Syphilis-Associated Perinatal and Infant Mortality in Rural Malawi
McDermott, J., Steketee, R., Larsen, S., Wirima, J., Bulletin of the World Health Organization
Syphilis infection in pregnancy causes high rates of fetal and early infant death and adversely affects the health of women. The prevalence of reactive syphilis serology among pregnant women in Africa ranges from 6% to 16%[1-5]. The findings of pregnancy studies conducted prior to instigating syphilis treatment programmes indicate that untreated syphilis has one of the following outcomes: approximately a third of cases result in second trimester spontaneous abortions or perinatal deaths; a third result in congenitally infected infants; and a third result in healthy uninfected infants. This high level of perinatal mortality has serious psychological and cultural implications for women in societies where fertility and childbearing are a critical part of social standing and self-image.
Although there is a simple screening test for syphilis, and penicillin continues to be effective for its treatment, the disease remains a largely ignored maternal and perinatal health problem in most sub-Saharan African countries. The implementation of an effective antenatal screening and treatment intervention requires a functioning health care system that collects blood specimens from women early in pregnancy, accurately carries out and interprets syphilis serology tests and reports them to the health facility, and promptly treats positive women and their partners. Currently, the resources required for antenatal syphilis screening and treatment programmes are either lacking or difficult to sustain. The need for such programmes or the need to sustain an existing programme by diverting resources frequently has to be justified by repeated demonstration of the extent of adverse outcomes. Also, although programmes of this type are available in large towns in many African countries, they are operationally ineffective because of centralized testing, which results in low follow-up and treatment of positive women.
In Malawi, despite national policy, routine antenatal syphilis screening programmes were discontinued in many rural district hospitals when they could not sustain the programmatic requirements. In such a setting and within the context of a population-based study of malaria prevention in pregnant women, we examined and quantified the prevalence of and risk factors for acquiring syphilis and its outcomes among a rural population of pregnant women.
Enrolment and follow-up
We carried out a prospective study of malaria chemoprophylaxis among pregnant women between 1987 and 1990 in a rural district in Malawi. Consecutive pregnant women who were attending any of four antenatal clinics were enrolled and followed up throughout pregnancy and then once every two months for more than a year after delivery. Because 93% of pregnant women in Malawi attend antenatal care (L. Schultz et al., unpublished results, 1993) and the study women and their infants were followed up in the community, the study can be considered to be population-based. The women received routine antenatal care, including monthly visits, tetanus toxoid, iron supplements, and malaria chemoprophylaxis, but there was no routine syphilis screening through the clinic system. Stored blood specimens collected during the antepartum period and at delivery as part of the malaria study were later tested serologically for syphilis.
Syphilis serology testing and case definition
The sera were tested in 1991-92 for syphilis, using Venereal Disease Research Laboratory (VDRL) or rapid plasmin reagin (RPR) tests and a microhaemagglutination assay for antibodies to Treponema pallidum (MHA-TP). All VDRL/RPR reactive sera were diluted to an endpoint titre.
The definitions of syphilis used in the study were based on the serological findings expected for the various stages of the disease. The last cases of yaws in Malawi occurred prior to 1977, and this condition is considered to have been eradicated from the country. Women with a VDRL/RPR titre [greater than or equal to]1:8 and reactive MHA-TP were considered to have recently acquired active syphilis. Larsen et al. and Gershman et al. have found that in areas where the eradication of yaws could be documented a titre of [greater than or equal to]1:16 in the RPR card test, confirmed by MHA-TP, indicated a recently acquired case of syphilis[12, 13]. Women who had both a nonreactive VDRL/RPR test and MHA-TP assay were considered to be syphilis negative; those with reactive VDRL/RPR tests and a nonreactive MHA-TP assay were considered to be syphilis false-positives (although this group may also have included some very early cases of syphilis); those with a reactive MHA-TP assay and a nonreactive VDRL/RPR test were considered to be syphilis resolved (old treated or untreated cases). Women with a VDRL/RPR titre <1:8 and a reactive MHA-TP assay included a mixture of old, treated or untreated infections and newly acquired ( <6 weeks) infections, and were considered to be syphilis old/new. Because the interpretation of inconsistent serological results is difficult (either the VDRL/RPR test or MHA-TP assay is discordant), we compared women whose serology was consistent with active syphilis (VDRL/RPR reactive with a titre [greater than or equal to]1:8 and reactive MHA-TP) with women who had no evidence of syphilis (nonreactive VDRL/RPR and MHA-TP assay) to identify and quantify the risk factors for active syphilis and to estimate the contribution of syphilis to poor pregnancy outcomes.
Analysis of results
To avoid potential confounding by human immunodeficiency virus (HIV) infection and multiple pregnancies, we included in the analysis only women who were botb HIV-negative by enzyme-linked immunosorbent assay (ELISA) within 2 months of delivery and who had singleton births.
Any infections were incident cases, since the case definition for syphilis was consistent with newly acquired infections. Odds ratios (OR) with 95% Cornfield or exact confidence intervals (CI) (as appropriate) were used in the univariate analysis to identify risk factors for syphilis and to quantify the association between the active disease and poor pregnancy outcome. Variables entered into logistic regression models were analysed in a forward stepwise fashion with an entry P-value = 0.05 and removal P-value = 0.06 using the PROC LOGISTIC procedure (SAS version 6.04). Primigravidae and multigravidae were modelled separately in identifying risk factors for active syphilis, since multigravidae had a set of potential risk factors related to their past obstetric history that were not applicable to the primigravidae (see Tables 2 and 3). For the poor pregnancy outcomes that were considered (see Tables 4 and 5), the referent group was live births who survived the post-delivery period considered in the outcome. For example, the referent group for stillbirth was live births, and that for neonatal death was live births who survived until at least the 28th day of life. The population-attributable risk percent (PAR%) was used to estimate the impact of syphilis on poor pregnancy outcomes; this measurement combines the strength of the risk factor (relative risk measure) and the prevalence of the risk factor in the population. The 95% CIs for the PAR% were calculated using the method described by Walters.
[TABULAR DATA OMITTED]
Of the 3591 women who were enrolled and tested in the study, 130 (3.6%) had serological results that were consistent with active syphilis (VDRL/RPR test reactive, with titre [greater than or equal to]1:8 and reactive MHA-TP assay), and 2968 (82.7%) had nonreactive VDRL/RPR tests and MHA-TP assays (Table 1).
[TABULAR DATA OMITTED]
Risk factors for syphilis
The maternal characteristics that were potential risk factors for active syphilis among primigravidae are shown in Table 2, while those for multigravidae are shown in Table 3. Primigravidae with active syphilis were more than twice as likely to be illiterate (OR = 2.31; 95% CI = 1.03, 5.56) than those who did not have syphilis. Multigravidae with active syphilis were twice as likely to have had a home delivery (OR = 1.9; 95% CI = 1.24, 2.91) or to have a history of child death in the previous pregnancy (OR = 2.05; 95% CI = 1.28, 3.29) than women without syphilis. Multigravidae enrolled in clinic C or D were 2-3 times more likely to have had active syphilis than women who enrolled in clinic A (the clinic with the lowest rate of active syphilis). A history of a stillbirth with the previous pregnancy was a very strong risk factor (OR = 7.84; 95% CI = 4.05, 15.18). The following factors were, however, not associated with active syphilis in either primigravidae or multigravidae: maternal age, alcohol use, parity, late antenatal care (begun <60 days before delivery), enrolment at clinic B, haematocrit <21% at enrolment, low socioeconomic status, and a history of a spontaneous abortion.
Odds ratios for syphilis and poor pregnancy
Compared with women without syphilis, those who had active syphilis were more likely to experience all types of stillbirths as well as the early and late neonatal death and even postneonatal death of their infants (Table 4).
Women with active syphilis were 11 times more likely to have a stillbirth (OR = 10.89; 95% CI = 6.61, 17.93), with the risk of a macerated stillbirth being 18 times greater and of a fresh stillbirth almost eight times greater. This increased risk associated with active syphilis extended not only into the neonatal (OR = 4.86; 95% CI = 2.73, 8.66) but also into the postneonatal period (OR = 2.24; 95% CI = 1.26, 3.99).
There was very little confounding between active syphilis and other factors, as evidenced in Table 4 by the similarity of the odds ratios for active syphilis obtained in the univariate and multivariate analyses. Smoking, alcohol consumption, anaemia, or parasitaemia at enrolment were not associated with any of the poor pregnancy outcomes.
Population-attributable risk percent for
Active syphilis accounted for 21% of the perinatal deaths and 8% of the infant deaths in the study population, as well as contributing to 26% of the stillbirths, 11% of the neonatal deaths, and 5% of the postneonatal deaths (Table 5).
In sub-Saharan Africa, syphilis is a common infection that has clear, detrimental effects on fetal and infant survival. Most studies of syphilis in Africa have examined urban or hospital-based populations and few studies have quantified the problem in rural population-based settings. In 1982-83 Lijestrand et al. used a nationwide study of maternal morbidity to estimate the prevalence of syphilis among pregnant women in villages in Mozambique, but did not estimate the contribution of syphilis to poor pregnancy outcome. Greenwood et al. carried out a clinical survey, which included syphilis serology, as part of a prospective population-based study of pregnancy outcome in 41 villages in the Gambia; although the prevalence of reactive syphilis serology was high, no association was found with perinatal mortality. This lack of association between maternal syphilis and adverse birth outcome could have arisen because the positive test outcomes were due to infections with nonvenereal treponemal infections, e.g., endemic syphilis or yaws.
Our findings quantify the significant impact of syphilis during pregnancy on fetal and infant mortality in rural Malawi. This impact was not limited to the fetal and neonatal periods but continued into the postneonatal period. In the study population nearly 1 out of 4 stillbirths, 1 out of 9 neonatal deaths, and 1 out of 25 postneonatal deaths were attributed to maternal syphilis. This strong association between reactive syphilis serology and poor outcome indicates that venereal syphilis was responsible, rather than other treponemal infections.
Syphilis during pregnancy has been recognized to be a public health problem in Africa for many years, but there have been few sustainable programmes to deal with it. Clinicians have attempted to identify women at risk by limiting screening to those in urban areas or those with a history of stillbirth, since these were assumed to be risk factors. The prevalence of active syphilis that we found in a rural area (3.6%) demonstrates that screening should not be limited only to urban areas. Although a history of a stillbirth as the outcome of the immediately preceding pregnancy was an important risk factor in the study population (OR = 7.84), this criterion identified none of the primigravidae and only 14 (15%) of the 96 multigravidae with active syphilis. Other risk factors for active syphilis (illiteracy, clinic site, home delivery, and the death of the immediately previous child) were not very useful for identifying women at high risk. Over 70% of the women were illiterate, and only 33% of those with active syphilis came from the two clinics with the highest rates of this condition. The factors that determined the birth location were unknown, and over 40% of the women delivered at home. A universal screening programme appears to be the most appropriate option.
The identification and immediate proper treatment of women with reactive serologies at the first visit to an antenatal clinic is the key element in a successful antenatal syphilis intervention programme. For this purpose, macroscopic rapid tests (RPR card or toluene red unheated serum test (TRUST)) are suitable. The results of such screening tests are available immediately and women with reactive tests can be treated at their first visit. Many of the logistic problems associated with identifying women who have reactive sera would thus be eliminated, and treatment would not be delayed until the next visit or missed altogether.
Contact-tracing and treatment of partners, while important, may be more difficult to accomplish. Although partner notification and treatment should not be ignored, initial attention should be focused on the early identification and treatment of pregnant women who are reactive in screening tests. After adequate coverage has been achieved and sustained for pregnant women, activities should then be extended to include treatment of their partners.
Several recent reports have described cost-benefit analyses of antenatal screening programmes, even in countries where the prevalence of syphilis is low (0.04-0.4%)[19-21]. In Lusaka, Zambia, Hira et al. have reported a 50% reduction in adverse pregnancy outcomes after implementation of an antenatal screening programme that included a strong health education component that promoted early attendance at antenatal clinics. The cost of the intervention (per 1000 women), including that of two tests per woman, treatment of women with reactive tests and their partners, health education materials, and laboratory equipment is estimated to range from US$ 4.20, in a country where the prevalence is 1%, to US$ 7.00 in a country where the prevalence is 15%. The cost of averting each adverse outcome (fetal loss, perinatal death, or congenitally infected infant) was estimated to be US$ 70.00 in the 1% prevalence country and US$ 9.28 in the 15% prevalence country. This compares with the estimated US$ 152.53 to avert a neonatal tetanus death and US$ 99.85 to avert a pertussis death using immunization interventions in the Gambia.
The potential for a programme to prevent congenital syphilis in the perinatal, neonatal, and postneonatal period is clear. In considering resource allocation to child survival programmes in areas where the prevalence of syphilis is high, those responsible need to include antenatal syphilis screening, using rapid tests and treatment at the first visit.
Mortalite perinatale et infantile associee a
la syphilis dans les regions rurales du
En Afrique, la prevalence notifiee de la syphilis chez les femmes qui se rendent aux consultations antenatales va de 6 a 16%, et l'on estime a 20 a 40% la proportion de femmes ayant une syphilis non traitee dont l'enfant decede au cours de la periode perinatale. Bien qu'il existe, pour cette maladie, un test de depistage sur le terrain et qu'elle puisse etre efficacement traitee par la penicilline, la syphilis reste, dans de nombreux pays en developpement, un probleme de sante maternelle et perinatale meconnu. Nous avons etudie la prevalence de la syphilis et les facteurs de risque d'infection, ainsi que les consequences de la maladie chez une population rurale de femmes enceintes au Malawi.
Chez 3591 femmes ayant eu un accouchement simple et dont la serologie pour le VIH (virus de l'immunodeficience humaine) etait negative, la prevalence d'une syphilis evolutive, definie par un titre [greater than or equal] 1:8 dans les tests VDRL (Venereal Disease Research Laboratory) ou RPR (rapid plasmin reagin) et par un resultat positif dans l'epreuve de micro-hemagglutination pour la recherche des anticorps diriges contre Treponema pallidum (MHA-TP), etait de 3,6%. Ces resultats montrent que la syphilis s'etend aux regions rurales et qu'il ne faut pas limiter le depistage aux zones urbaines.
Lors de l'analyse, les femmes atteintes de syphilis evolutive ont ete comparees a celles dont les tests VDRL/RPR et MHA-TP etaient negatifs. Les parametres associes a la syphilis evolutive ont ete analyses separement pour les primigestes et les multigestes. Bien que le fait que la grossesse immediatement precedente se soit soldee par une mortinaissance constitue un facteur de risque important dans la population etudiee (odds ratio (OR) = 7,84), ce critere n'aurait permis d'identifier aucune des primigestes atteintes de syphilis evolutive et seulement 14 (15%) des multigestes concernees. Les facteurs de risque suivants ont ete identifies pour la syphilis evolutive: analphabetisme (OR = 2,39); recrutement dans le dispensaire C (OR = 2,94) ou dans le dispensaire D (OR = 2,23); accouchement a domicile (OR = 1,9); deces neonatal lors de l'accouchement immediatement precedent (OR = 2,05). Ces facteurs ne permettaient toutefois pas de cibler efficacement les femmes a haut risque. Plus de 70% des femmes etaient illettrees, 33% seulement des femmes syphilitiques venaient des deux dispensaires ayant les taux de syphilis les plus eleves, et plus de 40% des femmes accouchaient a domicile. Ces observations plaident en faveur du depistage general de la syphilis dans les programmes antenatals.
Par rapport aux femmes non syphilitiques, les femmes atteintes de syphilis evolutive avaient une probabilite 10,89 fois plus forte d'avoir un enfant mort-ne, 4,86 fois plus forte de voir leur enfant mourir au cours de la periode neonatale et 2,24 fois plus forte de le voir mourir au cours de la periode post-neonatale. La syphilis evolutive est responsable de 21% des morts perinatales et de 8% des morts infantiles dans la population etudiee. II est manifeste qu'un programme destine a prevenir la syphilis congenitale au cours des periodes perinatale, neonatale et post-neonatale est necessaire. Le cout (pour 1000 femmes) de deux tests serologiques par femme, du traitement des femmes ayant un resultat positif et de leur partenaire, du materiel d'education pour la sante, et du materiel de laboratoire, va de US$ 4,20 dans un pays ou la prevalence de la syphilis est de 1% a US$ 7,00 dans un pays ou elle est de 15%. Le cout par issue indesirable evitee (mort foetale in utero, deces perinatal, ou nouveau-ne atteint de syphilis congenitale) a ete estime a US$ 70,00 dans un pays ou la prevalence de la syphilis est de 1% et a US$ 9,28 dans un pays ou elle est de 15%. On citera pour comparaison le cout, estime a US$ 152,53, d'un deces par tetanos neonatal evite et de US$ 99,85 par deces du a la coqueluche evite par une intervention vaccinale en Gambie. Lors de l'examen des attributions de ressources aux programmes de survie de l'enfant dans les regions ou la prevalence de la syphilis est elevee, il faut envisager d'inclure un depistage antenatal de la syphilis, avec emploi de tests rapides et traitement lors du premier contact de la mere avec le systeme de soins de sante.
[1.] Cooper-Poole, B. Prevalence of syphilis in Mbeya, Tanzania - the validity of the VDRL as a screening test. East African medical journal, 63: 646-650 (1986). [2.] Guinness, L.F. et al. Influence of antenatal screening on perinatal mortality caused by syphilis in Swaziland. Genitourinary medicine, 64: 294-297 (1988). [3.] Manning, B. et al. Syphilis in pregnant Black women. South African medical journal, 67: 966-967 (1985). [4.] Lijestrand, J. et al. Syphilis in pregnant women in Mozambique. Genitourinary medicine, 61: 355-358 (1985). [5.] Hira, S. et al. Syphilis intervention in pregnancy: Zambian demonstration project. Genitourinary medicine, 66: 159-164 (1990). [6.] Ingraham, N.R. The value of penicillin alone in the prevention and treatment of congenital syphilis. Acta dermato-venerologica, 31 (suppl. 24): 60-88 (1951). [7.] Rapid plasma reagin (RPR) 18-mm circle card test, toluidine red unheated serum test (TRUST), and microhemagglutination assay for antibodies to T. pallidum. In: Larsen, S. et al., ed. A manual of tests for syphilis. Washington, DC, American Public Health Association, 1990, pp. 100-108, 120-127, and 154-165. [8.] Guidelines for the prevention and control of congenital syphilis. Morbidity and mortality weekly review, 37(suppl. S-1), 1988. [9.] Diagnostic tests. In: Larsen, S. et al., ed. A manual of tests for syphilis. Washington, DC, American Public Health Association, 1990, p. 6. [10.] Larsen, S. Current status of laboratory tests for syphilis. In: Rippey, J. & Nakamuin, R., ed. Diagnostic immunology: technology assessment Skokie, IL, College of American Pathologists, 1983, pp. 162-170. [11.] Widy-Wirski, R. Surveillance and control of resurgent yaws in the African region. Reviews of infectious diseases, 7(suppl. 2): S227-S232 (1985). [12.] Larsen, S. A population-based serosurveillance of syphilis in Costa Rica. Sexually transmitted diseases, 18: 124-128 (1991). [13.] Gershman, K. Seroepidemiological characteristics of a syphilis epidemic in the Republic of the Marshall Islands, formerly a yaws endemic area. International journal of epidemiology, 21: 599-606 (1992). [14.] Kleinbaum, D. et al. Epidemiologic research. Belmont, CA, Lifetime Learning, 1982, pp. 304-305. [15.] Mehta, C.R. et al. Computing an exact confidence interval for common odds ratio in several 2 x 2 contingency tables. Journal of the American Statistical Association, 80: 969-973 (1985). [16.] Lilienfeld, A. & Lilienfeld, D. Observational studies: retrospective and cross-sectional studies. In: Foundations of epidemiology, 2nd ed. New York, Oxford University Press, 1980, p. 217. [17.] Walters, S. Calculation of attributable risks from epidemiological data. International journal of epidemiology, 7: 175-182 (1978). [18.] Greenwood, A. et al. Treponemal infection and the outcome of pregnancy in a rural area of The Gambia, West Africa. Journal of infectious diseases, 166: 842-846 (1992). [19.] Stray-Pedersen, B. Economic evaluation of maternal screening to prevent congenital syphilis. Sexually transmitted diseases, 10: 167-172 (1983). [20.] Garland, S. & Kelly, N. is antenatal screening for syphilis worthwhile? Medical journal of Australia, 151: 368-372 (1989). [21.] Bowell, P. et al. Serological screening tests for syphilis in pregnancy: results of a five-year study (1983-1987) in the Oxford region. Journal of clinical pathology, 42: 1281-1284 (1989). [22.] Schulz, K. et al. Maternal health and child survival: opportunities to protect both women and children from the adverse consequences of reproductive tract infections. In: German, A. et al., ed. Reproductive tract infections: global impact and priorities for women's reproductive health. New York, Plenum Press, 1992, pp. 145-154. [23.] Robertson, R. et al. Cost-effectiveness of immunization in the Gambia. Journal of tropical medicine and hygiene, 88: 343-351 (1985).…
Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: Syphilis-Associated Perinatal and Infant Mortality in Rural Malawi. Contributors: McDermott, J. - Author, Steketee, R. - Author, Larsen, S. - Author, Wirima, J. - Author. Journal title: Bulletin of the World Health Organization. Volume: 71. Issue: 6 Publication date: November-December 1993. Page number: 773+. © 1990 World Health Organization. COPYRIGHT 1993 Gale Group.
This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.