Patel, Sameer J. MD, MPH; Klinger, Ellen J. MPH; O'Toole, Dana BS; Schillinger, Julia A. MD, MSc
In 2007, the World Health Organization (WHO) articulated a strategy for the global elimination of congenital syphilis,1 emphasizing that congenital syphilis is a preventable disease. The WHO strategy highlighted the importance of both the availability and quality of maternal and newborn health services to limit missed opportunities for prevention. The Centers for Disease Control and Prevention has identified elimination of congenital syphilis as a “winnable battle.”2
The clinical manifestations of congenital syphilis range from asymptomatic infection to bone deformities, neurologic impairment, and stillbirth. Lack of prenatal screening and inadequate maternal therapy can lead to the diagnosis of congenital syphilis.3 Any stage of maternal syphilis infection during pregnancy may lead to intrauterine infection. Prompt penicillin therapy appropriate for maternal stage of infection is recommended on diagnosis and at least 30 days before delivery to prevent congenital syphilis.4 Serologic testing during pregnancy should be used to assess response to therapy, relapse, or reinfection. Previously described patient-level risk factors for congenital syphilis include poverty, crack cocaine drug use, lower educational status, lack of prenatal care, and black race.5–7 An increase in syphilis infections in women may presage an increase in congenital syphilis infections.
Nationally, and in New York City, rates of congenital syphilis cases were at a nadir in 2005.8 However, in the late 2000s, the national incidence of all manifestations of syphilis increased, with a congenital syphilis rate more than 10 per 100,000 live births per year from 2007 to 2009. New York City experienced a 24% increase in primary and secondary syphilis infections among women between 2008 (0.88 cases per 100,000) and 2009 (1.10 cases per 100,000).9
Because New York State laws mandate prenatal syphilis screening and treatment and follow-up of all women and newborns with syphilis, each case of congenital syphilis warrants examination to identify provider and structural factors that contributed to its occurrence. In this study, we examined newborn and maternal characteristics of congenital syphilis cases reported to the New York City Department of Health and Mental Hygiene from 2000 to 2009. We identified missed opportunities for prevention by evaluating provider compliance with the Centers for Disease Control and Prevention (CDC) and United States Preventive Task Force recommendations for management of pregnant women and newborns.4
PATIENTS AND METHODS
New York State mandates prenatal syphilis screening of pregnant women at the first prenatal visit (Public Health Law Section 23O8) and, with limited exceptions, of neonates born alive and fetuses born dead after 22 weeks of gestation (New York Codes, Rules and Regulations, Title 10, Section 69–2.2), as well as treatment and follow-up of all women and newborns with syphilis (New York Codes, Rules and Regulations, Title 10, Section 69–2.5). In addition, laboratories and physicians in New York City are mandated to report positive syphilis test results and clinical diagnoses of syphilis, respectively (New York City Health Code, codified at Rules of the City of New York, Title 24, Sections 13.O3 and 11.O3). For New York City, pregnancy status, if available, must also be reported on laboratory case reports [Rules of the City of New York, Title 24, Section 13.O3(a)(1)] and on provider case reports [Rules of the City of New York, Title 24, Section 11.05(b)].
The New York City Department of Health and Mental Hygiene reviews each reactive serologic test for syphilis reported for women of childbearing age (age 15–45 years) residing in New York City and uses previously reported serologic test results and treatment to determine if further investigation and treatment are needed. If a reported titer suggests a new infection, then the Department of Health and Mental Hygiene contacts women and their providers to assure treatment is administered and that sex partners are tested and treated to prevent repeat infection. Women with a reactive syphilis serology who are known to be pregnant are the highest priority for investigation and are tracked by a single designated staff member. For each neonate born to a woman with a reactive syphilis serology, a field worker reviews the maternal and newborn records and completes a CDC case report form, which includes information on the following: demographics; receipt and timing of prenatal care; maternal and newborn serologic testing and treatment histories; and significant clinical, laboratory, and radiographic findings. A single staff member reviews the maternal and newborn medical records and makes a determination of congenital syphilis case status using CDC definitions for probable or confirmed congenital syphilis (Table 1).3 New York City Department of Health and Mental Hygiene field staff are trained to determine the appropriateness of treatment regimen for stage of syphilis infection, and the CDC case report form allows field staff to specify when a nonpenicillin regimen was used but does not collect information on specific penicillin formulations or doses. Data are entered into a surveillance and case management registry. The registry also includes other sexually transmitted infections reported to the New York City Department of Health and Mental Hygiene. We retrospectively reviewed this registry for cases of congenital syphilis from January 1, 2000 to December 31, 2009.
For our study, race and ethnicity data were combined. Mothers were categorized as Hispanic or as the following non-Hispanic groups: black, white, Asian or Pacific Islander, Native American, and other (which included multiracial mothers). Academic hospitals were defined as hospitals primarily affiliated with medical schools. Receipt of prenatal care was defined as having at least one prenatal visit or serologic test before 30 weeks of gestation. Thirty weeks was selected to allow ample time for the clinician to have evaluated, screened, and treated mothers with reactive syphilis serology testing 30 days before a term delivery. Maternal syphilis relapse or treatment failure was defined as an increase in maternal nontreponemal titers after treatment or as a symptomatic newborn despite maternal therapy.
Missed opportunities for preventing congenital syphilis during pregnancy were defined as receipt of prenatal care plus at least one of the following: 1) lack of documented maternal treatment for syphilis infection diagnosed before the case pregnancy; 2) lack of serologic testing during the case pregnancy; 3) late maternal treatment (eg, less than 30 days before delivery); 4) maternal treatment with a nonpenicillin regimen; or 5) lack of maternal treatment for syphilis infection diagnosed during the case pregnancy. We analyzed women who received prenatal care but did not receive appropriate serologic testing or treatment to focus on provider deficits in preventing congenital syphilis. Women who did not receive prenatal care were not considered to have had a missed opportunity. Similarly, a positive serologic test after 30 weeks of gestation, after an initial negative test result earlier in pregnancy, was not considered a missed opportunity. When there were multiple missed opportunities to prevent congenital syphilis infection, we classified the cases by the earliest missed opportunity. Women who delivered twins with congenital syphilis were considered to have a single opportunity for prevention and were included in the prevention analysis only once. For women who delivered newborns with congenital syphilis over separate pregnancies, we only included the first pregnancy. When calculating case rates and describing the medical management of the newborns, we counted each unique newborn, including each newborn of a twin pair.
To evaluate the medical management of newborns with congenital syphilis, we used the CDC case report form and recommendations from the American Academy of Pediatrics to determine if diagnostic tests were performed (long bone radiographs, cerebrospinal venereal disease research laboratory testing, and cerebrospinal fluid [CSF] cell count) and if recommended penicillin therapy was administered to the newborns.10
Case rates (probable plus confirmed cases per 100,000 live births) were calculated for each year of the study period. Frequencies and percentages were calculated for categorical variables and medians and ranges were calculated for continuous variables. Although we calculated percentages that represented population parameters for New York City, we have calculated 95% confidence intervals (CIs) for generalizability outside of New York City. Odds ratios were calculated to evaluate potential associations between missed opportunities and maternal race or ethnicity, marital status, age, and delivery at an academic hospital, compared with nonacademic hospital, using multiple logistic regression. The a priori level of significance was set at P<.05. All models were run in SAS statistical software 9.5.2. The project was classified as nonresearch according to New York City Department of Health and Mental Hygiene guidelines on the use of public health surveillance data. One of the investigators was affiliated with Columbia University and the project was approved by the Institutional Review Board of the medical center.
From January 1, 2000 to December 31, 2009, 195 newborns reported to the New York City Department of Health and Mental Hygiene met the CDC surveillance case definition of probable or confirmed congenital syphilis. One hundred eighty-five women delivered one newborn with congenital syphilis, four mothers delivered twins, and one mother delivered two newborns over successive pregnancies. The congenital syphilis case rate over the course of the decade was 16.3 per 100,000 live births. As shown in Figure 1, rates decreased in the early 2000s but have shown an upward trend in the past 3 years. Fifty-four percent (95% CI 47–61%, 105 of 195) of cases were born at seven hospitals. Only 18% (95% CI 13–23%, 35 of 195) were born at academic hospitals associated with medical schools.
The median gestational age at birth was 39 weeks (range 23–42 weeks). The median birthweight was 3,145 g (range 942–4,670 g). Of the 195 newborns, two were stillborn and 193 were liveborn. Of the latter, 192 survived and 1 died in the neonatal period. Only the two stillborn fetuses met the CDC criteria for a confirmed case. The remaining 193 cases met the CDC case definition for probable cases. Only 5.6% (95% CI 2.4–8.9%, 11 of 195) of liveborn newborns had signs of congenital syphilis on physical examination, radiographs, or CSF venereal disease research laboratory testing. Of newborns who had CSF testing performed, 64% (95% CI 52–75%, 42 of 66) had an abnormal protein or cell count.
Among the 190 mothers, the median maternal age was 31.9 years (range 17.7–45.2 years). Seventy-seven percent (95% CI 71–83%, 146 of 190) were unmarried at the time of diagnosis in their newborns. Of mothers with available race and ethnicity data, 53% (95% CI 45–61%, 85 of 160) were non-Hispanic black and 36% (95% CI 28–43%, 56 of 159) were Hispanic.
Of the 190 mothers, few had primary (1%, 95% CI 0–1.6%, n=2) or secondary syphilis (2.6%, 95% CI 0.4–4.9%, n=5) infection. Twenty-six percent (95% CI 20–32%, 49 of 190) had early latent infection, 35% (95% CI 28–42%, 66 of 190) had late latent infection, and 36% (95% CI 29–43%, 68 of 190) had latent syphilis of unknown duration or syphilis of unknown stage. One hundred seventy-three mothers had syphilis first diagnosed during the case pregnancy. Of these, 21% (95% CI 15–27%, 36 of 173) were screened in the first trimester, 15% (95% CI 10–21%, 27 of 173) were screened in the second trimester, 24% (95% CI 18–31%, 42 of 173) were screened in the third trimester, and 38% (95% CI 32–47%, 68 of 173) were tested initially at the time of delivery, including 35 who received prenatal care.
After a negative test in the first or second trimester, 32 mothers had a subsequent positive serologic test result, indicating seroconversion during pregnancy. These 32 women included 16 instances of late therapy and 13 instances in which no therapy was provided within 30 days of or after delivery. Fourteen mothers or their newborns had evidence of relapse or treatment failure despite receipt of timely and appropriate therapy.
Figure 2 displays the continuum of care along which we considered missed opportunities for the 190 mothers, displayed in the chronologic order in which care might be expected to proceed: treatment of known infection before pregnancy; receipt of prenatal care during the case pregnancy; prenatal screening during case pregnancy; maternal treatment; and serologic follow-up during case pregnancy. Missed opportunities for prevention are highlighted along this continuum.
The majority (152 of 190, 80%, 95% CI 74–86%) of mothers attended at least one prenatal visit. This included 140 mothers without documented syphilis infection before pregnancy and 13 of the 17 mothers with documented syphilis infection before pregnancy. Of the 20% (95% CI 14–25%, 37 of 190) of women who did not receive prenatal care, 32 had syphilis first diagnosed during labor or immediately after delivery.
Numerous opportunities to prevent syphilis infection were missed among women who received prenatal care. Twenty-nine percent (95% CI 22–36%, 42 of 145) of mothers with no known history of syphilis infection or with previous adequately treated infection had no testing at all during the case pregnancy. Of note, eight of these mothers had gonococcal or chlamydial infection detected during pregnancy, yet they were not tested for syphilis infection. Of mothers who had timely serologic testing performed and had a positive serologic test result before 30 weeks of gestation, 13% (95% CI 5.5–18%, 12 of 103) received late penicillin therapy, 26% (95% CI 18–35%, 27 of 103) received no therapy, and 3% (95% CI 0–6.2%, 3 of 103) received inappropriate (nonpenicillin) therapy. Another 16 women received late therapy and 13 women received no therapy; however, these were not considered missed opportunities, because they were women who initially had a negative test result and seroconversion after 30 weeks of gestation. The nonpenicillin antibiotics used were erythromycin and doxycycline. Seven of the women who were inappropriately treated or untreated for known syphilis infection before their pregnancies were not tested and treated in a timely fashion during their subsequent pregnancies, despite receipt of prenatal care.
Overall, 63% (95% CI 56–71%, 96 of 152) of women who gave birth to newborns with confirmed or probable congenital syphilis had a missed opportunity for preventing congenital syphilis infection during prenatal care. Characteristics of the mothers with and without missed opportunities are listed in Table 2. Results of multiple logistic regression analysis are shown in Table 3. Age, race or ethnicity, marital status, and delivery at a nonacademic hospital were not associated with having a missed opportunity for congenital syphilis prevention.
Evaluation of case newborns was frequently incomplete. Long-bone radiographs were not performed for 46% (95% CI 39–53%, 89 of 193) of liveborn newborns. The CSF venereal disease research laboratory testing was not performed for 32% (95% CI 26–39%, 62 of 193) of newborns and CSF cell count was not performed for 47% (95% CI 40–54%, 91 of 193) of newborns. More than half (56%, 95% CI 33–79%, 10 of 18) of the newborns who had an rapid plasma reagin four-times or more their mother's titer did not have a complete diagnostic evaluation performed.
Seventeen percent (95% CI 12–22%, 33 of 193) of newborns received no treatment at all during their newborn hospitalization. Five percent (95% CI 2.1–8.3%, 10 of 193) received a nonpenicillin regimen and 5% (95% CI 1.7–7.6%, 9 of 193) received a penicillin course that was too short in duration (less than 10 days, despite clinical indication). Three percent (95% CI 0.7–5.6%, 6 of 193) of newborns had treatment initiated more than 2 weeks after birth. There was no association between the academic status of the hospital and failure to treat congenital syphilis after birth.
In our study of congenital syphilis cases in New York City over the past decade, several salient points emerge regarding patterns of infection and management of mothers and newborns. Although rare, cases of congenital syphilis in New York City have increased in recent years, similar to national trends.8 In New York City, the proportion of case mothers who received prenatal care was higher than the proportion of women nationally. However, missed opportunities were common among those who did have prenatal care, particularly with regard to antenatal screening and late and absent treatment. Newborns were rarely overtly symptomatic but commonly received inadequate evaluation and treatment. Importantly, our study described missed opportunities along a continuum of prenatal care, which should include treatment of known syphilis before conception.
The majority of the cases of congenital syphilis reported in New York City appear to be attributable to lack of provider adherence to screening laws and treatment guidelines or to some combination of provider-level and structural factors. Diagnosis of syphilis during pregnancy is dependent on interpretation of reactive serologic test results in the context of previous testing and treatment history—data that may not be fully reviewed or available at the time of a prenatal visit. Even with eventual appropriate treatment (more than 30 days before delivery), delays can result in treatment failure.11 Obstetricians may be able to obtain expert consultation from their local or state health department or to obtain a patient's serologic and treatment history from a registry, as is available in New York City, although many may not be aware of or use this resource.
Opportunities to prevent congenital syphilis existed across multiple clinical encounters, and some women were subject to multiple missed opportunities. Some women were known to have syphilis infection before conception and were inadequately treated at the time of that diagnosis. There were cases among women who had seroconversion during pregnancy that that could have been prevented if routine repeat third trimester screening had been performed. However, this is not currently mandated in New York City and may not be as cost-effective at current rates of disease, in contrast to previous decades when it was integral to congenital syphilis prevention efforts.12 At the least, our findings argue the need for careful history-taking at each prenatal visit, including asking about interim sexual partners and sexually transmitted infections, because this may identify women who may acquire infection during pregnancy.
Incomplete diagnostic evaluations of asymptomatic newborns can lead to missed diagnoses and have been associated with increased newborn mortality.13 We found that long-bone radiographs and CSF evaluations were frequently not performed on newborns born to mothers with undocumented treatment of syphilis. Even more surprisingly, a significant proportion of newborns with congenital syphilis received no treatment at all during their hospitalization. One possible explanation is that pediatricians may not be aware of the mother's syphilis status. Whereas this could be considered a provider-level factor contributing to congenital syphilis, this also may be the result of structural problems, such as poor communication between obstetrics and pediatrics staff, lack of laboratory alert values specific to obstetric patients, or lack of policies requiring documentation of maternal syphilis titers before discharging a newborn from the hospital.
There are limitations to surveillance data and, therefore, to the generalizability of our findings. Although we have records of whether women received prenatal care, we have no measure of patient adherence, and it is possible that some instances we classify as physicians' missed opportunities were not, in fact, under physicians' control. For example, women may not follow-up after their first prenatal visit, even if it occurs in the first trimester. In addition, we could not determine the duration or quality of prenatal care when present. We relied on retrospective data from a public health registry with predefined parameters and relatively few cases of congenital syphilis. Our findings are restricted to missed opportunities for prevention in newborns with probable or confirmed congenital syphilis infection, and we cannot use these data to make conclusions on the overall management of pregnant women in New York City with suspected or confirmed syphilis infection (ie, the number of congenital syphilis cases that were successfully prevented). All liveborn newborns had “probable infection” diagnosed rather than confirmed infection;” therefore, we may have included newborns who were not infected. The accessibility and quality of prenatal care may differ in other regions of the country, and the proportion of missed opportunities may differ in other parts of the United States. Finally, our methodology classified the congenital syphilis cases by the earliest missed opportunity, whereas in reality multiple opportunities may be missed for a given patient.
Our results inform provider and public health solutions that can help avert a future congenital syphilis case. Standardized forms can be used to obtain full sexual histories, including previous signs and symptoms of syphilis infection and new partners during pregnancy. Obstetricians should screen and treat for syphilis infection at the earliest possible opportunity. Public health agencies can play a critical role in preventing congenital syphilis through the timely investigation and follow-up of laboratory-reported reactive syphilis serologies and provider case reports of syphilis among women. Public health workers can help to find women who have reactive syphilis tests but who have not returned for care, and they also can help assure that providers adhere to treatment guidelines. To facilitate the role of public health agencies in preventing congenital syphilis, providers testing women for syphilis should record pregnancy status on laboratory requisition forms so that this information may be available to public health agencies receiving reports of reactive syphilis serologies, they should report cases of syphilis as required by law, and they should indicate pregnancy status. Pediatricians should discharge newborns only after the mother's syphilis status is determined and maternal and newborn therapy is provided, if needed. Because maternal treatment history affects the diagnostic evaluation of newborns, effective communication and integration of care between obstetricians and pediatricians must be maintained, including availability of medical records after hours. Providers should take advantage of access to serologic testing results contained in public health registry to facilitate timely and accurate staging of disease and appropriate therapy. It is critical that public health departments educate clinicians, particularly at hospitals where a significant proportion of congenital syphilis cases are diagnosed. Such education should address the diagnosis and treatment of syphilis, case definitions relevant to congenital syphilis, management guidelines including the importance of treating sex partners, and the changing epidemiology of syphilis, including, if available, local risk factors for infection in women of childbearing age. Access to initial prenatal care is only the first step. Access to quality care with adequate continuity is necessary to prevent congenital syphilis infections. Coordinated efforts between providers and public health departments will be necessary to prevent congenital syphilis infections.