Chlamydia trachomatis and Neisseria gonorrhea are the most common bacterial sexually transmitted infections (STIs) in the United States.1–3 In 2017, there were 748.8 cases of C trachomatis infection per 100,000 women in the South, compared with 682.9 in the country overall.3 Most infections are asymptomatic, and infection can cause maternal and neonatal complications.4–7 Studies have shown antibiotic treatment to be effective in decreasing adverse health outcomes.8–10
The Centers for Disease Control and Prevention recommends screening pregnant women younger than 25 years and older women with risk factors for C trachomatis and N gonorrhea infection.11 Risk factors include multiple sexual partners, inconsistent condom use, young age, being of non-Hispanic black race–ethnicity, and previous STI diagnoses.3,12 Nucleic acid amplification testing is preferred for screening, given its high sensitivity and specificity.11 Infection should be treated with antibiotics, followed by a test of reinfection 3–4 weeks later.13
Delayed treatment among nonpregnant women carries established risks.14 The primary objective of this study was to estimate the prevalence of C trachomatis and N gonorrhea infection and identify risk factors for not being tested and for testing positive for these infections in pregnancy in an urban teaching hospital. Our secondary objective was to explore patterns of treatment and tests of reinfection.
We conducted a retrospective cohort study of women who delivered at Grady Memorial Hospital under the supervision of Emory University clinicians. Grady Memorial Hospital is an urban teaching hospital that serves a diverse range of patients, including many who are indigent, uninsured, or underinsured.15 We included all women who delivered one or more fetuses after 20 weeks of gestation, regardless of viability, from July 1, 2016 to June 30, 2018. Women with at least one prenatal care or triage visit were considered eligible. For women with multiple pregnancy episodes in the study period, only the index delivery was included.
The electronic health record of all patients identified was reviewed for demographic and clinical characteristics. Study data were collected and managed using Research Electronic Data Capture.16 Researchers included medical and public health graduate students. All researchers were trained in abstraction and used a codebook to ensure consistency. The Emory University Institutional Review Board and the Grady Research Oversight Committee approved this study.
The primary outcomes of our study were 1) not being tested for C trachomatis and N gonorrhea infection and 2) testing positive for C trachomatis or N gonorrhea infection during pregnancy. Because women who receive care at Grady Memorial Hospital live in a high-prevalence area, all women, regardless of age, are considered at risk and are routinely screened in pregnancy. Testing is done through a single-probe nucleic acid amplification testing from a urine sample or cervical swab which offers high sensitivity and specificity as well as testing for both infections at once. A positive C trachomatis or N gonorrhea infection was defined as a positive result on nucleic acid amplification testing processed at Grady Memorial Hospital. Self-reported testing results from an outside hospital were classified as unknown. Women without a documented nucleic acid amplification testing were considered to not have had testing.
Our secondary outcomes were time to treatment and test of reinfection. Time to treatment was defined as the time from a positive nucleic acid amplification test result to the initiation of treatment before delivery. Date of treatment was defined as the date the prescription was written or medication administered. Time to test of reinfection was defined as the time from treatment initiation to subsequent nucleic acid amplification testing performed 21 or more days later. Tests conducted before 21 days after treatment were excluded owing to the potential for false-positive results due to nonviable organisms.17 Testing and treatment occurring after delivery were excluded.
Exposures of interest included age at delivery, race or ethnicity, preferred language, parity, history of chronic medical conditions, primary insurance type at admission for delivery, and diagnosis of symptomatic bacterial vaginosis during the current pregnancy. Because the majority of women in our population are non-Hispanic black, race–ethnicity was dichotomized into non-Hispanic black and other. Prenatal care adequacy was defined using the Kotelchuck index.18 History of STIs before the current pregnancy included self-reported history of C trachomatis, N gonorrhea, trichomoniasis, syphilis, human immunodeficiency virus (HIV), hepatitis B, or herpes simplex virus. Diagnosis of an STI other than C trachomatis and N gonorrhea in the current pregnancy included these same STIs. Substance use during pregnancy and history of intimate partner violence were determined by documentation in a clinic or triage note.
We used logistic regression to analyze factors associated with not being tested and for testing positive for C trachomatis or N gonorrhea infection in pregnancy. Variables that were significant in bivariate analyses at alpha less than 0.1 were entered into an adjusted model and retained regardless of significance in the adjusted model. The unknown or missing category was retained for variables missing more than 5% of data for any outcome category. The adjusted models were examined for multicollinearity using a variance inflation factor cutoff of 10. Chi-square, Fisher exact, and Kruskal-Wallis H tests were used to compare proportions and times to treatment and test of reinfection by STI diagnosis. To account for women with multiple infections, we used a shared frailty Cox proportional hazards model to investigate factors associated with time to treatment and tests of reinfection. The model was created by stepwise selection with the same variables examined in the bivariate logistic regression models. All analyses were conducted using SAS 9.4. Statistical tests were considered significant if the P-value was less than .05. We reviewed medical records to determine reasons for delays in treatment longer than 7 days and grouped these reasons into common themes.
There were 3,723 deliveries during the study period (July 1, 2016–June 30, 2018). After excluding 375 deliveries for women who did not have at least one prenatal care or triage visit and 83 subsequent deliveries for women who delivered more than once during the study period, 3,265 eligible deliveries remained (Fig. 1).
The median age at delivery was 28 years, and the majority of women were non-Hispanic black (Table 1). The majority of patients experienced less-than-adequate prenatal care. Nearly a third of women had been diagnosed with an STI before the current pregnancy, and a minority of women were diagnosed with an STI other than C trachomatis or N gonorrhea in the current pregnancy. However, testing for some of these infections, particularly herpes simplex virus and trichomoniasis, was conducted on only a subset of patients.
Among all eligible deliveries, 88 (3%) women were not tested for C trachomatis and N gonorrhea infection. In bivariate analyses, Spanish language, drug use, Medicaid insurance status, and diagnosis of an STI other than C trachomatis and N gonorrhea in this pregnancy were associated with an increase in testing. Unknown history of intimate partner violence and transfer of care or less-than-adequate prenatal care were associated with an increased odds of not being tested (Table 1). Prenatal care adequacy remained significant in the full model with an increased odds of not being tested for those who transferred care or had less-than-adequate prenatal care (adjusted odds ratio 2.83, P<.01) (Table 2). Those with Medicaid insurance and those who were diagnosed with a STI other than C trachomatis or N gonorrhea during the current pregnancy were at decreased risk.
Of the 3,177 women who were tested, 370 (12%) tested positive. Thirty-five (1%) women were diagnosed with N gonorrhea infection, 287 (9%) with C trachomatis infection, and 48 (2%) with both. These 370 women who tested positive contributed 440 unique diagnoses: 348 C trachomatis infections, 47 N gonorrhea infections, and 45 co-infections. Fifteen percent of women who tested positive had multiple infections in pregnancy: 43 had two infections, 12 had three, and one had four.
In bivariate analyses, younger age; being of non-Hispanic black race–ethnicity; alcohol or drug use; history of intimate partner violence; receiving less-than-adequate prenatal care; diagnosis of symptomatic bacterial vaginosis during the current pregnancy; being diagnosed with another STI during the current pregnancy; and STI history were each associated with testing positive (Table 3). Having a primary language other than English was associated with a decreased risk of infection. In multivariate analysis, being younger than 25 years of age, non-Hispanic black race–ethnicity, alcohol use, being diagnosed with another STI during the current pregnancy, and history of any STI before the current pregnancy remained associated with testing positive (Table 3). Having a primary language other than Spanish or English was associated with a decreased risk.
Of the 440 unique patients with C trachomatis and N gonorrhea infection identified, 95% received antibiotic treatment before delivery. The proportion treated differed by STI diagnosis (C trachomatis 96%, N gonorrhea 87%, co-infection 91%, P=.02). Time to treatment ranged from 0 to 221 days, with the majority (55%) of patients experiencing a delay of more than 1 week (Fig. 2). Treatment occurred during hospital admission for delivery in 25 of the treated patients (6%). Time to treatment did not differ by STI diagnosis (median days to treatment: C trachomatis 8, N gonorrhea 12, co-infection 15, P=.11). Among the 228 patients who experienced delays of greater than 1 week, the most common reasons included health care provider recognition and follow-up of the abnormal result (n=147, 65%) and difficulty contacting the patient (n=76, 33%). Less common reasons included delay in the patient presenting to clinic for treatment (n=19, 8%) and the patient not collecting or not taking the medication (n=18, 8%). These reasons were not mutually exclusive.
A test of reinfection was completed in 76% of all treated patients with C trachomatis and N gonorrhea infection. Among those retested, the test of reinfection was delayed by more than 1 month beyond the recommended 21–28-day period in 24% of patients (Fig. 3). Time to test of reinfection (median days to retesting: C trachomatis 37, N gonorrhea 43, co-infection 43, P=.83) and the proportion retested (C trachomatis 73%, N gonorrhea 85%, co-infection 85%, P=.06) did not differ significantly by STI diagnosis. The only demographic or clinical variable that was associated with time to treatment and tests of reinfection in bivariate analysis was increasing gestational age (in weeks) at diagnosis. In a Cox model with only gestational age, treatment hazard ratio was 1.03 (95% CI 1.02–1.04) and test of reinfection hazard ratio was 1.04 (95% CI 1.02–1.05).
A higher proportion of our study population was screened for C trachomatis and N gonorrhea infection in pregnancy (97%) compared with previous reports of screening rates of approximately 60% for both STIs based on U.S. laboratory data.19 Women with Medicaid insurance were more likely to be tested. Although factors associated with STI screening have not been specifically studied in pregnancy, our findings support previous research in nonpregnant patients, which has shown increased odds of accepting C trachomatis testing by those with public insurance.20 Patient language has also been associated with STI risk assessment, but language was not significant in our adjusted model.21
Less-than-adequate prenatal care and transfer of care were associated with not being tested for C trachomatis and N gonorrhea infection. This finding may be due in part to health care providers assuming that patients who transferred care were screened at the previous institution or patients reporting or providing documentation of previous screening. Additionally, women with inadequate prenatal care may be more likely to have other immediately pressing issues to address in a clinic visit of limited duration. Although it is unsurprising that prenatal care adequacy is associated with lack of testing, this finding reveals an opportunity to remind clinicians or implement standard checklists to ensure screening is offered.
Our measured prevalence of 11% for C trachomatis and 3% for N gonorrhea infection is higher than the national median of 8% and 1%, respectively.6 Several demographic and clinical characteristics were associated with testing positive including age, race–ethnicity, primary language, alcohol use, and history of STIs. Though younger age was a risk factor for infection, 39% of those who tested positive were 25 years of age or older. Limiting screening to risk algorithms may miss many infections, especially in similar high-risk populations.
Many women experienced delays in treatment and tests of reinfection, though these delays were not associated with demographic or clinical characteristics with the exception of gestational age at diagnosis. This association may reflect that testing later in pregnancy more often occurs in triage rather than during scheduled prenatal care visits at our institution, and the lack of continuity makes timely follow-up of abnormal results more difficult. Moreover, during the study period, results from triage visits were not sent directly to health care providers' inboxes as clinic test results were. Though the majority of patients were treated, fewer patients with N gonorrhea infection were treated compared with C trachomatis alone, possibly due to the need to present in person for intramuscular ceftriaxone.
Information from this study is being used at our hospital for health care provider education and to help streamline processes for identification and treatment of abnormal lab results. Results from triage tests are now sent to the ordering provider's inbox. Point-of-care testing may be an opportunity to expedite treatment by eliminating the major barriers to timely treatment identified in this study. There are several commercially available point-of-care testing modalities available for C trachomatis and N gonorrhea infection, and their use has been shown to decrease time to treatment in low-resource settings.22 These tests are highly specific, but sensitivity varies widely, ranging from 13% to greater than 90%.23–27 Emerging technology is focused on offering point-of-care tests with high sensitivity and specificity and validating their use and cost-effectiveness in the clinical setting.26,27 Moreover, fewer than a third of treated patients in our study received a test of reinfection during the recommended timeframe, and 15% of women diagnosed with C trachomatis or N gonorrhea infection had more than one infection during their pregnancies. Retesting after treatment and partner treatment are critical, and expedited partner therapy should be considered in appropriate settings.28
This study has some important limitations. First, we focused on a high-risk, medically underserved population, and, although our findings may not be generalizable to all pregnant women, they are relevant to understanding what factors associated with STI screening and positivity may be modifiable among women at greatest risk for STIs. Moreover, prenatal care is delivered at our hospital by trainees and rotating residents and our findings may not be generalizable outside of similar academic practices. Additionally, we excluded tests of reinfection before 21 days after treatment. However, a subset of women had a negative test of reinfection during this time period, and this may affect our evaluation of time to test of reinfection. Finally, reasons for delays in treatment were abstracted from the medical record and may not be fully understood. Strengths of our study include the large sample size and detailed information on patient demographics, clinical history, and STI diagnoses and treatment obtained from medical record review. Although the sample size was overall high, subsets of the analysis dealt with small numbers, raising the issue of decreased power for nonsignificant associations and possible over-fitting. To address this, we reran the adjusted models excluding variables with low counts and found that all adjusted point estimates changed by less than 10%.
This study highlights the need for continued STI risk assessment and follow-up of positive results, especially in high-risk populations. It adds to the existing STI literature by describing patterns of positivity, treatment, and retesting in a pregnant population and identifying potential avenues for improved clinical management. Strong promotion of prenatal care and proactive outreach is important to improve prenatal care adequacy as well as STI detection and treatment. Given the effect of treatment on reducing risks for adverse pregnancy outcomes, routine screening in areas with high STI prevalence is imperative.
1. Expedited partner therapy. ACOG Committee Opinion No. 737. American College of Obstetricians and Gynecologists. Obstet Gynecol 2018;131:e190–3.
2. Ruhl C. Update on chlamydia and gonorrhea screening during pregnancy. Nurs Womens Health 2013;17:143–6.
3. Centers for Disease Control and Prevention. Sexually transmitted disease surveillance 2017. Available at: https://www.cdc.gov/std/stats17/default.htm
. Retrieved December 28, 2018.
4. Andrews WW, Goldenberg RL, Mercer B, Iams J, Meis P, Moawad A, et al. The Preterm Prediction Study: association of second-trimester genitourinary chlamydia infection with subsequent spontaneous preterm birth. Am J Obstet Gynecol 2000;183:662–8.
5. Alger LS, Lovchik JC, Hebel JR, Blackmon LR, Crenshaw MC. The association of Chlamydia trachomatis, Neisseria gonorrhoeae, and group B streptococci with preterm rupture of the membranes and pregnancy outcome. Am J Obstet Gynecol 1988;159:397–404.
6. Centers for Disease Control and Prevention. STDs and pregnancy—CDC fact sheet. Available at: www.cdc.gov/std/pregnancy/stdfact-pregnancy.htm
. Retrieved December 28, 2018.
7. Rours GI, Duijts L, Moll HA, Arends LR, de Groot R, Jaddoe VW, et al. Chlamydia trachomatis infection during pregnancy associated with preterm delivery: a population-based prospective cohort study. Eur J Epidemiol 2011;26:493–502.
8. Cohen I, Veille JC, Calkins BM. Improved pregnancy outcome following successful treatment of chlamydial infection. JAMA 1990;263:3160–3.
9. Ryan GM Jr, Abdella TN, McNeeley SG, Baselski VS, Drummond DE. Chlamydia trachomatis infection in pregnancy and effect of treatment on outcome. Am J Obstet Gynecol 1990;162:34–9.
10. Martin DH, Eschenbach DA, Cotch MF, Nugent RP, Rao AV, Klebanoff MA, et al. Double-blind placebo-controlled treatment trial of Chlamydia trachomatis endocervical infections in pregnant women. Infect Dis Obstet Gynecol 1997;5:10–7.
11. Workowski KA, Bolan GA. Sexually transmitted diseases treatment guidelines, 2015. MMWR Recomm Rep 2015;64:1–137.
12. Datta SD, Sternberg M, Johnson RE, Berman S, Papp JR, McQuillan G, et al. Gonorrhea and chlamydia in the United States among persons 14 to 39 years of age, 1999 to 2002. Ann Intern Med 2007;147:89–96.
13. Centers for Disease Control and Prevention. 2015 sexually transmitted diseases treatment guidelines. Available at: https://www.cdc.gov/std/tg2015/default.htm
. Retrieved December 28, 2018.
14. Reekie J, Donovan B, Guy R, Hocking JS, Kaldor JM, Mak D, et al. Risk of ectopic pregnancy and tubal infertility following gonorrhoea and chlamydia infections. Clin Infect Dis 2019;69:1621–3.
15. Jamieson DJ, Haddad LB. What obstetrician-gynecologists should know about population health. Obstet Gynecol 2018;131:1145–52.
16. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009;42:377–81.
17. Renault CA, Israelski DM, Levy V, Fujikawa BK, Kellogg TA, Klausner JD. Time to clearance of Chlamydia trachomatis ribosomal RNA in women treated for chlamydial infection. Sex Health 2011;8:69–73.
18. Kotelchuck M. The adequacy of prenatal care utilization index: its US distribution and association with low birthweight. Am J Public Health 1994;84:1486–9.
19. Blatt AJ, Lieberman JM, Hoover DR, Kaufman HW. Chlamydial and gonococcal testing during pregnancy in the United States. Am J Obstet Gynecol 2012;207:55.e1–8.
20. Playforth KB, Coughlan A, Upadhya KK. The association between insurance status and acceptance of Chlamydia screening by teenagers who present for preventive care visits. J Pediatr Adolesc Gynecol 2016;29:62–4.
21. de Bocanegra HT, Rostovtseva D, Cetinkaya M, Rundel C, Lewis C. Quality of reproductive health services to limited English proficient (LEP) patients. J Health Care Poor Underserved 2011;22:1167–78.
22. Guy RJ, Ward J, Causer LM, Natoli L, Badman SG, Tangey A, et al. Molecular point-of-care testing for chlamydia and gonorrhoea in Indigenous Australians attending remote primary health services (TTANGO): a cluster-randomised, controlled, crossover trial. Lancet Infect Dis 2018;18:1117–26.
23. Guy RJ, Causer LM, Klausner JD, Unemo M, Toskin I, Azzini AM, et al. Performance and operational characteristics of point-of-care tests for the diagnosis of urogenital gonococcal infections. Sex Transm Infect 2017;93:S16–21.
24. Kelly H, Coltart CEM, Pant Pai N, Klausner JD, Unemo M, Toskin I, et al. Systematic reviews of point-of-care tests for the diagnosis of urogenital Chlamydia trachomatis infections. Sex Transm Infect 2017;93:S22–30.
25. Gaydos CA. Review of use of a new rapid real-time PCR, the Cepheid GeneXpert(R) (Xpert) CT/NG assay, for Chlamydia trachomatis and Neisseria gonorrhoeae: results for patients while in a clinical setting. Expert Rev Mol Diagn 2014;14:135–7.
26. Herbst de Cortina S, Bristow CC, Joseph Davey D, Klausner JD. A systematic review of point of care testing for Chlamydia trachomatis, Neisseria gonorrhoeae, and trichomonas vaginalis. Infect Dis Obstet Gynecol 2016;2016:4386127.
27. Jacobsson S, Boiko I, Golparian D, Blondeel K, Kiarie J, Toskin I, et al. WHO laboratory validation of Xpert((R)) CT/NG and Xpert((R)) TV on the GeneXpert system verifies high performances. APMIS 2018;126:907–12.
28. Haddad LB, Jamieson DJ. Let us be proactive rather than reactive: the time has come to implement expedited partner therapy. Obstet Gynecol 2019;133:413–15.