NAAT) of urine or vaginal swabs, serum hepatitis B surface antigen (SAg), hepatitis C antibody (HCV Ab), human immunodeficiency virus (HIV) antibody or HIV viral load], and subsequent birth outcomes were abstracted from the electronic medical record (EMR). Data obtained from women with multiple deliveries during the study period were restricted to their first pregnancy during each test period. Women who were pregnant during the syphilis test transition in November 2014 were categorized according to the algorithm used for their first positive syphilis screening test. Clinical syphilis staging, symptoms at the time of testing, history of syphilis, and sex partner information were not available. Infant outcomes were collected with linkage of maternal and newborn medical records in the EMR.
2.2. Syphilis testing
Testing outcomes included timing (trimester) of the initial syphilis screening test, gestational age (GA; trimester) of the initial positive screening test, and the number of syphilis testing episodes during pregnancy. Outpatient syphilis testing was conducted at the University of Alabama at Birmingham (UAB) Department of Obstetrics and Gynecology (OBGYN) Diagnostic Laboratory. The non‐treponemal test used was RPR (Macrovue) with quantification of reactive tests as titers determined by doubling dilution and ranging from 1:1 to 1:1024. RPR sensitivity is approximately 86% (95% confidence interval [CI], 75%–94%) and specificity 94% (95% CI, 84%–99%), with performance that varies by syphilis stage [16, 17]. For the reverse algorithm, the treponemal enzyme immunoassay (EIA) screen was Trinity Biotech Captia IgG with sensitivity 95%–100% and specificity 98%–100% in nonpregnant adults [15]. Confirmatory treponemal testing was TPPA [10]. Syphilis testing for inpatients and in the ER was performed in the UAB Main Diagnostic Laboratory using the same testing with one exception. The reverse algorithm screen was a qualitative EIA (TrepSure) with sensitivity 99% and specificity 83% [18]. During the study period, reflex TPPA testing for reverse algorithm discordant results (EIA+/RPR−) was in place in the main laboratory but not in the OBGYN Diagnostic Laboratory.
2.3. Study definitions and outcomes
Three main study outcomes were based on the diagnostic algorithm: initial screen reactivity (irrespective of follow‐up testing), confirmed infection (reactive screen and reactive confirmatory test), and false positive screening (reactive screen and nonreactive confirmatory test). The confirmed infection group includes active untreated infection as well as previously treated infection with persistently positive non‐treponemal testing (syphilis serofast). All participants with confirmed infection were treated for syphilis. Birth outcomes are also presented including a composite measure with CS, preterm delivery (<37 weeks GA), low birthweight (<2500 g), stillbirth (>28 weeks GA), IUFD (>20 weeks GA), neonatal death within the first 28 days of life, and spontaneous abortion (<28 weeks GA).
2.4. Statistical analysis
Demographic and clinical characteristics were reported with proportions for categorical variables and mean with standard deviation for continuous variables. Rates of screen positivity, false positive screening, confirmed infection, and negative testing were compared between traditional and reverse syphilis diagnostic algorithms.
Univariable and multivariable logistic regression analyses were conducted to assess whether testing algorithm was associated with false positive syphilis screening (primary outcome) and confirmed syphilis (secondary outcome). The following individual‐ and institutional‐level factors were included in multivariable analysis based on their clinical importance and published literature: age at delivery, race, insurance status, test timing and location, and coinfection with HIV, hepatitis B or C, and concomitant sexually transmitted infection (STI) (CT/NG/TV). Association was expressed in terms of odds ratio (OR; crude and adjusted) and 95% CI. Variance inflation factor was <1.6 for both outcomes, indicating no multicollinearity. Firth correction was used wherever applicable. Statistical significance was set at 0.05 (two‐tailed). Analysis was conducted using SAS statistical software, version 9.4. The study was approved by the UAB Institutional Review Board with a waiver of informed consent.
3. RESULTS
A total of 26,519 pregnant women were tested for syphilis during the study period: 8781 with the traditional algorithm and 17,738 with the reverse algorithm. Participant characteristics were similar in both groups (Table 1). Mean age was 27.9 years old, 85.3% were screened for syphilis during the first trimester, 77.6% were screened as outpatients, 22.1% were screened as inpatients, and 0.3% in the ER, and 97% of pregnancies ended with a live birth. On average, women were tested for syphilis 2.3 times during pregnancy with similar frequency between algorithm types. Syphilis was diagnosed throughout pregnancy (42% 0–14 weeks; 31% >14–27 weeks; 28% >28 weeks) and a higher proportion of women tested with the reverse algorithm had syphilis diagnosed during the third trimester (31% vs. 12% for women screened with the traditional algorithm) despite high rates of first trimester initial screening in both groups. Some women tested with both algorithms may have screened RPR negative early in pregnancy and later positive after November 2014, when the treponemal screening test was used.
TABLE 1.
Characteristics of participants by syphilis diagnostic algorithm.
| Characteristic |
Total N = 26,519 n (%) or mean (SD) |
Traditional algorithm n = 8781 n (%) or mean (SD) |
Reverse algorithm n = 17,738 n (%) or mean (SD) |
|---|---|---|---|
| Demographics | |||
| Age at delivery (years) | 27.9 (6.0) | 27.3 (5.9) | 28.2 (6.0) |
| Age at delivery (years) | |||
| ≤17 | 684 (2.6) | 274 (3.1) | 410 (2.3) |
| 18–29 | 16,023 (60.4) | 5671 (64.6) | 10,352 (58.4) |
| 30–39 | 9113 (34.3) | 2664 (30.3) | 6449 (36.4) |
| 40+ | 699 (2.6) | 172 (2.0) | 527 (3.0) |
| Race | |||
| Black | 12,615 (47.6) | 4423 (50.4) | 8192 (46.2) |
| White | 8769 (33.1) | 2695 (30.7) | 6074 (34.2) |
| Other | 4822 (18.2) | 1582 (18.0) | 3240 (18.3) |
| Unknown | 313 (1.2) | 81 (0.9) | 232 (1.3) |
| Hispanic ethnicity | 3936 (14.8) | 1292 (14.7) | 2644 (14.9) |
| Insurance | |||
| Private | 8575 (32.3) | 2475 (28.2) | 6100 (34.4) |
| Public | 15,696 (59.2) | 5675 (64.6) | 10,021 (56.5) |
| None | 2248 (8.5) | 631 (7.2) | 1617 (9.1) |
| Syphilis testing | |||
| Location a | |||
| Outpatient | 20,590 (77.6) | 6893 (78.5) | 13,697 (77.2) |
| Inpatient | 5684 (22.1) | 1873 (21.3) | 3991 (22.5) |
| Emergency room | 65 (0.3) | 15 (0.2) | 50 (0.3) |
| Timing of initial screening test | |||
| 0–13 weeks | 22,609 (85.3) | 8242 (93.9) | 14,367 (81.0) |
| 14–27 weeks | 3139 (11.8) | 506 (5.6) | 2633 (14.8) |
| 28+ weeks | 771 (2.9) | 33 (0.4) | 738 (4.2) |
| Gestation at first positive syphilis test | n = 335 | n = 51 | n = 284 |
| 0–13 weeks | 139/335 (41.5) | 28 (54.9) | 111 (39.1) |
| 14–27 weeks | 103/335 (30.7) | 17 (33.3) | 86 (30.3) |
| 28+ weeks | 93/335 (27.8) | 6 (11.8) | 87 (30.6) |
| # Syphilis testing episodes in pregnancy | 2.3 (0.9) | 2.2 (1.0) | 2.5 (0.9) |
| Coinfection in pregnancy | |||
| HIV | 23 (0.1) | 0 (0.0) | 23 (0.1) |
| Hepatitis B virus surface antigen + | 18 (0.1) | 0 (0.0) | 18 (0.1) |
| Hepatitis C antibody + | 158 (0.6) | 2 (0.0) | 156 (0.9) |
| Chlamydia NAAT + | 1866 (7.0) | 640 (7.3) | 1226 (6.9) |
| Gonorrhea NAAT + | 506 (1.9) | 168 (1.9) | 338 (1.9) |
| Trichomonas NAAT + | 99 (0.4) | 1 (0.0) | 98 (0.6) |
| Pregnancy outcomes a | |||
| Live birth (without neonatal death) | 25,693 (96.9) | 8521 (97.0) | 17,172 (96.8) |
| Spontaneous abortion (<20 weeks) | 3 (0.0) | 3 (0.03) | 0 (0.0) |
| Intrauterine fetal demise (≥20 weeks) | 586 (2.2) | 172 (2.0) | 414 (2.3) |
| Neonatal death (<28 days) | 81 (0.3) | 27 (0.3) | 54 (0.3) |
| Unknown | 156 (0.6) | 58 (0.7) | 98 (0.6) |
| Congenital syphilis | 2 (0.0) | 1 (0.0) | 1 (0.0) |
| Preterm delivery (<37 weeks) | 5947/25,678 (23.2) | 1911/8781 (21.8) | 4036/16,897 (23.9) |
| Low birthweight (<2500 g) | 4462/23,833 (18.7) | 1290/7510 (17.2) | 3172/16,323 (19.4) |
| Any adverse birth outcomes | 5703 (21.6) | 1653 (18.9) | 4050 (23.0) |
Abbreviations: HIV, human immunodeficiency virus; NAAT, nucleic acid amplification testing.
Missing data: location 3, pregnancy outcomes 156, preterm delivery 841, birthweight 2053.
For additional STI/HIV testing performed in our center, positivity rates were 7.0% for chlamydia, 1.9% for gonorrhea, and 0.1% for HIV. A composite measure of adverse birth outcomes was 23% during reverse algorithm testing (2015–2019) compared to 19% during traditional algorithm testing (2012–2014). Two asymptomatic infants were diagnosed with possible CS due to maternal treatment within 30 days of delivery and both were treated, one during each algorithm period. In the first case, the woman was asymptomatic and staged as early latent syphilis with RPR increase from 1:8 to 1:32 at 32 weeks and the second case was staged as secondary syphilis since the woman had a palmar skin lesion and RPR 1:128. Both women were treated as soon as lab results were available.
Screen reactivity at the individual level for the traditional algorithm was 0.6% (51/8781) compared to 1.6% (283/17,738) for the reverse algorithm (p < 0.001) (Figures 1 and 2). The overall proportion with confirmed infection was similar between testing algorithms (0.2% traditional vs. 0.3% reverse). Overall, the proportion of women with false positive screening was 0.3% for the traditional algorithm and 0.5% for the reverse algorithm. Among women who screened reactive with nonreactive follow‐up testing, 52.4% were classified as false positive screens in the traditional algorithm (RPR reactive/TPPA nonreactive), and 60.4% in the reverse algorithm (EIA reactive/RPR nonreactive/TPPA nonreactive). In the reverse algorithm, 18.7% of women with discordant results (EIA+/RPR nonreactive) did not have recommended follow‐up treponemal testing performed. These women were diagnosed with syphilis without a confirmatory test. The lack of a gold‐standard reference test for comparison is important.
FIGURE 1.
Traditional algorithm syphilis testing in pregnancy. The values in bold are proportion of total. RPR, rapid plasma reagin; TPPA, Treponema pallidum particle agglutination.
FIGURE 2.
Reverse algorithm syphilis testing in pregnancy. The values in bold are proportion of total. EIA, enzyme immunoassay; RPR, rapid plasma reagin; TPPA, Treponema pallidum particle agglutination.
Unadjusted and adjusted models conducted to assess the association between algorithm type and other factors with false positive syphilis screening are shown in Table 2. False positive screening was associated with the reverse algorithm in the crude model (OR, 2.2; 95% CI, 1.4–3.5) but not in the adjusted model (adjusted OR [aOR], 1.1; 95% CI, 0.6.1–7). After adjustment for participant and test characteristics, false positive screening was associated with delayed testing in pregnancy (aOR, 22.8; 95% CI, 14.1–36.8; for ≥28 weeks compared to <14 weeks), inpatient/ER screening location (aOR, 22.7; 95% CI, 13.4–38.4 vs. clinic), Black race (aOR, 5.4; 95% CI, 3.2–9.2 vs. White), other STI in pregnancy (aOR, 2.2; 95% CI, 1.2–4.1), and lack of private insurance (aOR, 1.8; 95% CI, 1.2–2.9). In our secondary model to identify factors associated with confirmed syphilis as a comparison, the same factors were associated with the outcome except for STI coinfection (aOR, 1.6; 95% CI 0.9–30) (Table 3).
TABLE 2.
Factors associated with false positive syphilis screen in pregnancy.
|
Univariable model Odds ratio (95% CI) |
Multivariable model Odds ratio (95% CI) |
|
|---|---|---|
| Testing algorithm | ||
| Traditional | REF | REF |
| Reverse | 2.2 (1.4–3.5) | 1.1 (0.6–1.7) |
| Age at delivery (years) | ||
| <30 | 1.3 (0.9–1.9) | 1.3 (0.9–1.9) |
| 30+ | REF | REF |
| Race | ||
| White | REF | REF |
| Black | 3.5 (2.1–5.8) | 5.4 (3.2–9.2) |
| Unknown/other | 1.7 (0.9–3.3) | 3.8 (1.9–7.5) |
| Insurance | ||
| Private | REF | REF |
| Public/none | 1.1 (0.8–1.7) | 1.8 (1.2–2.9) |
| Screening location | ||
| Clinic | REF | REF |
| Inpatient/emergency room | 21.1 (12.6–35.3) | 22.7 (13.4–38.4) |
| Timing of initial syphilis test | ||
| 0–13 weeks | REF | REF |
| 14–27 weeks | 5.8 (3.6–9.3) | 6.0 (3.7–9.7) |
| 28+ weeks | 39.2 (25.5–60.2) | 22.8 (14.1–36.8) |
| HIV infection | 4.7 (0.3–83.1) | 6.2 (0.3–125.2) |
| Hepatitis B or C coinfection | 1.3 (0.2–9.2) | 2.8 (0.5–17.5) |
| Other STI in pregnancy (CT/NG/TV) | 1.5 (0.9–2.6) | 2.2 (1.2–4.1) |
Note: Bold values highlight statistically significant values.
Abbreviations: CI, confidence interval; CT, Chlamydia trachomatis; HIV, human immundeficiency virus; NG, Neisseria gonorrhoeae; STI, sexually transmitted infection; TV, trichomonas vaginalis.
TABLE 3.
Factors associated with confirmed syphilis in pregnancy.
|
Univariable model Odds ratio (95% CI) |
Multivariable model Odds ratio (95% CI) |
|
|---|---|---|
| Testing algorithm | ||
| Traditional | REF | REF |
| Reverse | 1.4 (0.9–2.5) | 1.3 (0.8–2.1) |
| Age at delivery (years) | ||
| <30 | 0.9 (0.6–1.5) | 0.6 (0.4–1.0) |
| 30+ | REF | REF |
| Race | ||
| White | REF | REF |
| Black | 5.2 (2.5–11.0) | 4.3 (2.1–8.8) |
| Unknown/other | 2.6 (1.0–6.3) | 2.1 (0.9–5.1) |
| Insurance | ||
| Private | REF | REF |
| Public/none | 3.8 (1.9–7.6) | 3.4 (1.7–6.7) |
| Location of initial syphilis test | ||
| Clinic | REF | REF |
| Inpatient/emergency room | 1.9 (1.2–3.0) | 2.3 (1.5–3.7) |
| Timing of initial syphilis test | ||
| 0–13 weeks | REF | REF |
| 14–27 weeks | 1.8 (1.0–3.2) | 1.8 (1.0–3.2) |
| 28+ weeks | 4.8 (2.3–9.7) | 4.7 (2.3–9.4) |
| HIV Infection | 7.0 (0.4–122.8) | 4.5 (0.3–77.4) |
| Hepatitis B or C Infection | 0.9 (0.1–15.1) | 1.1 (0.1–17.0) |
| Other STI in Pregnancy (CT/NG/TV) | 2.0 (1.1–3.6) | 1.6 (0.9–3.0) |
Note: Bold values highlight statistically significant values.
Abbreviations: CI, confidence interval; CT, Chlamydia trachomatis; HIV, human immundeficiency virus; NG, Neisseria gonorrhoeae; STI, sexually transmitted infection; TV, Trichomonas vaginalis.
Younger maternal age (<30 years) was associated with higher rates of false positive screening (aOR, 1.3; 95% CI, 0.9–1.9) and lower rates of confirmed infection (aOR, 0.6; 95% CI, 0.4–1.0), but age did not reach statistical significance. In sensitivity analyses, models without HIV/HBV/HCV (with relatively small numbers) and screening location (with performance differences between tests used), associations were similar for both outcomes.
4. DISCUSSION
Among 26,519 pregnant women screened for syphilis at our facility, mostly as outpatients during the first trimester, the initial screening test was positive in 0.6% with the traditional algorithm and 1.6% with the reverse algorithm. A majority of these initially reactive screens in both algorithms were found to be false positive tests when follow‐up testing was performed. Confirmed syphilis rates with positive treponemal and non‐treponemal testing were similar in both algorithms (0.2% traditional vs. 0.3% reverse) as was the number of infants diagnosed with CS (one during each algorithm period). Since most pregnant women with active syphilis are asymptomatic, antenatal screening rates and diagnostic algorithm performance are crucial to identifying whom to treat.
For our primary question exploring the association between diagnostic algorithm type and testing outcomes, the reverse algorithm was not associated with the odds of false positive syphilis screening in adjusted models. Other maternal factors and test characteristics we identified as associated with false positive syphilis screening (delayed initial screening, inpatient testing, Black race, other STI in pregnancy, and lack of insurance) were also associated with confirmed syphilis. Race is likely a surrogate for unmeasured factors that may include access to care, social determinants of health, and/or characteristics of sexual partner networks.
To our knowledge, this is one of the few studies that compare the performance of the recommended traditional and reverse diagnostic algorithm in pregnancy. Since syphilis can only be cured if it is diagnosed and CS rates are rising in the United States, questions about test performance have added urgency. Falsely positive screening results were about twice as common in women tested with the treponemal‐based reverse algorithm compared to those screened with the RPR‐based traditional algorithm. Benefits of positive treponemal screening include detection of early or untreated infection, where antenatal treatment is necessary to prevent adverse outcomes. Limitations of serologic antibody detection include false positive testing that can complicate antenatal care since positive results can be stigmatizing, require additional testing and counseling, and may lead to unnecessary treatment.
Our study findings are consistent with other published data that documented high rates of false positive syphilis screening in pregnancy. In 2015, the false positive screening rate was 80% among 194 pregnant women in the United States with discordant reverse algorithm syphilis testing [T. pallidum chemiluminescence immunoassay (CIA)+/RPR−/TPPA−)] [13]. Among those with repeat CIA testing in this study, 53% reverted to negative antibody testing. A similar retrospective study in Ohio (n = 127), where the syphilis Bioplex IgG assay was used for reverse algorithm screening, showed a false positive screening rate of 65% in pregnancy [14].
We anticipate that our results would be generalizable to facilities that use other treponemal antibody tests since a recent systematic review found that most EIA/CIA immunoassays have similar performance [18]. One of our EIA screening tests (Trep‐sure) had lower specificity (83%) than others. This may have contributed to elevated false positive screening rates and the association with test location since it was only used in the inpatient/ER setting. The TPPA used in our study has excellent performance as a confirmatory test: 90%–100% sensitivity and 100% specificity in the setting of secondary or early latent syphilis stages [18].
The reverse algorithm consistently has more false positive results compared to the traditional algorithm (0.6% vs. 0%), including among low‐prevalent nonpregnant groups [19, 20]. Another limitation of the reverse algorithm in diagnosing active infection occurs in populations with high syphilis prevalence, where treponemal antibodies often reflect lifetime exposure and successful cure of syphilis in the past [21, 22]. In terms of cost‐effectiveness, two CDC studies suggested that use of the reverse algorithm in low‐prevalence setting was less cost effective compared to the traditional algorithm and led to overtreatment of uninfected individuals [23, 24]. Since these analyses focused on nonpregnant populations, results may not generalize to pregnancy given differences in health outcomes. Additional data in this area would be useful.
Updated 2024 ACOG guidelines recommend syphilis screening three times in pregnancy: at entry to care, after 28 weeks, and at delivery [25]. It is important for pregnancy providers to know the details and limitations of the syphilis algorithm used at their facility. Use of treponemal diagnostic tests for syphilis screening is increasing. In 2019, 36% of US lab managers surveyed reported using the reverse algorithm for syphilis testing (up from 16% a few years prior) [10, 26]. Treponemal tests are useful since they are reactive earlier in the natural history of syphilis compared to non‐treponemal tests and they can detect latent syphilis that may not have been diagnosed and treated in the past: two characteristics with high utility in pregnancy [27]. In contrast, non‐treponemal tests are useful in that they are quantitative with titers that can indicate burden of infection and assess response to therapy. In the 2024 CDC Laboratory Guidelines for Syphilis, the advantages and disadvantages of the traditional and reverse algorithms are described. Either algorithm is acceptable in pregnant and nonpregnant populations [10]. Guidelines note that ensuring a complete algorithm with both screening and confirmatory testing is critical to diagnostic accuracy and reflex testing can be useful. Automated non‐treponemal testing is now available in support of the traditional algorithm.
Although our findings show high rates of false positive treponemal EIA screening, RPR testing can also be falsely reactive (averaging 0.2%–0.8%) in a variety of settings, including pregnancy itself [28, 29, 30, 31, 32]. No matter which algorithm is used, reflex confirmatory testing should be in place to avoid missed opportunities if a provider fails to order the appropriate follow‐up test in a timely manner. Based on factors associated with confirmed syphilis in pregnancy (in this study and other published data), centers should ensure that ACOG guidelines for repeat syphilis testing in the third trimester and at delivery are followed universally, irrespective of insurance status, race, or timing of entry to care. Women should be encouraged to participate in timely antenatal care and supported to receive penicillin treatment for syphilis as soon as it is offered.
Due to the reactivity of non‐treponemal antibody testing in pregnancy and the lifelong persistence of treponemal antibodies after syphilis infection, serologic testing has critical limitations in a population where early and accurate diagnosis of active infection is needed to prevent adverse birth outcomes. After four decades of research, the diagnosis of many bacterial diseases now relies on direct detection of the pathogen using molecular methods such as quantitative polymerase chain reaction (PCR) [33]. In the past, T. pallidum culture was not possible and the utility of PCR testing was thought to be limited since spirochetemia occurs briefly during early infection when many patients are asymptomatic. The option of T. pallidum direct detection via PCR is now emerging with the recognition that organism shedding occurs at the oral cavity in significant proportions of asymptomatic adults with early, untreated syphilis [34, 35]. Although no PCR test for syphilis has been cleared by US Food and Drug Administration (FDA) to date, ongoing studies to assess the utility and performance of T. pallidum PCR testing on lesion and oral swabs in pregnancy may advance the field.
For improved serologic testing, some studies have suggested the utility of assay signal strength (defined by antibody index threshold) for treponemal IgG tests to identify samples in need of confirmatory TPPA testing and improve the cost‐effectiveness of the reverse algorithm [21, 36]. New advances in serologic testing that could allow for stratification into recent or distant infection would also represent an exciting advance to inform treatment recommendations. A direct comparison of the performance of traditional and reverse algorithms with samples collected for PCR testing could be conducted as a multicenter trial within the long‐standing National Institute of Child Health and Human Development (NICHD)‐funded Maternal Fetal Medicine Units network (MFMU) or National Institute of Allergy and Infectious Diseases (NIAID)‐funded International Maternal Pediatric Adolescent AIDS Clinical Trials (IMPAACT) network with potential for high impact.
Strengths of our study include the focus on pregnancy, sample size, screening rates, and access to robust EMR data including testing details and birth and neonatal outcomes over a 7‐year period. Limitations of the study include the lack of information about syphilis history, partner information to assess exposure risk, symptoms at the time of testing, and all comorbid conditions associated with biologic false positive testing (i.e., injection drug use or active autoimmune disease). Due to our retrospective study design, we did not perform traditional algorithm and reverse algorithm testing in parallel to compare differences. Also, different treponemal testing was used in the two labs and reflex follow‐up treponemal testing was not uniformly in place. As a result, some women without true infection were likely treated for infection based on false positive screening results. Since our catchment area is women who reside in a US region with elevated syphilis rates, study findings may not generalize to all pregnant women although our screening and positivity rates were similar to many other published studies.
5. CONCLUSIONS
In summary, the traditional and reverse syphilis testing algorithms had similar performance in pregnancy in detecting antibodies consistent with active infection. However, rates of false positive screening were high and reflex confirmatory testing must be ensured. Improved diagnostic methods for active syphilis in pregnancy are needed.
FUNDING INFORMATION
The authors received no specific funding for this work.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflicts of interest.
ACKNOWLEDGMENTS
The authors thank the UAB Department of Pathology and UAB Hospital Laboratories.
REFERENCES
- 1. Adhikari, E. H. 2020. “Syphilis in Pregnancy.” Obstetrics and Gynecology 135(5): 1121–35. 10.1097/aog.0000000000003788. [DOI] [PubMed] [Google Scholar]
- 2. Stafford, I. A. , Workowski K. A., and Bachmann L. H.. 2024. “Syphilis Complicating Pregnancy and Congenital Syphilis.” New England Journal of Medicine 390(3): 242–53. 10.1056/NEJMra2202762. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Prevention CfDCa . 2024. Sexually Transmitted Infections Surveillance 2023. https://www.cdc.gov/sti‐statistics/annual/index.html.
- 4. Eppes, C. S. , Stafford I., and Rac M.. 2022. “Syphilis in Pregnancy: An Ongoing Public Health Threat.” American Journal of Obstetrics and Gynecology 227(6): 822–38. 10.1016/j.ajog.2022.07.041. [DOI] [PubMed] [Google Scholar]
- 5. US Preventive Services Task Force . 2018. “Screening for Syphilis Infection in Pregnant Women: US Preventive Services Task Force Reaffirmation Recommendation Statement.” JAMA 320(9): 911–7. 10.1001/jama.2018.11785. [DOI] [PubMed] [Google Scholar]
- 6. Workowski, K. A. , Bachmann L. H., Chan P. A., Johnston C. M., Muzny C. A., Park I., Reno H., Zenilman J. M., and Bolan G. A., 2021. “Sexually Transmitted Infections Treatment Guidelines, 2021.” MMWR Recommendations and Reports 70(4): 1–187. 10.15585/mmwr.rr7004a1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Warren, H. P. , Cramer R., Kidd S., and Leichliter J. S.. 2018. “State Requirements for Prenatal Syphilis Screening in the United States, 2016.” Maternal and Child Health Journal 22(9): 1227–32. 10.1007/s10995-018-2592-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. US Preventive Services Task Force . 2022. “Screening for Syphilis Infection in Nonpregnant Adolescents and Adults: US Preventive Services Task Force Reaffirmation Recommendation Statement.” JAMA 328(12): 1243–9. 10.1001/jama.2022.15322. [DOI] [PubMed] [Google Scholar]
- 9. Amerson, E. H. , Castillo Valladares H. B., and Leslie K. S.. 2022. “Resurgence of Syphilis in the US‐USPSTF Reaffirms Screening Guidelines.” JAMA Dermatology 158(11): 1241–3. 10.1001/jamadermatol.2022.3499. [DOI] [PubMed] [Google Scholar]
- 10. Papp, J. R. , Park I. U., Fakile Y., Pereira L., Pillay A., and Bolan G. A.. 2024. “CDC Laboratory Recommendations for Syphilis Testing, United States, 2024.” MMWR Recommendations and Reports 73(1): 1–32. 10.15585/mmwr.rr7301a1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Nah, E. H. , Cho S., Kim S., Cho H. I., and Chai J. Y.. 2017. “Comparison of Traditional and Reverse Syphilis Screening Algorithms in Medical Health Checkups.” Annals of Laboratory Medicine 37(6): 511–5. 10.3343/alm.2017.37.6.511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Rourk, A. R. , Nolte F. S., and Litwin C. M.. 2016. “Performance Characteristics of the Reverse Syphilis Screening Algorithm in a Population with a Moderately High Prevalence of Syphilis.” American Journal of Clinical Pathology 146(5): 572–7. 10.1093/ajcp/aqw182. [DOI] [PubMed] [Google Scholar]
- 13. Mmeje, O. , Chow J. M., Davidson L., Shieh J., Schapiro J. M., and Park I. U.. 2015. “Discordant Syphilis Immunoassays in Pregnancy: Perinatal Outcomes and Implications for Clinical Management.” Clinical Infectious Diseases 61(7): 1049–53. 10.1093/cid/civ445. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Williams, J. E. P. , Bazan J. A., Turner A. N., Thung S. F., Hanlon C., Pettus T. R., and Sánchez P. J., 2020. “Reverse Sequence Syphilis Screening and Discordant Results in Pregnancy.” Journal of Pediatrics 219: 263–266.e1. 10.1016/j.jpeds.2019.11.035. [DOI] [PubMed] [Google Scholar]
- 15. Park, I. U. , Tran A., Pereira L., and Fakile Y.. 2020. “Sensitivity and Specificity of Treponemal‐Specific Tests for the Diagnosis of Syphilis.” Clinical Infectious Diseases 71(Suppl 1): S13–20. 10.1093/cid/ciaa349. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Lee, J. H. , Lim C. S., Lee M. G., and Kim H. S.. 2014. “Comparison of an Automated Rapid Plasma Reagin (RPR) Test with the Conventional RPR Card Test in Syphilis Testing.” BMJ Open 4(12): e005664. 10.1136/bmjopen-2014-005664. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Golden, M. , O'Donnell M., Lukehart S., Swenson P., Hovey P., Godornes C., Romano S., and Getman D., 2019. “Treponema Pallidum Nucleic Acid Amplification Testing to Augment Syphilis Screening Among Men Who Have Sex With Men.” Journal of Clinical Microbiology 57(8): e00572–19. 10.1128/jcm.00572-19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Park, I. U. , Fakile Y. F., Chow J. M., Gustafson K. J., Jost H., Schapiro J. M., Novak‐Weekley S., et al. 2019. “Performance of Treponemal Tests for the Diagnosis of Syphilis.” Clinical Infectious Diseases 68(6): 913–8. 10.1093/cid/ciy558. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Ortiz, D. A. , Shukla M. R., and Loeffelholz M. J.. 2020. “The Traditional or Reverse Algorithm for Diagnosis of Syphilis: Pros and Cons.” Clinical Infectious Diseases 71(Suppl 1): S43–51. 10.1093/cid/ciaa307. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Binnicker, M. J. , Jespersen D. J., and Rollins L. O.. 2012. “Direct Comparison of the Traditional and Reverse Syphilis Screening Algorithms in a Population With a Low Prevalence of Syphilis.” Journal of Clinical Microbiology 50(1): 148–50. 10.1128/jcm.05636-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Dionne‐Odom, J. , Van Der Pol B., Boutwell A., Biligowda N., Schmid D. G., and Hook E. W. 3rd. 2021. “Limited Utility of Reverse Algorithm Syphilis Testing in HIV Clinic Among Men Who Have Sex With Men.” Sexually Transmitted Diseases 48(9): 675–9. 10.1097/olq.0000000000001386. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Eaton, E. F. , Joe W., Kilgore M. L., and Muzny C. A.. 2018. “Reverse Syphilis Screening Algorithm Fails to Demonstrate Cost Effectiveness in Persons Living with HIV.” International Journal of STD & AIDS 29(6): 563–7. 10.1177/0956462417743409. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Owusu‐Edusei, K. Jr , Koski K. A., and Ballard R. C.. 2011. “The Tale of Two Serologic Tests to Screen for Syphilis—Treponemal and Nontreponemal: Does the Order Matter?.” Sexually Transmitted Diseases 38(5): 448–56. 10.1097/OLQ.0b013e3182036a0f. [DOI] [PubMed] [Google Scholar]
- 24. Owusu‐Edusei, K. Jr , Peterman T. A., and Ballard R. C.. 2011. “Serologic Testing for Syphilis in the United States: A Cost‐Effectiveness Analysis of Two Screening Algorithms.” Sexually Transmitted Diseases 38(1): 1–7. 10.1097/OLQ.0b013e3181ec51f1. [DOI] [PubMed] [Google Scholar]
- 25. Gynecologists ACoOa . Screening for Syphilis in Pregnancy: A Practice Advisory. 2024. https://www.acog.org/clinical/clinical‐guidance/practice‐advisory/articles/2024/04/screening‐for‐syphilis‐in‐pregnancy.
- 26. Rhoads, D. D. , Genzen J. R., Bashleben C. P., Faix J. D., and Ansari M. Q.. 2017. “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.” Archives of Pathology & Laboratory Medicine 141(1): 93–97. 10.5858/2016-0110-cp. [DOI] [PubMed] [Google Scholar]
- 27. Ho, E. L. , and Lukehart S. A.. 2011. “Syphilis: Using Modern Approaches to Understand an Old Disease.” Journal of Clinical Investigation 121(12): 4584–92. 10.1172/jci57173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Korentzelos, D. , Baloda V., Jung Y., Wheeler B., Shurin M. R., and Wheeler S. E.. 2022. “COVID‐19 mRNA Vaccines May Cause False Reactivity in Some Serologic Laboratory Tests, Including Rapid Plasma Reagin Tests.” American Journal of Clinical Pathology 158(2): 162–6. 10.1093/ajcp/aqac025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Swica, Y. , and Westhoff C.. 2009. “Comparison of Serum Markers for Thrombophilia and Autoimmune Disease in Reproductive Age Women With and without False Positive Rapid Plasma Reagin Tests.” American Journal of Reproductive Immunology 62(2): 73–7. 10.1111/j.1600-0897.2009.00719.x. [DOI] [PubMed] [Google Scholar]
- 30. Liu, F. , Liu L.‐L., Guo X.‐J., Xi Y., Lin L.‐R., Zhang H.‐L., Huang S.‐J., et al. 2014. “Characterization of the Classical Biological False‐Positive Reaction in the Serological Test for Syphilis in the Modern Era.” International Immunopharmacology 20(2): 331–6. 10.1016/j.intimp.2014.03.011. [DOI] [PubMed] [Google Scholar]
- 31. Yang, J. , Tacker D. H., Wen S., and LaSala P. R.. 2024. “Retrospective Comparison of False‐Positive Result Frequencies of 3 Syphilis Serology Screening Tests in Pregnant and Nonpregnant Patients at an Academic Medical Center in Appalachia.” Laboratory Medicine 55(3): 293–8. 10.1093/labmed/lmad078. [DOI] [PubMed] [Google Scholar]
- 32. Larsen, S. A. , Steiner B. M., and Rudolph A. H.. 1995. “Laboratory Diagnosis and Interpretation of Tests for Syphilis.” Clinical Microbiology Reviews 8(1): 1–21. 10.1128/cmr.8.1.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Schmitz, J. E. , Stratton C. W., Persing D. H., and Tang Y. W.. 2022. “Forty Years of Molecular Diagnostics for Infectious Diseases.” Journal of Clinical Microbiology 60(10): e0244621. 10.1128/jcm.02446-21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Towns, J. M. , Leslie D. E., Denham I., Wigan R., Azzato F., Williamson D. A., Lee D., et al. 2021. “Treponema Pallidum Detection in Lesion and Non‐Lesion Sites in Men Who Have Sex With Men with Early Syphilis: A Prospective, Cross‐Sectional Study.” Lancet Infectious Diseases 21(9): 1324–31. 10.1016/s1473-3099(20)30838-0. [DOI] [PubMed] [Google Scholar]
- 35. Dionne, J. A. , Giacani L., Tamhane A., Workowski K., Lieberman N. A. P., Greninger A. L., Perlowski C., Newman L., and Hook E. W., 2024. “Prevalence and Predictors of Oral Treponema Pallidum Detection by Quantitative Polymerase Chain Reaction in Early Syphilis.” Journal of Infectious Diseases 229(6): 1628–36. 10.1093/infdis/jiad582. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Berry, G. J. , and Loeffelholz M. J.. 2016. “Use of Treponemal Screening Assay Strength of Signal to Avoid Unnecessary Confirmatory Testing.” Sexually Transmitted Diseases 43(12): 737–40. 10.1097/olq.0000000000000524. [DOI] [PubMed] [Google Scholar]