Frequency of pregnant women with HBsAg in a Brazilian community1

Geraldo Duarte,2 Marisa Márcia Mussi-Pinhata,3 Roberto Martinez,4 Cláudia Lemos,2 Eduardo M. Leite Figueiredo,2 & Silvana Maria Quintana2

 SUMMARY The work reported here points up the real benefits provided by neonatal immunoprophylaxis of newborns delivered by mothers who are seropositive for the hepatitis B virus surface antigen HBsAg and underscores the need to properly identify such mothers in Brazil so that immunoprophylaxis can be undertaken. To help determine levels of hepatitis B virus (HBV) infection and seropositivity for various HBV markers among pregnant women in Southeast Brazil, investigators studied 7992 pregnant women delivering at the Clinical Hospital of the University of São Paulo's Ribeirão Preto School of Medicine in Ribeirão Preto, Brazil. Seroreactivity for HBsAg was determined first by serologic screening with an enzyme-linked immunosorbent assay (ELISA) procedure in which the sera were incubated for 2 hours and then by confirmation with another ELISA in which the sera were incubated for 18 hours. Subsequently, tests for anti-HBsAg, HBeAg, anti-HBeAg, and anti-HBcAg markers were conducted using confirmed positive samples. Initial screening found 84 of the 7992 samples (1.05%, 95% CI: 0.84-1.30) to be positive for HBsAg; however, this HBsAg positivity was confirmed in only 76 (0.95%, 95% CI: 0.75-1.19). The positivity rate was significantly higher among subjects whose pregnancies terminated in miscarriage (1.84%) than among those with live births (0.83%) (c2, Yates correction = 7.6; P = 0.005). Anamnesis was able to identify HBV risk factors in only 27.6% of the confirmed HBsAg-positive subjects or close household contacts. However, 21.3% (95% CI: 1.04-30.56) of the confirmed HBsAg-positive subjects were found positive for HBeAg, indicating a high risk of vertical transmission of the virus. These results demonstrate a need to conduct specific serologic research at term, in order to provide effective neonatal immunoprophylactic benefits.

It is known that hepatitis caused by B virus (HBV) occurs with greater frequency among young adults whose occupations, lifestyles, or behaviors (particularly illicit intravenous drug use and unprotected homosexual or heterosexual activities) place them at increased risk of contracting the infection. Among other things, HBV transmission through heterosexual relations accounts for some 25% of all hepatitis B cases, a higher percentage than that transmitted through homosexual relations (1-3). More specifically, in the case of Brazil the high level of heterosexual transmission has resulted from an increase in the number of HBV-infected women of reproductive age (4).

One of the most important routes of HBV transmission and sources of HBV perpetuation in the community, of concern to both pediatricians and obstetricians, is mother-child transmission (5, 6). Transcending specific epidemiology, such perinatal HBV transmission can occur as a result of either acute or chronic maternal infection. Other researchers (7) have found the risk of vertical HBV transmission to be 71% when acute maternal infection occurs at the end of pregnancy versus 3% when it occurs at the onset of pregnancy. It has also been shown that the fetuses or newborns of women who are chronic HBsAg and HBeAg carriers are more susceptible to vertical transmission (80-90% risk), a finding that highlights the importance of assessing all the serologic markers of HBV infection (8, 9).

Approximately 6% of the cases of perinatal HBV transmission occur during pregnancy (10), while the remainder occur during or shortly after birth. This high prevalence of relatively late transmission permits immunoprophylaxis initiated after pregnancy to serve as an effective measure for reducing neonatal infectionas well as preventing acute and chronic complications (11, 12).

According to data released by the World Health Organization's Consulting Group on Hepatitis (13), active immunoprophylaxis is indicated for all newborn babies in communities where the prevalence of HBV is intermediate or high (over 2%). In the meantime, until universal prophylaxis is instituted in communities where HBV prevalences are high and also where the prevalences are below 2%, the viable alternative is detection of pregnant women who are HBsAg carriers and provision of prophylaxis, a procedure that not only determines the seroprevalence of infection but also permits development of regional strategies for addressing the problem and selective application of neonatal HBV immunoprophylaxis (11). This measure is directed at reducing and controlling vertical HBV transmission (14).

Although identifying pregnant women at risk of hepatitis B based on their clinical histories used to be controversial, there is no longer any doubt that serologic screening for HBsAg is clearly superior to anamnesis for this purpose (10, 15-19), or that it shows a cost-benefit ratio that supports its use (14, 20-22).

In Brazil, HBV infection is most prevalent in the Amazon region, reaching levels as high as 12% in some communities (23). The prevalence is lower in the southeastern area, ranging from 0.1% to 4.0% (24). However, information about the HBV prevalence among pregnant or postpartum women in Brazil is scanty (25), which is one of the prime reasons for the study reported here.

The aims of this study were as follows: (1) to determine the rate of HBsAg seropositivity among women whose pregnancies were resolved at the Clinical Hospital of the University of São Paulo's Ribeirão Preto School of Medicine in Ribeirão Preto, Brazil; (2) to assess the contribution of anamnesis to the detection of HBV risk factors among women carrying HbsAg; and (3) to confirm the accuracy and significance of a positive test for HBsAg by conducting tests to verify the presence of this antigen in the serum and to detect other serologic markers of HBV infection.

MATERIALS AND METHODS

The study was conducted at the Clinical Hospital of the Ribeirão Preto School of Medicine from 1 March 1991 through 31 December 1993. During this period the 7992 women whose pregnancies were resolved at the hospital were examined for seroreactivity to HBV surface antigen (HBsAg).

Detection of HBsAg seroreactivity was achieved in two steps (screening, followed by confirmatory testing) using an enzyme-linked immunosorbent assay (Auzyme ELISA, Abbott Laboratories). In the first procedure (screening), the serum samples tested were incubated for two hours. To exclude possible false-positives, the samples testing positively were then retested using the same ELISA procedure with a longer incubation period (18 hours). In addition, samples that were repeatedly positive for HBsAg were tested for other serologic markers of HBV infection using other immunoenzymatic methods: anti-HBsAg (Ausab EIA, Abbott Laboratories, procedure B), anti-HBcAg (Corzyme, Abbott Laboratories, procedure B), HBeAg and anti-HBeAg (HBe-rDNA-EIA, Abbott Laboratories.)

The screening test results were available six to 18 hours after delivery, which made it possible to provide active/passive immunoprophylaxis for the newborns (12). Even when maternal reactivity was not confirmed, the opportunity to complete immunization of the infant was always taken, with additional booster vaccinations being scheduled for 1 and 6 months of age. Cases of confirmed reactivity were evaluated individually, and the women involved were informed about the importance of contraception, use of condoms, and seeking out close contacts (both sexual partners and others) for purposes of vaccination.

Routine evaluation of all new mothers included the gathering of information regarding their professions, habits, sexual behaviors, and illnesses both before and during pregnancy. The following clinical information was obtained from patients identified as HBsAg carriers:

• Past personal illnesses (hepatitis, chronic renal insufficiency, and other sexually transmitted diseases);
• Problems during the current pregnancy (hepatitis, other sexually transmitted diseases, renal insufficiency, jaundice and/or choluria);
• Professional or behavioral risks (being a health care provider, day care center worker, or sex worker; use of illicit intravenous drugs; alcoholism; imprisonment).

The relationship between seroreactivity for HBsAg and the way in which pregnancy terminated was tested by means of the chi-square method, assuming a significance level of a = 5%. The results were expressed in 95% confidence intervals (CI).5

RESULTS

As Table 1 shows, the initial screening test found 84 serum samples from the 7992 study women (1.05%, 95% CI: 0.84-1.30) to be positive. The confirmatory test found 76 of these 84 samples to be positive (0.95% of the total screened, 95% CI: 0.75-1.19). The remaining 8 samples were negative, indicating a false-positive rate of 9.5% (95% CI: 4.2-17.9) for the initial tests employing the shorter incubation period.

The Table 2 data indicate that the prevalence of HBsAg positivity was higher among women whose pregnancies ended in miscarriage (1.84%, 95% CI: 1.11-2.99) than among those whose pregnancies ended in successful delivery (0.83%, 95% CI: 0.64-1.08). This difference was statistically significant (c2 = 7.76; P = 0.005).

The routine anamnesis following resolution of pregnancy identified situations and behaviors placing 15 women confirmed positive for HBsAg and 10 close household contacts of women confirmed positive at risk of HBV infection. Since there was some overlapping (in four cases a woman participating in the study and a close household contact of the same woman were identified as being at risk), the anamnesis identified only 21 (27.6%) of the 76 women confirmed seropositive. Among the risk factors, illicit intravenous drug use was the one most frequently identified among the pregnant women carrying HBsAg, and prior diagnosis of viral hepatitis was the one most frequently identified among their contacts (Table 3). No cases were detected where the infected women were health professionals, day care center employees, prostitutes, or prisoners.

Table 4 summarizes the results of tests for other serologic markers of HBV infection. For technical reasons, it was not possible to subject all 76 HBsAg-positive samples to all of the other test procedures. However, the presence of anti-HBcAg confirmed the presence of HBV infection in 65 (90.3%) of 72 cases found repeatedly positive for HBsAg. On the other hand, the anti-HBsAg marker was found in only 3 (4.5%) of 67 sera from confirmed HBsAg-positive cases; and the HBeAg marker was found in 16 (21.3%) of 75 sera from confirmed HBsAg-positive cases.

DISCUSSION AND CONCLUSIONS

The medical, health, and economic importance of vertical HBV transmission (the primary form of transmission of this virus during the first decade of life) is undeniable (26). Among other things, it is known that most children infected with HBV early in life become chronic carriers of the infection capable of spreading it to the community (8, 9); and chronic HBV infection is known to cause cirrhosis and hepatocellular carcinoma in adults (27). In a large number of cases, these complications could be prevented by neonatal immunoprophylaxis, a measure that has proved beneficial to both children and their communities.

In the current case, where a positive screening result was obtained in 1.05% and confirmed in 0.95% of the 7992 study women, the screening test made it possible to adopt measures such as active and passive neonatal immunoprophylaxis in the first hours after delivery, a time at which the efficacy of this treatment is more evident than at others (28). A negative confirmatory test could have precluded continuation of the immunization schedule, but instead the opportunity was taken to carry on and complete the newborns' immunization. In this study the confirmatory tests made it possible to effectively identify cases where it was necessary to search for contacts. Contacts of the infected study subjects received counseling on HBV serologic evaluation and anti-HBV vaccination. Sexual partners received additional information about use of condoms during sexual intercourse until the occurrence of postvaccination seroconversion. And both the mothers and their partners were instructed about the need to receive periodic booster vaccinations every 5-10 years, because of the potential decline in protective antibody levels over time (12).

Considering that 9.5% of those found positive in the screening test for HBsAg at the time of delivery were not confirmed by confirmatory testing, it would appear logical to focus research efforts on the prenatal period, thus avoiding both passive and active immunoprophylaxis in unconfirmed cases and markedly reducing costs. However, the low utilization of prenatal services by women in Brazil constitutes a constraint upon implementation of such a strategy.

As previously noted, the observed prevalence of HBV carriers (HBsAg seropositives) was significantly higher among women whose pregnancies terminated in miscarriage (1.84%) than among those who delivered live infants (0.83%). The clinical hospital attended by the study population sees a high number of miscarriages induced by low-income women who are also exposed to relatively high risks of contracting sexually transmitted diseasesa circumstance that could account for the higher observed HBV infection rates among the study women who miscarried.

The small percentage (27.6%) of infected women in whom HBV risk factors or behaviors were identified by anamnesis is in accord with most of the relevant studies in the literature (15, 19)a circumstance indicating that no such epidemiologic measure can replace serologic screening. Part of the discrepancy observed is associated with omission of information about situations and behaviors currently being repressed by society, such as use of illicit drugs and sexual promiscuity (29). Another partial explanation is that many patients may not know that their partners are or have been exposed to high risk of infection.

While hepatitis is not as discrimination-provoking as the acquired immunodeficiency syndrome (AIDS), the two diseases share several common features. For instance, the current study found intravenous drug use (a major AIDS risk factor) to be the risk factor most frequently encountered among women infected with HBV. (Other leading risk factors were a history of hepatitis, jaundice, and HIV infection.) Again as in the case of AIDS, alcoholism was considered a risk factor, because alcoholics are more prone than nonalcoholics to intravenous drug use and sexual promiscuity.

With regard to serologic markers of HBV infection other than HBsAg, the finding of anti-HBcAg in 65 (90.3%) of the samples from 72 of the 76 women repeatedly positive for HBsAg suggests that the test for this marker could be used as an alternative in screening for maternal HBV infection. Special attention should also be given to the fact that 16 (21.3%) of the 75 HBsAg-positive samples tested were found positive for HBeAg.

Even though some authors recommend universal screening of all obstetric populations for HBs antigens (30), it is important to stress that routine screening yields a better cost-benefit situation when the infection is endemic or when a selective screening program fails to isolate a significant number of HBV carriers among the general population. Population data for the United States indicate that universal screening appears to be cost-effective (considering both direct and indirect costs) when the prevalence of the carrier state exceeds 0.06% (31). However, in developing countries it is essential to balance the cost-efficiency aspects and the potential impact of a screening policy aimed at pregnant women against other primary health care problems that continue unresolved.

The World Health Organization (WHO) recommends that for populations with medium to high prevalences (exceeding 2%) of chronic HBV infection, appropriate intervention should consist of early vaccination of the newbornsince transmission occurs primarily among lactating and very young babies. In such circumstances, the vaccination schedule should begin shortly after birth, principally in hyperendemic areas (where the prevalence of chronic HBV exceeds 10%). In areas with prevalences below 2%, WHO recommends including the option of screening pregnant women and actively immunizing the newborn babies of infected mothers, realizing that simultaneous administration of immunoglobulin against hepatitis B could present operational and cost problems (13).

In the specific obstetric population studied, even though the prevalence of HBV-positive women (0.95%) may be considered low, the high observed prevalence of HBeAg carriers and the failure to detect anti-HBeAg antibodies in 34.7% of 75 who were confirmed positive for HBsAg points to a significant opportunity for vertical transmission of the virus. Virtually all those born to mothers carrying HBeAg will be infected at birth, and 90% of the infected babies will become chronic carriers (32-34). If one considers that some 70% of the infections in the confirmed positive study women would not have been detected through a routine clinical anamnesis (which would have permitted maternal-child HBV transmission intervention at the earliest opportunity), universal screening for HBsAg would appear justifiable in this population. Since the cost of each routine screening test is approximately US$5.00, as compared to US$ 99.00 for the three-dose vaccination series, such screening would provide an alternative to routine vaccination of all newborns. At present, obstetric screening at the time of delivery is operationally feasible in the study community, despite the additional costs and effort implied by the screening test's 9.5% false-positive rate. It should be noted, however, that routine vaccination could become a more economical alternative if mass vaccination of infants were implemented in Brazil and if vaccination costs were reduced, which has not yet happened.

Maximum protection against vertical transmission of HBV is achieved when a combination of hyperimmune HBV human immunoglobulin (HBIG) and vaccine is administered shortly (not more than 48 hours) after birth. This treatment has a protective efficacy superior to 90-95%, even when the mother carries the HBe antigen and does not have anti-HBe antibodies (33). Use of hepatitis B vaccine alone for prevention of vertical transmission has rates of protective efficacy in the range of 70-85%, depending on the frequency of the HBeAg marker in the obstetric population (35). Considering that no anti-HBe antibodies were detected in 26 (34.7%) of 75 women confirmed positive for HBsAg, the safest intervention for this population was general inclusion of HBIG in all cases of neonatal prophylaxis. This procedure was adopted in the study and continues to be used at the hospital. Due to its high cost, however, it is not a measure that could easily be extended to all regions of the country.

We believe that this study has made an important contribution to the existing pool of information about the prevalence of HBV carriers and corresponding serologic markers among obstetric populations in Southeastern Brazil. Dissemination of its findings should help to alert epidemiologists, obstetricians, and neonatologists about the need to conduct assessments of their patients in their communities and to allocate resources for planning public health measures at the regional level. On the basis of these findings, it is proposed that routine screening of HBV carriers be instituted as early as possible by means of HBsAg detection, followed by both active and passive immunoprophylaxis when economic and operating conditions permit. The early immunization of all newborn infants, as is proposed for developed countries (36), is still costlier than serologic screening of pregnant women and seems to us unsuited to the present reality of Brazil and other Latin American countries.

REFERENCES

1. Alter MJ, Coleman PJ, Alexander WJ, Kramer E, Miller JK, Mande E, et al. Importance of heterosexual activity in the transmission of hepatitis B and non-A, non-B hepatitis. JAMA 1989;262:1201-1205.

2. Alter MJ. Heterosexual activity: a leading risk factor in the transmission of hepatitis B. In: Piot P, Andre FE, eds. Hepatitis B: a sexually transmitted disease in heterosexuals. New York: Excerpta Medica; 1990:17-22.

3. Piot P, Andre FE, eds. Hepatitis B: a sexually transmitted disease in heterosexuals. New York: Excerpta Medica; 1990.

4. Duarte G. Doenças sexualmente transmissíveis durante o ciclo grávido-puerperal. In: Morais EN, ed. Temas de obstetrícia. São Paulo: Roca; 1992:385-406.

5. Snydman DR. Hepatitis in pregnancy. N Engl J Med 1985;313:1398-1401.

6. Beat SV, Boxall EH, Watson RM, Tarlow MJ, Kelly DA. Fulminant hepatitis B in infants born to anti-HBe hepatitis carrier mothers. Br Med J 1992;304:1169-1170.

7. Tong MJ, Thusby M, Rakela J, McPeack C, Edwards VM, Mosley JW. Studies on the maternal-infant transmission of viruses which cause acute hepatitis. Gastroenterology 1981; 80:999-1004.

8. Lidman K, Magnius L, Sterner G. Viral hepatitis in pregnant women at term. Scand J Infect Dis, Suppl 1990;71:39-44.

9. Schaefer E, Koeppen H, Wirth S. Low level virus replication in infants with vertically transmitted fulminant hepatitis and their anti-HBe positive mothers. Eur J Pediatr 1993; 152:581-584.

10. Summers PR, Biswas MK, Pastorek II JG, Pernoll ML, Smith LG, Bean BE. The pregnant hepatitis B carrier: evidence favoring comprehensive antepartum screening. Obstet Gynecol 1987;69:701-704.

11. Hall CB, Halsey NA. Control of hepatitis B: to be or not to be? Pediatrics 1992;90:274-277.

12. West DJ, Margolis HS. Prevention of hepatitis B virus infection in the United States: a pediatric perspective. Pediatr Infect Dis J 1992; 11:866-874.

13. Anonymous. WHO progress in the control of viral hepatitis: memorandum from a WHO meeting. Bull World Health Organ 1988;66: 443-455.

14. Ginsberg GM, Berger S, Shouval D. Cost-benefit analysis of a nationwide inoculation programme against viral hepatitis B in an area of intermediate endemicity. Bull World Health Organ 1992;70:757-767.

15. Soulie JC, Larsen M, Goudeau A, Parnet F, Dubois F, Pinon F, et al. Perinatal transmission of the hepatitis B virus in Paris region. Pathol Biol 1991;39:264-270.

16. Mateo JR, Sever JL. Perinatally acquired infections and screening. Curr Opin Obstet Gynecol 1990;2:662-667.

17. Okun NB, Larke RP, Waters JR, Joffres MR. Success of a program of routine prenatal screening for hepatitis B surface antigen: the first two years. Can Med Assoc J 1990;143: 1317-1321.

18. Brook MG, Lever AM, Kelly D, Rutter D, Trompeter RS, Griffith P, et al. Antenatal screening for hepatitis B is medically and economically effective in the prevention of vertical transmission: three years' experience in a London hospital. J Med 1989;71:313-317.

19. Greenspoon JS, Martin J, Greenspoon RL, McNamara BT. Necessity for routine obstetric screening for hepatitis B surface antigen. J Reprod Med 1989;34:655-658.

20. Thomas IL. Cost effectiveness of antenatal hepatitis B screening and vaccination of infants. Aust N Z J Obstet Gynecol 1990;30: 331-335.

21. Arizmendi JFJE, Ruiz CTR, Alonso EB, Mansanet MLM, Rico EV, Ramos MDP, et al. Metodología y rentabilidad de la utilización de un protocolo de despistaje del AgHB materno del parto. Enf Trans Sex 1992;6: 107-111.

22. Soulie JC, Goudeau A, Larsen M, Parnet F, Dubois F, Pinon F, et al. Prevention of perinatal hepatitis B virus transmission. Epidemiology and cost/efficacy ratio in the Paris region. Press Med 1991;20:939-944.

23. Juarez E, Amaral C Jr, Pinheiro F, Biellik R. Hepatitis B and delta vaccination program in the Western Amazon region of Brazil. Prog Clin Biol Res 1991;364:343-349.

24. Passos AD, Gomes UA, Figueiredo JF, Nascimento MM, Oliveira JM, Gaspar AM, et al. Influência da migração na prevalencia de marcadores sorológicos de hepatite B em comunidade rural. Análise da prevalencia segundo local de nascimento. Rev Saude Publica 1993;27:30-35.

25. Sabino EC, Guerra EM, Oba IT, Spina AMM, Vaz AJ. Freqüencia de marcadores de hepatite B em gestantes de primeira consulta em centros de saúde da área metropolitana de São Paulo, Brasil. Rev Inst Med Trop São Paulo 1992; 34:535-541.

26. Stevens CE, Beasley RP, Tsui J, Lee WC. Vertical transmission of hepatitis B antigen in Taiwan. N Engl J Med 1975;292:771-774.

27. Beasley RP. Hepatitis B virus. The major etiology of hepatocellular carcinoma. Cancer 1988; 61:1942-1956.

28. Sehgal A, Gupta I, Bhakoo ON, Ganguly NK. Use of hepatitis B vaccine alone or in combination with hepatitis B immunoglobin for immunoprophylaxis of perinatal hepatitis B infection. J Trop Pediatr 1992;38:247-250.

29. Duarte G, Mussi-Pinhata MM, Del Lama J, Takeda E, Pasti MJ, Costa JC. Valor de questionário específico na identificação de parturientes de risco para infecção pelo vírus da imunodeficiencia humana (HIV). J Bras Ginecol 1991;101:169-174.

30. Kumar ML, Dawson NV, McCullough AJ, Radiovoyervitch M, King KJ, Hertz R, et al. Should all pregnant women be screened for hepatitis B? Ann Intern Med 1987;107:273.

31. Arevalo JA, Washington AE. Cost-effectiveness of prenatal screening and immunization for hepatitis B virus. JAMA 1988;259:365.

32. Lee CY, Huang LM, Chang MH, Hsu CY, Wu SJ, Sung JL, et al. The protective efficacy of recombinant hepatitis B vaccine in newborn infants of hepatitis B and antigen-positive-hepatitis B surface antigen carrier mothers. Pediatr Infect Dis J 1991;10:299-303.

33. Poovorawan Y, Sanpavat S, Pongpunglert W, Sentrakul P, Vandepapelière P, Safary A. Long-term efficacy of hepatitis B vaccine in infants born to hepatitis Be antigen-positive mothers. Pediatr Infect Dis J 1992;11:816-821.

34. Ip HM, Wong VCW, Lelie PN, Kuhns MC, Reesing HW. Prevention of hepatitis B virus carrier state in infants according to maternal serum levels of HBV DNA. Lancet 1989;1: 406-409.

35. Wheeley SM, Boxall EH, Tarlow MJ, Gatrad AR, Anderson J, Bissenden J, et al. Hepatitis B vaccine in the prevention of perinatally transmitted hepatitis B virus infection: final report on a West Midlands pilot study. J Med Virol 1990;30:113-116.

36. Centers for Disease Control and Prevention. Hepatitis B virus: a comprehensive strategy for eliminating transmission in the United States through universal childhood vaccination: recommendations of the Immunization Practice Advisory Committee. MMWR 1991; 40(RR-13):1-25.

Manuscript received on 11 May 1994. Accepted for publication (following revision) in Portuguese in the Boletín de la Oficina Sanitaria Panamericana on 28 August 1995 and for publication in English in the Revista Panamericana de Salud Pública/Pan American Journal of Public Health on 4 October 1995.

1 The work reported here was conducted in Brazil at the Clinical Hospital of the University of São Paulo's Ribeirão Preto School of Medicine. This article has previously been published in Portuguese in the Boletín de la Oficina Sanitaria Panamericana, Vol. 120, No. 3, 1996, pp. 189-196.

2 Department of Gynecology and Obstetrics, Ribeirão Preto School of Medicine, University of São Paulo, Ribeirão Preto, São Paulo, Brazil. Mailing address for reprint requests and other correspondence: Geraldo Duarte, Departamento de Ginecologia e Obstetricia da Faculdade de Medicina de Ribeirão Preto (USP), Avenida Bandeirantes, No. 3900, CEP 14049-900, Ribeirão Preto, SP, Brazil.

3 Department of Puericulture and Pediatrics, Ribeirão Preto School of Medicine, University of São Paulo.

4 Department of Clinical Medicine, Ribeirão Preto School of Medicine, University of São Paulo.

5 Although confidence intervals were calculated to get some idea of the variability of the estimates, it is important to note that the 7992 women studied did not constitute a random sample of the general population of women, so that due caution must be exercised in making inferences about population-based values.

Organización Panamericana de la Salud Washington - Washington - United States
E-mail: contacto_rpsp@paho.org