Utilization and Performance of a Rapid Treponemal Antibody Test for Diagnosis of Syphilis in a High-Risk Population at an Urban County Hospital

Abstract

Background:

Rapid detection of syphilis infection enables timely antibiotic treatment to prevent long-term complications and further transmission. We implemented a treponemal rapid lateral flow immunochromatographic test that was performed by trained clinical laboratory personnel on a 24/7 basis in an urban county hospital.

Methods:

From March 2023 to February 2024, 461 rapid syphilis tests (RSTs) were performed on serum from 449 patients. Specimens were also tested by a conventional reverse algorithm beginning with an enzyme immunoassay, followed by a rapid plasma reagin test, and when necessary, an additional treponemal test to resolve discordant results.

Results:

Most patients (349, 77.7%) initially presented to the emergency department. The most common patient risk factors for syphilis were substance use disorder (51.9%) and unstable housing (42.8%). Compared with the reverse algorithm, rapid syphilis test (RST) sensitivity was 52.2%, specificity 98.2%, positive predictive value 83.7%, and negative predictive value 92.1%. After excluding 45 patients who had a documented prior history of syphilis, RST sensitivity was 83.3%, specificity 98.2%, positive predictive value 74.1%, and negative predictive value 99.0%. Nearly all of the 33 patients testing negative by the RST but positive by enzyme immunoassay were more likely to have been treated for a previous infection: 27 had a documented history of treatment, and 30 had undetectable or low rapid plasma reagin titers (≤1:2). The median RST turnaround time was 1.6 hours compared with 197.8 hours for reverse algorithm screening.

Conclusions:

RST is a useful option for high-risk patients in urban settings. Other testing methods are preferable for previously treated syphilis, and conventional follow-up testing is required to confirm results.

Syphilis is a sexually and vertically transmitted infection caused by the spirochete Treponema pallidum subspecies pallidum. Untreated infections can progress through four clinical stages with sometimes nonspecific symptoms that can be mistaken for other diseases. Severe clinical manifestations include neurosyphilis, ocular syphilis, and cardiovascular syphilis, and untreated maternal infection frequently results in pregnancy loss or significant birth defects.¹ Although syphilis is effectively treated with benzathine penicillin G, challenges inherent to both clinical and laboratory diagnosis have hampered public health efforts to control the spread of disease.²

The currently recommended approach for the diagnosis of syphilis involves algorithmic use of indirect serological tests, and additional direct methods such as darkfield microscopy, molecular assays, and immunohistochemistry are applied in specific settings.³ Serologic assays are categorized as treponemal or nontreponemal (also referred to as lipoidal). Nontreponemal tests such as the rapid plasma reagin (RPR)⁴ detect antibodies produced in response to antigens such as cardiolipin, phosphatidylcholine, and cholesterol originating from both T. pallidum and damaged host tissue during infection.⁴ Sequential RPR titers can be compared to assess for possible re-infection with rising titers or a successful treatment response with decreasing titers over time. Treponemal tests, including enzyme immunoassays (EIA), chemiluminescence immunoassays (CMIA), and T. pallidum particle agglutination assays (TPPA), detect antibodies specific for T. pallidum.^3,4 In contrast to nontreponemal antibodies, treponemal antibodies typically remain detectable for life even after successful antibiotic treatment.^3,4

Both classes of serological tests have known limitations that must be considered when interpreting results. Serological tests can yield false negative results, particularly in cases of early primary infection.⁴ Biological false positive results can occur due to nontreponemal test cross-reactivity in patients with other conditions including autoimmune diseases, other bacterial or viral infections, and in older or pregnant indivduals.^4,5 Additionally, treponemal tests exhibit lower positive predictive value when the prevalence of infection is low.⁶ As no single assay is sufficiently sensitive and specific to be used as a standalone test, algorithmic testing is recommended.^4,7 The traditional or forward algorithm for testing begins with a nontreponemal test with reflexive performance of a confirmatory treponemal test when positive.^4,8 Alternatively, an increasing number of laboratories utilize a reverse algorithm that begins with a treponemal test performed on an automated platform, with positive results confirmed by a nontreponemal test. An additional treponemal test, ideally using antigens different from the ones tested in the initial step, is used as a third step as needed to arbitrate discrepant results.⁸ Previously infected patients can have detectable treponemal antibodies for life, regardless of treatment status.⁴ Test results must be considered together with a patient’s symptoms, exposures, and treatment history, and in some cases retesting is warranted.

The overall number of syphilis cases reported worldwide has continued to increase in recent years across all stages of disease, including congenital disease.⁹ Syphilis treatment has been complicated by episodic supply chain shortages of sterile injectable benzathine penicillin G.¹⁰ The rising incidence of syphilis, including congenital syphilis, and the challenge of providing follow-up care to medically underserved populations, prompted our laboratory to introduce a rapid syphilis testing solution at our urban hospital in King County, Washington.^10,11 Our primary goal was to enable same-day diagnosis and treatment of syphilis, particularly in patients with limited access to healthcare.

The Syphilis Health Check (Diagnostics Direct, LLC) is a lateral flow immunochromatographic test that detects IgG and IgM antibodies to T. pallidum antigens.¹² This rapid syphilis test (RST) is waived under the Clinical Laboratory Improvement Amendments and cleared by the US Food and Drug Administration (FDA) for rapid testing in clinical and point-of-care settings with a visual readout of a qualitative result after a 10 to 15 minute incubation period.¹² Several studies have found test performance to be superior using serum rather than whole blood or plasma specimens, and reported sensitivity using serum samples ranges from 77.8% to 100%, with 50% to 100% specificity.¹² We introduced this RST as a rapid preliminary test available 24 hours/7 days a week for providers to order on a STAT basis. Here, we retrospectively reviewed the performance and clinical impact of the RST during the first 12 months after the implementation at our hospital.

METHODS

Study Design

This retrospective study was conducted at Harborview Medical Center, an academic public medical center and county hospital in Seattle, WA. Patient serum specimens with orders for rapid syphilis testing collected between March 1, 2023 and February 29, 2024 were included. All specimens were required to have concurrent orders for confirmatory syphilis testing by conventional serology as described. Eight specimens were excluded from the study due to insufficient sample volume. A total of 461 specimens from 449 patients were included. The study was approved by the University of Washington, School of Medicine Human Subjects Division with a consent waiver (STUDY00019552).

Demographic Information and Chart Review

Demographic data, including sex, age at time of collection, race, and ethnicity, were collected from patient electronic medical records. Chart notes and previously obtained microbiology test results were reviewed for documentation of treatment with penicillin or doxycycline, prior history of syphilis, and risk factors for infection at the time of patient presentation, including HIV status, pregnancy, substance use disorder, unstable housing, incarceration, and detection of other sexually transmitted infections including trichomoniasis, Chlamydia, and gonorrhea.

Rapid Syphilis Testing

Serum samples were tested using the Syphilis Health Check assay (Diagnostics Direct, LLC, Stone Harbor, NJ). Testing was available 24 hours/day, 7 days/week, and performed in a high-complexity clinical microbiology laboratory by trained clinical microbiology laboratory personnel. Positive test results were considered critical values and were called to the ordering provider with the required read-back of results. The test order was available to all providers. Providers in the emergency department were instructed to order this test for patients who are pregnant or who are at risk for loss to follow-up, particularly those with housing instability or unreliable contact information.

Confirmatory Syphilis Testing Algorithm

Conventional syphilis serology testing was performed according to a reverse algorithm (Supplemental Fig. S1, https://links.lww.com/OLQ/B362). In-house testing was performed 5 to 6 days per week and batched once per day shift. Specimens were first screened with a treponemal BioPlex enzyme immunoassay (EIA; Bio-Rad Laboratories, Hercules, CA). Specimens positive by the BioPlex underwent additional testing by qualitative RPR (Arlington Scientific, Springville, UT) and were sent to a reference laboratory for supplementary testing. The reference laboratory performed a treponemal chemiluminescent microparticle immunoassay (CMIA; Abbott Alinity I) followed by a quantitative RPR. If the CMIA and RPR results were discordant, a TPPA was performed. Results were interpreted as positive by the reverse algorithm if the EIA was reactive and at least one additional treponemal test (CMIA or TPPA) or nontreponemal test (RPR) was positive.⁴ Test turnaround time was calculated as the time between specimen receipt in-laboratory and final result reporting time in the electronic medical record. For rapid syphilis testing, positive results were notified to providers immediately before reporting. For in-house conventional testing, the time of the EIA result was used if the RPR was not reflexively performed. Laboratory staff placed courtesy calls to notify providers of positive test results in patients with no known history of syphilis. Timestamps were obtained from the Laboratory Information System.

Statistics

Test performance characteristics, including sensitivity, specificity, positive predictive value, negative predictive value, and confidence intervals using the Wilson score method, were calculated with the epiR package (version 2.0.88) in R (version 4.5.2).^13,14

RESULTS

During the initial 12-month period after rapid syphilis testing became available (March 2023 to February 2024), 43 out of 461 total specimens (9.3%) from a total of 449 individuals were positive by RST (Fig. 1). In addition to the 43 patients with positive RST results, 33 patients tested negative by RST but ultimately tested positive for syphilis by the confirmatory reverse algorithm (Fig. 1). Emergency department providers ordered 360/461 (78.1%) of the tests (Table 1). Of the patients tested, 198/449 (44.1%) were 26 to 40 years of age, 141/449 (31.4%) were 41 to 60 years of age. Among these groups, 32/198 (16.2%) and 25/141 (17.7%) tested positive for syphilis, respectively (Table 1). More male patients were tested than female patients (265/449, 59%).

Figure 1.

Summary of rapid syphilis and confirmatory reverse algorithm testing results. All rapid syphilis test (RST) results were subsequently confirmed by serology using the reverse algorithm as described. Patient charts were reviewed for documented history of syphilis, and for patients with a positive RST, antibiotic treatment with penicillin or doxycycline was administered during their clinical encounter based on the RST result.

TABLE 1.

Demographic Data for Patients With Rapid Syphilis Testing Orders

Patient Characteristic	All Tested		Positive Result*
	n = 449	(%)	n = 74	(%)
Age (years)
<10	0	(0)	0	(0)
10–17	4	(0.9)	0	(0)
18–25	61	(13.6)	12	(16.2)
26–40	198	(44.1)	32	(43.2)
41–60	141	(31.4)	25	(33.8)
>60	44	(9.8)	5	(6.8)
Unknown	1	(0.2)	0	(0.0)
Sex
Female	184	(41.0)	35	(47.3)
Male	265	(59.0)	39	(52.7)
Race
American Indian or Alaska Native	17	(3.8)	4	(5.4)
Asian	27	(6.0)	1	(1.4)
Black or African American	140	(31.2)	28	(37.8)
Native Hawaiian or Other Pacific Islander	8	(1.8)	0	(0.0)
Unknown	5	(1.1)	1	(1.4)
White	227	(50.6)	37	(50.0)
Other	25	(5.6)	3	(4.1)
Ethnicity†
Hispanic or Latino	79	(17.6)	12	(16.2)
Non-Hispanic or Latino	365	(81.3)	61	(82.4)
Unknown	5	(1.1)	1	(1.4)
Risk Factors‡
HIV	20	(4.5)	14	(18.9)
Pregnant	20	(4.5)	6	(8.1)
Unstable housing	192	(42.8)	38	(51.4)
Substance use disorder	233	(51.9)	46	(62.2)
Incarcerated	31	(6.9)	7	(9.5)
Positive for other STI	30	(6.7)	11	(14.9)
No risk factors	151	(33.6)	12	(16.2)
1+ risk factor	298	(66.4)	62	(83.8)
2+ risk factors	193	(43.0)	45	(60.8)
Ordering department	n = 461	(%)	n = 76	(%)
Emergency department	360	(78.1)	60	(78.9)
Other	101	(21.9)	16	(21.1)

^*Number of unique patients positive by either rapid syphilis or reverse algorithm testing.

^†Medicine/Surgical, ICU, Abuse and Trauma, Psychiatry, OB/GYN, Neurosurgery.

^‡Not mutually exclusive.

Notably, 298/449 (66.4%) of patients tested had at least one documented risk factor associated with syphilis infection. This included 192 patients (42.8%) with unstable housing, 233 patients (51.9%) with documented substance use disorder, 30 patients (6.7%) positive for at least one other sexually transmitted infection at time of presentation, and 193 patients (43.0%) with at least two risk factors (Table 1; Supplemental Table S4, https://links.lww.com/OLQ/B362). Twenty (4.5%) patients were pregnant at the time of testing, and of the pregnant patients, 6/20 (30.0%) were positive by RST and/or reverse algorithm testing. Of the 43 patients who tested positive by RST, 31 (72.1%) were treated with antibiotics during their hospital encounter (Fig. 1).

Using the reverse algorithm as the gold standard (Fig. S1, https://links.lww.com/OLQ/B362), RST sensitivity was 52.2% (95% CI 40.6%–63.5%), and specificity was 98.2% (Fig. 1, Table 2). The positive predictive value was 83.7% and the negative predictive value was 92.1% (Table 2). There were 45 patients in this study with a self-reported previous history of syphilis or a previously positive result documented in their medical record (Fig. 1). Of these, 16 were positive by both RST and the reverse algorithm (true positives), and 29 were positive by the reverse algorithm only (false negatives) (Fig. 1, Table 2).

TABLE 2.

Rapid Syphilis Test Performance Characteristics

	Reverse Algorithm
RST	Positive	Negative
RST vs. Conventional Reverse Algorithm, All Tests Performed
Positive	36	7
Negative	33	385
Sensitivity	52.2% (95% CI 40.6–63.5)
Specificity	98.2% (95% CI 96.4–99.1)
Positive predictive value	83.7% (95% CI 70.0–91.9)
Negative predictive value	92.1% (95% CI 89.1–94.3)
RST vs. Conventional Reverse Algorithm, Excluding Tests From Patients With Known History of Syphilis
Positive	20	7
Negative	4	385
Sensitivity	83.3% (95% CI 64.1–93.3)
Specificity	98.2% (95% CI 96.4–99.1)
Positive predictive value	74.1% (95% CI 55.3–86.8)
Negative predictive value	99.0% (95% CI 97.4–99.6)

Sensitivity, specificity, positive predictive value, and negative predictive value were calculated using the overall interpretation by reverse algorithm testing as the reference standard.

Thirty-nine specimens tested negative by RST but positive by EIA, and 33 of these EIA-positive specimens were subsequently confirmed to also be positive by RPR and/or CMIA (Supplemental Table S1, https://links.lww.com/OLQ/B362 and Table 2). Six specimens were EIA-positive but negative by both RPR and CMIA, and overall were interpreted as negative (Fig. S1, https://links.lww.com/OLQ/B362). Of the RST-negative/reference method-positive specimens, 29/33 (87.9%) were from patients with a documented history of prior syphilis, and 30/33 (90.9%) were either undetectable by RPR or weakly positive with titers of 1:1 or 1:2, consistent with a remote history of treated infection (Fig. 2, Table 2; Supplemental Table S1, https://links.lww.com/OLQ/B362). Of the 3 patients with negative RST results and RPR titers >1:2, two were presumed to have new infections based on a rising RPR titer over time, and one had known serofast status (Supplemental Table S2, https://links.lww.com/OLQ/B362). We recalculated RST performance excluding patients who were previously diagnosed with syphilis. Test sensitivity improved to 83.3% (95% CI 64.1%–93.3%), and test specificity was 98.2% (Table 2), with positive predictive and negative predictive values of 74.1% and 99.0%, respectively (Table 2).

Figure 2.

RPR titer of specimens with false-negative RST results. Results from patients who did not have a previously documented history of syphilis are in purple, patients with a documented history of syphilis but unknown treatment status in yellow, and patients with a documented history of syphilis with treatment in gray.

RST was performed on specimens from 20 patients who were pregnant at the time of testing. Of those, 5/20 (25.0%) were positive by RST and subsequently confirmed positive by the reverse algorithm, that is, positive by screening EIA followed by a reactive RPR or second treponemal test (CMIA/TPPA) (Supplemental Table S3, https://links.lww.com/OLQ/B362). One pregnant patient was negative by RST but positive by EIA with a low RPR titer of 1:2 and a previously documented history of syphilis (Supplemental Table S3, https://links.lww.com/OLQ/B362). The RST result prompted antibiotic treatment in 4 of the positive pregnant patients, 3 of whom had a prior history of syphilis with persistently detectable RPR titers (Supplemental Table S3, https://links.lww.com/OLQ/B362). One patient with a positive RST result was not treated during their initial visit due to self-directed discharge but was later contacted by their provider to receive follow-up treatment.

The median syphilis test turnaround time was calculated as the time from specimen collection to result reporting. The median turnaround time was 1.6 hours (range 0.5–34.8 hours) for RST. In comparison, the median turnaround time was 32.0 hours (range 5.5–146.2 hours) for the initial EIA and RPR performed in-house according to the reverse algorithm (Table 3). Specimens with a discrepant EIA and RPR were sent to an outside reference laboratory for additional testing; in these cases, the median time from initial specimen collection to final result, inclusive of specimen transportation time, was 198.7 hours (range 95.9–475.3 hours).

TABLE 3.

Test Result Turnaround Time

Time From Collection to Result	Median (Hours)	Range (Hours)
Rapid syphilis test	1.6	0.5–34.8
In-house reverse algorithm*	32.0	5.5–146.2
Send-out reverse algorithm†	198.7	95.9–475.3

^*Time to result for EIA and qualitative RPR.

^†Time to result for CMIA, quantitative RPR, and TPPA when applicable.

DISCUSSION

The burden of syphilis has remained high in recent years. The World Health Organization (WHO)estimates there were 800 million incident cases among adolescents and adults in the 15 to 49 year age range in 2022.¹⁵ Of particular concern, the incidence of congenital syphilis in the United States has continued to increase over 12 consecutive years, with 3941 cases reported in 2024.⁹ Widespread screening in an urban emergency department setting has been shown to significantly increase the number of infections detected, including in pregnant patients.¹⁶ However, the rate of treatment completion, even among identified cases, is suboptimal, with many patients untreated and lost to follow-up care.¹⁷ At our institution, the RST was frequently ordered for patients with risk factors such as housing instability, substance use disorder, or other factors associated with infection and uncertain access to follow-up care, in line with trends found from another study in King County, Washington.¹¹ Patients with intersecting risk factors have been shown in other studies to be particularly vulnerable¹⁸; in the current study, 43.0% of patients undergoing testing presented with at least two risk factors. Although our guidance was to reserve the RST for patients with concern for loss to follow-up, we note that 33.6% of the patients in our study did not have documented risk factors. The need for rapid testing in this population is less certain.

Although many public health efforts to address syphilis have focused on men who have sex with men, the recent epidemic has seen a significant increase in infections in women, including pregnant women, both nationally and locally.^9,11 One study found that 36.8% of congenital syphilis cases in the United States in 2022 were associated with a lack of testing, and of the cases in which testing was performed, 39.7% were still inadequately treated.¹⁹ Rapid testing has the potential to both increase overall test access and expedite treatment interventions in high-risk and pregnant patients. Although only 20 pregnant patients were tested in this study, RST positivity prompted same-day antibiotic treatment in 4 of the 5 positive pregnant patients.

The Syphilis Health Check was the first point-of-care syphilis test to be cleared by the FDA and is considered waived under Clinical Laboratory Improvement Amendments.¹² Other European Conformity marked and WHO pre-qualified RST devices are also available on the market, as well as several tests approved for in-home use.²⁰ Most real-world evaluations have been in point-of-care settings (outpatient sexual health clinics, community centers), where patient whole blood specimens are obtained by fingerstick.^21–23 The few clinical studies to-date that have evaluated the Syphilis Health Check from serum have been from the assay’s FDA application and included several STI clinics and a hospital clinic as study sites.¹² Test sensitivity was reported to range from 77.8% to 100%, with specificity 50.0% to 100%.¹² In our study, RST sensitivity was 83.3% and specificity 98.2%. The wide range in performance between studies may be attributable to local disease prevalence, differences in patient population, and reference testing methodology. One laboratory-based evaluation on archived serum specimens found the assay’s sensitivity to be lower for patients living with human immunodeficiency virus (HIV).²⁴

Temporal trends may also impact disease prevalence and test performance. In contrast to most other real-world evaluations of the RST currently in the literature, which mostly took place before 2020, the present study was conducted from 2023 to 2024 during a period in which the incidence of syphilis was rapidly rising.^2,21,25,26 Another recent study in which RST was performed in incarcerated individuals in King County, Washington from 2022 to 2023 reported a lower positive predictive value of 64.7%. However, this was performed in a point-of-care setting using whole blood obtained by fingerstick.²⁷ We note that our study took place during a time period in our region where syphilis diagnoses began to decline in several populations including cisgender men and women after the introduction of doxycycline post-exposure prophylaxis (doxy-PEP).²⁸ This may have also impacted disease prevalence and test performance.

After excluding patients with a prior history of syphilis, the RST had a positive predictive value of 74.1%. Seven patients had a positive RST result that was ultimately determined to be a false positive after confirmatory testing. Four of these patients were treated with antibiotics. False positive results can potentially cause patient confusion, anxiety, and adverse effects from unnecessary treatment.²⁹ Our review of the medical histories did not reveal obvious risk factors that could explain the false positive RST results. In the FDA clinical trials for the RST, several false-positive results were documented from serum.¹² In another study comparing the performance of several treponemal tests, false positives were associated with a weaker EIA signal.³⁰ The negative predictive value was higher at 99.0%, and of the 4 patients who were initially negative by the RST but positive by confirmatory testing, only 1 had a detectable RPR titer at 1:2. It is possible these patients were previously treated for syphilis, even though this was not formally documented in their medical record. Alternatively, these patients may have very early syphilis. Overall, these cases highlight the need to couple RST use with definitive confirmatory testing according to conventional algorithmic methods. Healthcare providers, laboratories, and public health agencies considering the adoption of RST technology will have to assess whether the risk of overtreatment due to false positive results is sufficiently outweighed by the potential benefit of capturing more actionable positive cases. Congenital syphilis, in particular, is associated with significant morbidity and mortality, including pre-term birth, miscarriage, and lifelong disability, including deafness and blindness.¹ Preventing these cases can justify the use of RSTs to overcome current barriers to timely results, especially in the context of an ongoing epidemic.

Operator training and adherence to quality control can also impact test performance. A survey of STI clinic staff in one study revealed challenges with visually reading the reactivity line when whole blood was tested.^30,31s One unique aspect of our study was that the RST was performed by trained personnel in a high-complexity hospital laboratory, which can help ensure diagnostic accuracy and consistency. As this test was performed on a STAT basis, the median test turnaround time was only 1.6 hours, compared with 32.0 and 198.7 hours for in-house and send-out confirmatory testing, respectively, demonstrating its applicability to fast-paced clinical and point-of-care settings. This can enable rapid treatment, and in this study 31/43 patients with positive RSTs were treated during their hospital encounter. It is possible that more patients would have been treated if the RST were performed more quickly and directly in the emergency department. However, in a study in Detroit, Michigan, only 11/20 (55%) of patients who tested positive by point-of-care RST were treated on-site.²⁵ Another practical consideration for implementation is cost. Beyond the cost of reagents, the overall expense of testing in hospital, clinical, and public health settings also depends on anticipated test volume and existing staffing to support desired turnaround times.^32s In some cases, automated instruments already being used for other testing can also be used for treponemal and nontreponemal testing.

We note that despite taking place in a hospital setting, 45 patients with a previously documented history of syphilis still had RST ordered, even though treponemal tests are not recommended for assessing re-infection.⁴ The assay sensitivity rose from 52.2% to 83.3% when those patients were removed from the analysis. A study from the Florida state public health system similarly found that 20.8% of their tests were incorrectly ordered on patients with a prior syphilis diagnosis, even though a screening tool was employed to limit testing on those patients.²² We found that 29/33 patients who tested negative by RST but positive by EIA had a known history of syphilis, and 30/33 had low RPR titers ≤1:2. This indicates that the RST may be less sensitive for detection of prior disease compared with the EIA, particularly in treated patients with low RPR titers. It is also possible that some of these patients could have tertiary disease. The WHO considers a sensitivity of at least 80% as acceptable and at least 90% as preferred for point-of-care treponemal tests.^33s A meta-analysis comparing the performance of various rapid syphilis tests (RSTs) found that sensitivity ranged from 75% to 100% in serum.^34s The tradeoff in turnaround time for a test with lower sensitivity in some cases runs the risk of misleading patients and providers with negative rapid results. Required confirmatory testing, provider education, and decision support tools can together guide appropriate testing, reduce the risk for harm associated with missed diagnoses, and optimize assay performance.

To our knowledge, this study represents the largest real-world evaluation of the performance and clinical impact of the Syphilis Health Check RST on serum in a hospital laboratory-based setting. We acknowledge several study limitations. First, prior syphilis infection may not have been documented in all cases, particularly if a prior diagnosis was made at an outside institution and not disclosed by the patient. Second, as the RST was performed on serum by trained laboratory personnel, the findings from this work may be less generalizable to more community-based point-of-care settings utilizing fingerstick specimens. Third, the sample size was not sufficient to determine test performance in specific patient subpopulations, such as pregnant patients, patients with HIV, or patients grouped by infection stage. Likewise, we did not assess whether any patient comorbidities, such as autoimmune disease, were associated with discrepant results between the RST and EIA. Additional studies in various hospital and community settings are needed to better understand test limitations, best use case scenarios, and optimal ordering practices. Nonetheless, with careful implementation, rapid syphilis testing can serve as an effective tool in hospital settings to address the current syphilis epidemic in high-risk populations.