The study population consisted of live-born infants and their mothers from the maternity ward of Sao Lucas Hospital, associated with the Pontificia Universidade Catolica do Rio Grande do Sul School of Medicine in Porto Alegre (1.5 million inhabitants), capital of the southernmost state in Brazil. This teaching general hospital serves mainly an urban, low-income population where hospital delivery is the universal practice. In the obstetric ward, all patients are tested for Venereal Disease Research Laboratory (VDRL) as a routine procedure. For this study, performed from January 1, 1998, through December 31, 2000, we defined 3 groups of patients: group I (cases of congenital syphilis) consisting of newborn infants that met the Centers for Disease Control and Prevention surveillance case definition of congenital syphilis. 15 Group II (mothers treated for syphilis) consisted of neonates without congenital syphilis whose mothers had been adequately treated for syphilis before or during pregnancy. Group III (control subjects) was a selected random sample, from the same population, of newborn babies whose mothers never had syphilis. All live-born infants in the hospital were eligible for the study, and only those whose mothers refused or were unable to participate were excluded.
Information on characteristics of the mothers was collected during their stay in the maternity ward through a personal interview and a review of the antenatal and hospital records and included age, ethnicity (in Brazil usually defined as “color of skin”), marital status, occupation, education, monthly per capita income, number of sexual partners during pregnancy and in the last 2 years, smoking habits, alcohol consumption, drug use by the mother or partner, HIV status, obstetric history, length of gestation, number of prenatal visits, and time of the first prenatal visit. We defined prenatal care to be present if the mother had at least 2 visits and the first one occurred before 36 weeks of gestation. Variables were dichotomized for analysis.
A case-control study included group I and group III and was conducted from January 1, 1998, to December 31, 1999. Characteristics of the mother of cases were compared with those of mothers of control subjects, initially in a univariate analysis using χ2 test or Fisher exact test as appropriate, performed using Epi-Info 6.0. 16 Odds ratio (OR) was estimated with 95% confidence interval (CI). Multivariate analysis was made by logistic regression using SPSS. 17 All factors that were associated with congenital syphilis at the 5% level in the univariate analysis were included in the logistic regression, except those with possibly inaccurate estimates. An additional multivariate analysis, including only the control group, was conducted to investigate which of the maternal characteristics tended to occur together. The study on group II included patients from January 1, 1998, to December 31, 2000. Characteristics of mothers from this group were compared with those from groups I and III. The study obtained approval from the ethics committee and the mothers who participate were asked to sign an informed consent.
During the 2-year period 1998–1999, there were 7022 live births in Sao Lucas Hospital. We identified 87 cases of congenital syphilis, of which 77 mothers (88.5%) were included in the study as case mothers (3 mothers refused to participate and 7 left the hospital before the team was able to interview them). In 25 (32.5%) of the 77 infants classified as congenital syphilis, the mothers have not received prenatal care (9 of them had at least 1 contact with a medical service during pregnancy, but their syphilis was not diagnosed). In 17 cases (22%), VDRL tests were negative during pregnancy, but the mother acquired syphilis in the last weeks before delivery. The other 35 mothers (45.5%) received no or inadequate treatment as a result of other causes: in 11, penicillin was prescribed in an incorrect dose; in 7, treatment was not prescribed despite positive tests for syphilis; 6 patients refused or did not complete the treatment; 4 received other antibiotics; in 2 cases, treatment was not documented; and in 1 case, a serologic test for syphilis was not ordered prenatally.
Throughout 1998–1999, we found 29 newborns without congenital syphilis whose mothers had been adequately treated for syphilis, and we succeeded in interviewing 25 mothers. Because this group was very small for the purpose of analysis, we additionally identified 18 patients up to December 31, 2000, who fulfilled the criteria for group II, and we were able to interview 16 of their mothers. Group II was hence constituted by 41 patients. All except 1 (97.6%) of the mothers from this group have received prenatal care (the one that had not had regular prenatal care was hospitalized more than 4 weeks before delivery and was adequately treated during that period). When considering attendance to prenatal care as an exposure variable, comparison between groups I and II (only mothers who have had syphilis) shows a strong protective association of prenatal care with congenital syphilis. Fifty-two of 77 mothers had prenatal care performed in group I (67.5%) and 40 of 41 in group II (97.6%) (OR, 0.05; 95% CI, 0.00–0.39).
We had to select 264 potential controls to recruit 240 patients (90.9%) who fulfilled the criteria for inclusion in the control group (based on a ratio of 3 controls for each case), because 19 mothers refused to participate and 5 were not able to answer the interview as a result of health conditions. A random selection of all potential control subjects was made monthly according to the expected number of cases throughout the period 1998 to 1999.
The first 2 columns in Table 1 show the percent distribution of maternal characteristics in group I (cases) and group III (control subjects). The next 2 columns show the crude and adjusted ORs and 95% CIs for the case-control study that included these 2 groups. Maternal factors that remained significant in the final model were being unmarried, monthly per capita income under US $30, and number of prenatal visits less than 6. Use of illegal drugs was of borderline significance.
The last column in Table 1 shows the percent distribution of maternal characteristics in group II (mothers adequately treated for syphilis). We can observe that the frequencies of age and socioeconomic characteristics of the mothers from group II are similar to those from group III (mothers without syphilis), whereas for behavioral characteristics and marital status, mothers from group II are closer to mothers from group I (infants with congenital syphilis).
Table 2 shows statistically significant associations between maternal characteristics, including only the control group. Socioeconomic and behavioral characteristics tended to form 2 distinct blocks of characteristics. Monthly per capita income below US $30 was associated with few years of education, being black, having no prenatal care, and having fewer than 6 prenatal visits. None of the mothers with this low per capita income had a skilled occupation. Low income was not associated with HIV positivity or with high frequency of risk behaviors, like having had several sexual partners or having used illegal drugs. History of drug use by mothers or partners was associated with young age, being unmarried, smoking during pregnancy, and HIV positivity. Low maternal age was associated with being unmarried, having several sexual partners, and use of drugs.
Much has been published about risk factors for congenital syphilis, and those identified in our study are, in general, similar to what has been reported in the medical literature. 6,10–14 However, previous studies, except one, 9 have relied solely on medical chart and birth certificate review, whereas we were able to interview the mothers prospectively. This article is also original in showing that, in the same population, we find 2 maternal profiles that include distinct risk factors for congenital syphilis. Our proposed theoretical model is graphically represented in Figure 1. Although there is some overlap between the 2 arms of this dichotomous model, it is useful to help us think about why congenital syphilis occurs. The first arm included mainly socioeconomic variables (income, level of education, occupation, and color of the skin) and the second arm included mainly behavioral variables (use of illegal drugs, tobacco and alcohol use, sexual activity, marital status, and age) and HIV positivity.
The most important parameter as a risk factor for congenital syphilis was the small number of prenatal visits, which is consistent with published research. 3,6,9–14 This highlights the importance of absent or insufficient prenatal care as an independent risk factor for congenital syphilis. Comparison between groups I and II (only mothers who have had syphilis) shows that a pregnant woman that had or have syphilis and did not receive prenatal care had nearly a 20 times greater risk of delivering an infant that met the criteria of a definition of congenital syphilis.
Prenatal care is not only the best opportunity to treat gestational syphilis; it is also important for the control of a pregnant woman who had received documented adequate treatment for syphilis before pregnancy, and necessary for the interpretation of a positive serologic test at delivery. Records during prenatal control will contain the serologic status during pregnancy, and therefore knowledge of whether treatment was efficient and reinfection did not occur. If the serologic test for syphilis at delivery is positive, even at a low titer, and the serologic status during pregnancy is unknown, we cannot exclude treatment failure or reinfection, and the newborn has to be investigated and treated for congenital syphilis.
Despite the strong protective association between prenatal care and congenital syphilis, more than 50% of the mothers of case patients have received prenatal care. Even if we consider only the cases receiving prenatal care, a great proportion of the causes of failure to diagnose or treat gestational syphilis could be considered preventable. Some of these were related to noncompliance of treatment by the pregnant woman, but a substantial proportion reflect failures or mistakes on the part of prenatal caregivers, like prescription of penicillin in incorrect doses, nonprescription of treatment in dubious cases, and delay in diagnosis. Unfortunately, women with greater needs begin prenatal care later and have fewer visits, but also sometimes receive less care in priority procedures during the pregnancy. 18,19 Additionally, among the mothers who did not receive regular prenatal care, approximately 1 in 3 had some contact with a health service during pregnancy but was not evaluated for syphilis. These facts have been repeatedly reported 4,20–22 and should deserve greater attention by health providers.
The difference in the odds ratios between the univariate and the multivariate analysis demonstrates the confounding effect on the exposure factors. A risk factor could in fact be a proxy for others, and the multivariate analysis is very useful to demonstrate this. 23 The black color of the skin, for example, has been named as a risk factor for acquired and congenital syphilis. 6,13,24–27 The ethnicity could act as a proxy for other risk factors, in our population more socioeconomic than behavioral in nature. Studies that have found an independent association between black color and congenital syphilis did not include in the analysis a variable that represented the family income, making the color the main socioeconomic factor. 6,13 One of our strata was a very low per capita family income category, which was strongly associated with the presence of congenital syphilis. The highest frequency of syphilis in very poor women and their children was in general not dependent on behavioral factors, but on poor education and economic obstacles. These women were more likely to have been infected by their husbands, who are likely to not practice safe sex and refuse syphilis treatment. After being infected, the very poor pregnant women had a lower probability to receive adequate treatment, because of lack of information and difficulties to access the health system.
The association between drug addiction and congenital syphilis is well-established. 9–11,20,28,29 In our study, the frequency of use of illicit drugs could be underestimated, because it was elicited from the mother’s history and we know that some patients deny current drug use. 28 The use of tobacco and alcohol during pregnancy, like other maternal characteristics, could be indicators of a lifestyle that includes other risk behaviors.
History of some anterior stillbirth was associated with congenital syphilis, but we could not demonstrate a significant association with any of the 2 profiles described or with use of legal or illegal substances, probably as a result of the small number of cases in that category. It is worth noting that in group II, the percentage of mothers with a previous stillbirth is higher than in the other groups. Another study found a protective association of previous pregnancy loss with congenital syphilis; losses other than induced abortions might lead a woman with a desired pregnancy to seek better prenatal care and prompt treatment for syphilis. 6
A recent population-based study identified factors associated with risk behaviors for acquiring sexually transmitted diseases/AIDS among urban Brazilian women. 30 Our results show that the same maternal behavioral characteristics are associated with congenital syphilis; however, the observation of the characteristics profile of mothers from group II, who received adequate treatment and control for syphilis, gives us some important additional keys. Although the small size of group II has limited the power to demonstrate statistically significant differences from the other groups, we demonstrated that some women who had syphilis before or during pregnancy, most of them with a behavioral risk profile, did not have a child with congenital syphilis because they used the health system better and received adequate prenatal care. These women were, in general, older, less poor, and more educated than the mothers of infants with congenital syphilis.
In conclusion, our results point to the existence of 2 different maternal profiles of characteristics associated with congenital syphilis, one mainly consisting of low socioeconomic status and the other mainly consisting of behavioral risk factors. Probably the socioeconomic risk factors interfere more with prenatal care and are most responsible by the lack of adequate treatment and control for maternal syphilis. We believe that these findings can be extrapolated to other similar urban populations in which health actions should be improved for pregnant women who have a greater risk for congenital syphilis and who do not adequately use the traditional healthcare systems. Any contact of the pregnant woman with a health service should be an opportunity for a possible diagnosis and adequate treatment of gestational syphilis. The quality of prenatal care specifically directed to the poorest pregnant women and those with risk behaviors needs to improve. The strategies should be targeted to each profile to become more effective.
1. Finelli L, Berman SM, Koumans EH, Levine WC. Congenital syphilis. Bull World Health Organ 1998; 76( suppl 2): 126–128.
2. Walker DG, Walker GJA. Forgotten but not gone: the continuing scourge of congenital syphilis. Lancet Infect Dis 2002; 2: 432–436.
3. Williams PB, Ekundayo O. Study of distribution and factors affecting syphilis epidemic among inner-city minorities of Baltimore. Public Health 2001; 115: 387–393.
4. Warner L, Rochat RW, Fichtner RR, Stoll BJ, Nathan L, Toomey KE. Missed opportunities for congenital syphilis prevention in an urban southeastern hospital. Sex Transm Dis 2001; 28: 92–98.
5. Gust DA, Levine WC, Louis ME, Braxton J, Berman SM. Mortality associated with congenital syphilis in the United States, 1992–1998. Pediatrics 2002; 109: E79–E79.
6. Mobley JA, McKeown RE, Jackson KL, Sy F, Parham JS, Brenner ER. Risk factors for congenital syphilis in infants of women with syphilis in South Carolina. Am J Public Health 1998; 88: 597–602.
7. Nakashima AK, Rolfs RT, Flock ML, Kilmarx P, Greenspan JR. Epidemiology of Syphilis in the United States, 1941–1993. Sex Transm Dis 1996; 23: 16–23.
8. Mascola L. The rising incidence of congenital syphilis: back to the future. NY State J Med 1990; 90: 485–486.
9. Southwick KL, Guidry HM, Weldon MM, Mertz KJ, Berman SM, Levine WC. An epidemic of congenital syphilis in Jefferson County, Texas, 1994–1995: inadequate prenatal syphilis testing after an outbreak in adults. Am J Public Health 1999; 89: 557–560.
10. Greenberg MSZ, Singh T, Htoo M, Schultz S. The association between congenital syphilis and cocaine/crack use in New York City: a case-control study. Am J Public Health 1991; 81: 1316–1318.
11. Webber MP, Lambert G, Bateman DA, Hauser WA. Maternal risk factors for congenital syphilis: a case-control study. Am J Epidemiol 1993; 137: 415–422.
12. Bishop C, Crovari E, Murillo E. Casos y controles de la sifilis congenita en el Hospital del Niño (1989–1992). Rev Hosp Niño Panama 1992; 11: 75–81.
13. Desenclos JCA, Scaggs M, Wroten JE. Characteristics of mothers of live infants with congenital syphilis in Florida, 1987–1989. Am J Epidemiol 1992; 136: 657–661.
14. Risser WL, Hwang LY. Congenital Syphilis in Harris County, Texas, USA, 1990–92: incidence, causes and risk factors. Int J STD AIDS 1997; 8: 95–101.
15. Centers for Disease Control and Prevention. Guidelines for treatment of sexually transmitted disease. MMWR Morb Mortal Wkly Rep 1998; 47( RR-1): 1–118.
16. Dean AG, Dean JA, Coulombier D. Epi Info, version 6. Atlanta: Centers for Disease Control and Prevention, 1994.
17. SPSS release 6.0. Chicago: SPSS Inc, 1993.
18. Hansell MJ. Sociodemographic factors and the quality of prenatal care. Am J Public Health 1991; 81: 1023–1028.
19. Halpern R, Barros FC, Victora CG, Tomasi E. Prenatal care in Pelotas, Rio Grande do Sul, Brazil, 1993. Cad Saude Publica 1998; 14: 487–492.
20. Centers for Disease Control and Prevention. Epidemic of congenital syphilis—Baltimore, 1996–1997. MMWR Morb Mortal Wkly Rep 1998; 47: 904–907.
21. Ernst AA, Romolo R, Nick T. Emergency department screening for syphilis in pregnant women without prenatal care. Ann Emerg Med 1993; 22: 781–785.
22. Thompson BL, Matszak D, Dwyer DM, Nakashima A, Pearce H, Israel E. Congenital syphilis in Maryland, 1989–1991: the effect of changing the case definition and opportunities for prevention. Sex Transm Dis 1995; 22: 364–369.
23. Rothman K. The Role of Statistics in Epidemiology. Boston: Little Brown, 1986: 115–129.
24. Thomas JC, Kulik AL, Schoenback VJ. Syphilis in the South: rural rates surpass urban rates in North Carolina. Am J Public Health 1995; 85: 1119–1122.
25. St. Louis ME, Farley TA, Aral SO. Untangling the persistence of syphilis in the South. Sex Transm Dis 1996; 23: 1–4.
26. Centers for Disease Control and Prevention. Congenital syphilis—United States. MMWR Morb Mortal Wkly Rep 1999; 48: 757–761.
27. Centers for Disease Control and Prevention. Congenital Syphilis—United States, 2000. MMWR Morb Mortal Wkly Rep 2001; 50: 573–577.
28. Minkoff HL, McCalla S, Delke I, Stevens R, Salwen M, Feldman J. The relationship of cocaine use to syphilis and human immunodeficiency virus infections among inner city parturient women. Am J Obstet Gynecol 1990; 163: 521–526.
29. Sison CG, Ostrea EM, Reyes MP, Salari V. The resurgence of congenital syphilis: a cocaine-related problem. J Pediatr 1997; 130: 289–292.
30. Silveira MF, Beria JU, Horta BL, Tomasi E, Victora CG. Factors associated with risk behaviors for sexually transmitted disease/AIDS among urban Brazilian women: a population-based study. Sex Transm Dis 2002; 29: 536–541.