hiv/aids, tuberculosis, and malaria in...

45
HIV/AIDS, Tuberculosis, and Malaria in Pregnancy Journal of Pregnancy Guest Editors: Oliver Ezechi, Karen Odberg-Petterson, and Josaphat Byamugisha

Upload: vodien

Post on 20-Jan-2019

220 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

HIV/AIDS, Tuberculosis, and Malaria in Pregnancy

Journal of Pregnancy

Guest Editors: Oliver Ezechi, Karen Odberg-Petterson, and Josaphat Byamugisha

Page 2: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

HIV/AIDS, Tuberculosis, and Malaria inPregnancy

Page 3: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Journal of Pregnancy

HIV/AIDS, Tuberculosis, and Malaria inPregnancy

Guest Editors: Oliver Ezechi, Karen Odberg-Petterson,and Josaphat Byamugisha

Page 4: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Copyright © 2012 Hindawi Publishing Corporation. All rights reserved.

This is a special issue published in “Journal of Pregnancy.” All articles are open access articles distributed under the Creative CommonsAttribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is prop-erly cited.

Page 5: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Editorial Board

Vincenzo Berghella, USASean Blackwell, USAHein Bruinse, The NetherlandsRosa Corcoy, SpainR. L. Deter, USAGraziano Di Cianni, ItalyGian Carlo Di Renzo, ItalyKeith A. Eddleman, USAFabio Facchinetti, ItalyAntonio Farina, ItalyAlbert Fortuny, Spain

Sinuhe Hahn, SwitzerlandMordechai Hallak, IsraelJeffrey Keelan, AustraliaJustin Konje, UKDavid F. Lewis, USAE. R. Lumbers, AustraliaSam Mesiano, USARichard K. Miller, USARoberta B. Ness, USACees Oudejans, The NetherlandsAttila Pal, Hungary

Vorapong Phupong, ThailandFulgencio Proverbio, VenezuelaM. Rogers, Hong KongDaniel S. Seidman, IsraelLee P. Shulman, USAJ. L. Simpson, USAMarlene Sinclair, UKMark S. Sklansky, USADeborah A. Wing, USATamas Zakar, Australia

Page 6: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Contents

HIV/AIDS, Tuberculosis, and Malaria in Pregnancy, Oliver Ezechi, Karen Odberg-Petterson,and Josaphat ByamugishaVolume 2012, Article ID 946815, 3 pages

Pattern and Determinants of Antiretroviral Drug Adherence among Nigerian Pregnant Women,S. O. Ekama, E. C. Herbertson, E. J. Addeh, C. V. Gab-Okafor, D. I. Onwujekwe, F. Tayo, and O. C. EzechiVolume 2012, Article ID 851810, 6 pages

Tuberculosis in Pregnancy: A Review, Olabisi M. Loto and Ibraheem AwowoleVolume 2012, Article ID 379271, 7 pages

The Impact of HIV on Maternal Morbidity in the Pre-HAART Era in Uganda,Harriet Nuwagaba-Biribonwoha, Richard T. Mayon-White, Pius Okong, Peter Brocklehurst,and Lucy M. CarpenterVolume 2012, Article ID 508657, 6 pages

A Model of Tuberculosis Screening for Pregnant Women in Resource-Limited Settings Using XpertMTB/RIF, Eleanor R. Turnbull, Nzali G. Kancheya, Jennifer B. Harris, Stephanie M. Topp,German Henostroza, and Stewart E. ReidVolume 2012, Article ID 565049, 5 pages

Understanding Methods for Estimating HIV-Associated Maternal Mortality, James E. Rosen,Isabelle de Zoysa, Karl Dehne, Viviana Mangiaterra, and Quarraisha Abdool-KarimVolume 2012, Article ID 958262, 8 pages

Risk Factors Associated with False Positive HIV Test Results in a Low-Risk Urban Obstetric Population,Tamara T. Chao, Jeanne S. Sheffield, George D. Wendel Jr., M. Qasim Ansari, Donald D. McIntire,and Scott W. RobertsVolume 2012, Article ID 841979, 4 pages

Page 7: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Hindawi Publishing CorporationJournal of PregnancyVolume 2012, Article ID 140826, 3 pagesdoi:10.1155/2012/140826

Editorial

HIV/AIDS, Tuberculosis, and Malaria in Pregnancy

Oliver Ezechi,1 Karen Odberg Petterson,2 and Josaphat Byamugisha3

1 Maternal and Reproductive Health Research Unit, Division of Clinical Sciences, Nigerian Institute of Medical Research,101212 Yaba Lagos, Nigeria

2 Division of Global and International Health, Faculty of Medicine, Lund University, 20502 Malmo, Sweden3 Department of Obstetrics and Gynecology, School of Medicine, Makerere University, College of Health Sciences,P.O. Box 7072, Kampala, Uganda

Correspondence should be addressed to Oliver Ezechi, [email protected]

Received 2 February 2012; Accepted 2 February 2012

Copyright © 2012 Oliver Ezechi et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

HIV/AIDS, tuberculosis (TB), and malaria (ATM) are 3major global public health threats that undermine develop-ment in low- and middle-income countries [1].

Approximately 5 million lives are lost annually as a resultof these infections, with substantial humanitarian, economic,and social impact, which is still not fully measured [1]. Thethree infections are not only associated with poverty but alsooccur in the same geographic zone and have major publichealth implications [2]. The consequences of interactionsbetween the diseases are particularly serious for reproductivehealth. They often intersect in pregnancy, resulting in poormaternal and fetal outcomes. As access to their treatmentis increasing in low-income countries and new and cheaperalternative drugs are being deployed, there is also the po-tential for interactions between their treatments, whichaffects the efficacy of each drug.

The co-infected pregnant women are at a very high risk ofanaemia and infection of the placenta; hence a considerableproportion of children born to women with HIV, TB andmalaria infection have low birth weight and are more likelyto die during infancy [3–5]. Also, there is the risk of motherto child transmission of the infections.

Globally, an estimated 342,900 maternal deaths occurannually and more than 99% of these occur in low- andmiddle-income countries [6]. It is barely four years to thetarget for the attainment of Millennium Development Goal(MDG) 5, which seeks to reduce the maternal mortalityratio (MMR) by three-fourths. The three leading causesof these deaths have remained hemorrhage, infections, andhypertensive disorders [6, 7]. Local and regional variations

do exist, however in sub-Saharan Africa, and Asia, infectionas a result of HIV epidemic, resurgence of TB, and theubiquitous malaria is a major cause [7].

Malaria is the most common of the three infections andit is estimated that over 50 million women are exposed tothe risk of malaria in pregnancy annually [8]. Malaria infec-tion in pregnancy results in substantial maternal, fetal, andinfant morbidity and mortality, accounting for 75 000–200 000 infant deaths yearly [9]. It causes diverse adversepregnancy outcomes, including maternal anaemia and lowbirth weight due to preterm delivery and fetal growthrestriction. Pregnant women are more susceptible than non-pregnant women to malaria, and this susceptibility is greatestin the first and second pregnancy. Susceptibility to malariaduring pregnancy probably represents a combination of im-munological and hormonal changes associated with preg-nancy, combined with the unique ability of a subset ofinfected erythrocytes to sequester in the placenta [9].

In the same settings of malaria, HIV infection has con-tinued to remain a major public health challenge and thushave been described as a deadly combination, especially inpregnancy [5]. Available evidence confirms that HIV-positivewomen are at greater risk of maternal death and morbidity[10]. In high burden areas, which also have the highestrates of maternal mortality, HIV is beginning to reverse theprogress in safe motherhood [10]. HIV-infected women con-front stigma, discrimination, and violence that compromisetheir health and may put their lives and their offspring atrisk. In some situations they have been refused services forlabour and delivery and/or have been given substandard

Page 8: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

2 Journal of Pregnancy

services [10]. The estimates of the contribution of HIV andTB infections to overall maternal mortality vary greatly byregion [11]. All are certainly underestimates because the HIVstatus of many women is not known, and HIV contributesto increased mortality related to pregnancy even beyond 42days postpartum, the period during which maternal healthis calculated [5]. Areas with high HIV prevalence and highmaternal mortality also have lower rates of antiretroviraltreatment and higher levels of diseases and conditions thatcan be exacerbated by both HIV infection and pregnancy[12].

There are several ways in which HIV is believed to con-tribute to elevated maternal mortality and morbidity. HIVinfection is associated with an increase in direct complica-tions such as anaemia, postpartum haemorrhage, and puer-peral sepsis [11, 13]. In addition, both HIV infection andpregnancy may increase susceptibility to other diseases thataffect pregnancy outcome such as tuberculosis, malaria, andpneumonia [9, 12, 14]. Pregnancy itself renders women morevulnerable to acquiring HIV infection. Finally, women whoare HIV positive and pregnant may receive lower quality carefor both pregnancy and HIV due to discrimination by healthcare providers [10].

Globally, tuberculosis has increased over the past twodecades, with the largest burden of disease borne by impov-erished communities in the same geographic areas wheremalaria and HIV are endemic [13]. While it is known thatthe deterioration in socioeconomic conditions, poorly func-tioning national tuberculosis-control programmes, and mul-tidrug resistance have to varying extents contributed tothe resurgence of tuberculosis, the main driving force hasbeen the pandemic of HIV-1 infection [13]. A considerableoverlap exists in the age groups commonly affected bytuberculosis and HIV [15, 16].

The current understanding of the human immune re-sponse to malaria, HIV, and TB leads us to expect the threeinfections to influence the clinical outcome of each other aswell as that of pregnancy. Each of the infections is expectedto accelerate the progression of the other and thus impactsheavily on pregnancy, not only in terms of the morbidityand mortality but also on the transmission of the infectionsto the unborn child [3–5, 8]. The effects of HIV infectionon maternal health are superimposed on those of malariain low-income countries, where over 1 million pregnanciesannually are complicated by malaria-HIV coinfection [3,4]. With the documented 20–30% rate of HIV and TBcoinfection in these same settings, it is safe to state that atany given time, more than 300,000 pregnancies are triplyinfected by HIV, malaria, and TB. Such pregnancies areassociated with higher malaria parasite densities, HIV viralload, and a higher risk of maternal anaemia, low birth-weight, and perinatal and infants morbidity and mortality[3–5, 8]. While HIV and TB infection affects all pregnancies,the HIV-associated risk of malaria is consistently greater inmultigravida [3–5, 8]. Furthermore there are risks of poorerresponse to malaria prophylaxis and treatment with HIV andalso adverse drug reaction during pregnancy [9].

At the clinical level, the interactions between the infec-tions are complex and bidirectional [9]. Although the

morbidity and mortality attributable to each infection in thepresence of the other are not fully understood, interactionsthat are likely to have significant clinical consequences havebeen described [9, 14]. Each infection exacerbates the other:Malarial and TB infections of the placenta increase placentalviral load, thus increasing the risk of vertical transmissionof HIV. On the other hand, HIV-infected pregnant womenhave reduced antimalarial and TB immunity and higherprevalence rates of malaria and TB [15–17, 16]. TB-infectedpregnant women are more susceptible to HIV and malariainfection compared to their noninfected counterparts [3–5].

Though the burden and the consequences of the inter-action and intersection of the three infections in pregnancyare frightening, the real tragedy is the failure to make lowcost and effective interventions tools available to the women.These tools and interventions include prevention of motherto child transmission of HIV (PMTCT), Intermittent Pro-phylactic Therapy (IPTp), Long-Lasting Insecticide TreatedNets (LLINs), and infection control practices (ICPs). Theprize of the slow or nondeployment of evidence-based andlow cost tools is the avoidable human tragedy called maternaland infant mortality.

The regions with high HIV, malaria, and TB burdenfurther suffer from the world’s most pronounced crisis inhuman resources for health [18]. This shortage is com-pounded by the discussed infections especially the HIV/AIDSpandemic. It is therefore only reasonable to consider humanresource issues while trying to prevent and control the infec-tions. One strategy that has been suggested and evaluated isthe reassignment of clinical roles by shifting tasks to differentcadres of health workers. For examples nurses and midwivesmay become involved in prescribing drugs, lay counselorsinvolved in testing, and patients may be engaged in takingover some elements of their own care [18].

Halting and reversing the human tragedy due to thedeadly trio of HIV, malaria, and TB requires not only thedeployment of the evidence-based prevention and man-agement tools already available but also the integration ofvarious services into the existing maternal and child healthcare services utilizing available human resource. However,the best model for deployment of such tools in low-in-come countries has not yet been identified. Operationalresearch is therefore required to determine the best models ofintegration and deployment of these services through testingof existing models

Effort at addressing the challenge of HIV, TB, and malariain pregnancy should address all sexual and reproductivehealth services as one entity rather than in units. Many ofthe interventions that would improve their outcomes such asfamily planning, antenatal and delivery care, including accessto emergency obstetric services are general issues applying toall women. The integration of TB, malaria, and HIV servicesinto existing maternal, newborn, and child health care pro-grammes and task shifting seems to be the ideal strategy toreverse the human toll due to the three infections.

Oliver EzechiKaren Odberg Pettersson

Josaphat Byamugisha

Page 9: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Journal of Pregnancy 3

References

[1] M. Vitoria, R. Granich, C. F. Gilks et al., “The global fightagainst HIV/AIDS, tuberculosis, and malaria: current statusand future perspectives,” American Journal of Clinical Pathol-ogy, vol. 131, no. 6, pp. 844–848, 2009.

[2] World Health Organization, “Global malaria control andelimination,” Report of a Technical Review, World HealthOrganization, Geneva, Switzerland, 2008.

[3] World Health Organization, “Malaria and HIV/AIDS inter-actions and implications,” Conclusions of a Technical Con-sultation Convened by WHO, World Health Organization,Geneva, Switzerland, 2004, http://www.who.int/malaria/pub-lications/atoz/who hiv 2004 08/en/index.html.

[4] C. Ticconi, M. Mapfumo, M. Dorrucci et al., “Effect of mater-nal HIV and malaria infection on pregnancyand perinataloutcome in Zimbnabwe,” Journal of Acquired Immune Defi-ciency Syndromes, vol. 34, no. 3, pp. 289–294, 2003.

[5] L. M. Mofenson and B. E. Laughon, “Human immunode-ficiency virus, Mycobacterium tuberculosis, and pregnancy: adeadly combination,” Clinical Infectious Diseases, vol. 45, no.2, pp. 250–253, 2007.

[6] M. C. Hogan, K. J. Foreman, M. Naghavi et al., “Maternal mor-tality for 181 countries, 1980–2008: a systematic analysis ofprogress towards millennium development goal 5,” TheLancet, vol. 375, no. 9726, pp. 1609–1623, 2010.

[7] C. Ronsmans and W. J. Graham, “Maternal mortality: who,when, where, and why,” The Lancet, vol. 368, no. 9542, pp.1189–1200, 2006.

[8] S. J. Rogerson, L. Hviid, P. E. Duffy, R. F. Leke, and D. W.Taylor, “Malaria in pregnancy: pathogenesis and immunity,”The Lancet Infectious Diseases, vol. 7, no. 2, pp. 105–117, 2007.

[9] C. J. Uneke and A. Ogbonna, “Malaria and HIV co-infectionin pregnancy in sub-Saharan Africa: impact of treatment usingantimalarial and antiretroviral agents,” Transactions of theRoyal Society of Tropical Medicine and Hygiene, vol. 103, no.8, pp. 761–767, 2009.

[10] J. McIntyre, “Maternal health and HIV,” Reproductive HealthMatters, vol. 13, no. 25, pp. 129–135, 2005.

[11] J. McIntyre, “Mothers infected with HIV,” British MedicalBulletin, vol. 67, pp. 127–135, 2003.

[12] O. C. Ezechi, D. A. Oladele, C. V. Gab-Okafor et al., “Child-bearing desires and intentions of Nigerians living with HIVinfection,” in Proceedings of the 45th annual conference of theSociety of Obstetrics and Gynaecology of Nigeria, pp. 21–22,Ibadan, Nigeria, November 2011.

[13] E. L. Corbett, C. J. Watt, N. Walker et al., “The growing burdenof tuberculosis: global trends and interactions with the HIVepidemic,” Archives of Internal Medicine, vol. 163, no. 9, pp.1009–1021, 2003.

[14] V. Mwapasa, S. J. Rogerson, M. E. Molyneux et al., “The effectof Plasmodium falciparummalaria on peripheral and placentalHIV-1 RNA concentrationin pregnant Malawian women,”AIDS, vol. 18, no. 7, pp. 1051–1059, 2004.

[15] A. M. Mount, V. Mwapasa, S. R. Elliott et al., “Impairmentof humoral immunity to Plasmodiumfalciparum malaria inpregnancy by HIV infection,” The Lancet, vol. 363, no. 9424,pp. 1860–1867, 2004.

[16] F. H. Verhoeff, B. J. Brabin, C. A. Hart, L. Chimsuku, P.Kazembe, and R. L. Broadhead, “Increased prevalence of ma-laria in HIV-infected pregnantwomen and its implications formalaria control,” Tropical Medicine & International Health,vol. 4, no. 4, pp. 5–12, 1999.

[17] S. Khoo, D. Back, and P. Winstanley, “The potential for inter-actions between antimalarial and antiretroviral drugs,” AIDS,vol. 19, no. 10, pp. 995–1005, 2005.

[18] M. Callaghan, N. Ford, and H. Schneider, “A systematic reviewof task- shifting for HIV treatment and care in Africa,” HumanResources for Health, vol. 8, p. 8, 2010.

Page 10: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Hindawi Publishing CorporationJournal of PregnancyVolume 2012, Article ID 851810, 6 pagesdoi:10.1155/2012/851810

Research Article

Pattern and Determinants of Antiretroviral Drug Adherenceamong Nigerian Pregnant Women

S. O. Ekama,1 E. C. Herbertson,1 E. J. Addeh,1 C. V. Gab-Okafor,1

D. I. Onwujekwe,1 F. Tayo,2 and O. C. Ezechi1

1 Clinical Sciences Division, Nigerian Institute of Medical Research, P.M.B 2013, Yaba, Lagos, Nigeria2 Department of Clinical and Biopharmacy, Faculty of Pharmacy, University of Lagos, P.M.B 12003, Surulere, Lagos, Nigeria

Correspondence should be addressed to S. O. Ekama, [email protected]

Received 4 August 2011; Revised 24 November 2011; Accepted 9 December 2011

Academic Editor: Karen Odberg-Petterson

Copyright © 2012 S. O. Ekama et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Background. The need for a high level of adherence to antiretroviral drugs has remained a major hurdle to achieving maximalbenefit from its use in pregnancy. This study was designed to determine the level of adherence and identify factors that influenceadherence during pregnancy. Method. This is a cross-sectional study utilizing a semistructured questionnaire. Bivariate andmultiple logistic regression models were used to determine factors independently associated with good drug adherence duringpregnancy. Result. 137 (80.6%) of the interviewed 170 women achieved adherence level of ≥95% using 3 day recall. The desireto protect the unborn child was the greatest motivation (51.8%) for good adherence. Fear of being identified as HIV positive(63.6%) was the most common reason for nonadherence. Marital status, disclosure of HIV status, good knowledge of ART, andhaving a treatment supporter were found to be significantly associated with good adherence at bivariate analysis. However, aftercontrolling for confounders, only HIV status disclosure and having a treatment partner retained their association with goodadherence. Conclusion. Disclosure of HIV status and having treatment support are associated with good adherence. Maternaldesire to protect the child was the greatest motivator for adherence.

1. Introduction

The introduction of highly active antiretroviral therapy(HAART) has not only improved longevity in HIV-infectedindividuals but in addition has had a significant impact onthe rate of mother-to-child transmission of HIV infection(MTCT) [1]. Mother-to-child transmission is one of themodes of HIV transmission. Vaginal delivery contributes 60–70% of MTCT, breastfeeding contributes 20–30%, while in-utero infection could occur in less than 10% of MTCT [2].Without preventive intervention, about 25–40% of infantsborn to HIV-positive mothers will contract the virus [2].Following introduction of HAART, the rates of mother-to-child transmission of HIV infection has practically crashed toless than 2% [3, 4]. However, the success of HAART, like anymedication, is dependent on both the intrinsic properties ofthe drugs and the individual’s ability to take the medicationas prescribed [5]. This is particularly true in the preventionof mother-to-child transmission, where the consequence offailing to achieve viral suppression is the transmission of the

virus to the baby [5]. Adequate adherence to the prescribedantiretroviral medications is essential to achieving maximalviral suppression necessary to prevent MTCT [4].

Adherence rates exceeding 95% are necessary, in order tomaximize the benefits of antiretroviral therapy. Higher levelsof drug adherence are associated with improved virological,immunological, and clinical outcome [6].

Poor adherence to antiretroviral drugs during pregnancycan lead to suboptimal viral suppression, development ofviral resistance, higher risk of mother-to-child transmission,and mother-to-child transmission of resistant HIV strains[5].

Interrupting medication permits the virus to resume rap-id replication and as many as 1010 viral particles will be pro-duced per day [7].

This allows resistant mutant strains to be generatedwhich are no longer responsive to available antiretroviraldrugs, posing a public health danger [7].

Adherence to antiretroviral drugs poses unique challen-ges to HIV infected persons particularly in pregnant women

Page 11: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

2 Journal of Pregnancy

[7]. Improving adherence among pregnant women thereforerequires knowledge of the factors that influence adherence.

Several methods have been used to measure adherence,but no gold standard has been established [8]. Each of thesemethods has its respective strengths and weaknesses. Avail-able methods include pill counts, self-report, prescriptionrefills, medication event monitoring system (MEMS), bio-logical markers, and assays [4, 9].

Though several studies in Nigeria have evaluated thefactors associated with nonadherence to antiretroviral ther-apy among HIV-positive adults [10–12], only one fromliterature search studied antiretroviral adherence issues inHIV-positive pregnant women [4].

They deduced from their study that the determinants ofnonadherence to antiretroviral drugs in HIV-positive preg-nant women were low level of education, nondisclosure ofHIV status, and longer duration of therapy [4].

Good adherence is imperative for the success of PMTCT.This study, therefore, seeks to find out if there is good adher-ence to antiretroviral drugs among HIV-positive pregnantwomen.

The main objective of this study is to determine the leveland factors that influence adherence to antiretroviral drugsamong HIV-positive pregnant women accessing PMTCTservices in Lagos Nigeria.

Research Questions

(1) What is the level of adherence to antiretroviral drugsamong pregnant women?

(2) What are the barriers to adherence in pregnancy?

(3) What factors facilitate adherence among pregnantwomen?

2. Method

2.1. Study Setting. The study was conducted at the HIV treat-ment Centre of the Nigerian Institute of Medical Research(NIMR), Lagos, a Federal Government of Nigeria Compre-hensive HIV Care and Treatment Centre. NIMR is supportedby AIDS Prevention Initiative in Nigeria (APIN) throughthe United States president’s emergency fund for AIDS relief(PEPFAR), funding since 2004. The Centre provides anoutpatient Prevention of Mother-To-Child-Transmission ofHIV Infection (PMTCT) services, in addition to adult andpediatric HIV services. Pregnant women who tested positiveto HIV are referred from the NIMR HIV counseling andtesting centre or other public and private health institutionsto the PMTCT clinic which takes place once a week, onWednesdays. Intrapartum care is provided for these womenin collaboration with some hospitals within Lagos metropoliswhich include Lagos University Teaching Hospital Idi Araba,Lagos State University Teaching Hospital Ikeja, GeneralHospital Surulere, General Hospital Apapa, General HospitalIkorodu, Havana Specialist Hospital Surulere, and RaoSpecialist Hospital Surulere. Antenatal, postnatal, and infantpost exposure prophylaxes are provided by NIMR. Healthworkers from the collaborating centres have been trained onintrapartum care of HIV-positive mothers both by NIMR

and the national HIV programme. Each woman on thePMTCT programme is referred to any of the collaboratingcentres, close to their place of residence at 36 weeks or assoon as possible for late booking. The referral note usuallycontains detailed information about their chosen mode ofdelivery, infant feeding choice, client’s Viral Load, and CD4count. Infant postexposure prophylaxis drugs and mother’santiretroviral drugs are given to the women on referral. Thewomen are sent back to the centre at 2 weeks postdeliverywith a completed Case Record Form (CRF) designed specifi-cally to capture all delivery-related information. Informationon the CRF is used to complete the postnatal data base. TheHome-Based Care Team contacts any of the women who didnot report back 2 weeks after the expected date of delivery.This is to ascertain the reasons for the default.

2.2. Study Design. This is a cross-sectional study that wascarried out between 1st September 2009 and 31st November2009. A semistructured questionnaire was administered tothe women who met the inclusion criteria and gave consentto participate in the study. Knowledge of ART was measuredby asking some basic questions, which were graded. A scoreabove 70% was classified good while below 70% was notgood. Adherence was measured by expressing the numberof doses taken as a percentage of the number of dosesprescribed. For example if 20 doses are prescribed and 19doses are taken, adherence is 95% [13].

2.3. Study Population. All pregnant women with knowngestational age either by date or early ultrasound seen duringthe study period and who gave informed consent wereenrolled. At the time of the conduct of the study, a total of 273pregnant HIV-positive women were registered for PMTCTservices. The study aimed for a precision of ±5% for aproportion of 50% using 95% confidence interval, therefore,a sample size of 70 subjects was required. In order to increasethe power of the study, we decided to enroll all pregnant HIV-positive women who attended PMTCT clinic between 1stSeptember and 31st November 2009 provided the minimumsample size of 70 is met.

2.4. Data Management. The enrolled women were inter-viewed using a semistructured questionnaire containingclosed and open-ended questions. The open-ended questionswere included to give the women opportunity to freelyexpress themselves without boxing them into closed answers.Information on demographics, socioeconomic characteris-tics, knowledge of HIV, and antiretroviral drug medication,adherence pattern, reasons for missing drugs, and factorsthat encourage adherence were also contained in the ques-tionnaire. Data management was with SPSS for windowsversion 17. The P value was based on 95% confidenceintervals (CI); a P value > 0.05 was not significant (NS).

Descriptive analyses were first performed followed bybivariate analyses of the determinant factors associated withgood adherence. The variables that were found significantat this level were added to a multivariate logistic regressionmodel and those with a P-value < 0.05 were considered

Page 12: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Journal of Pregnancy 3

significant in the final multivariate model, calculating oddsratios (OR) and 95% confidence interval (CI).

2.5. Ethical Issues. Ethical approval for the study was ob-tained from the Institutional Review Board, Nigerian Insti-tute of Medical Research, Lagos, Nigeria. Written informedconsent was obtained from all women for the use of theirdata for study however, women who declined consent toparticipate in study were provided care but excluded fromresearch.

The clinic patients are organized into an independentsupport group of people living with HIV (Positive LifeOrganization of Nigeria) that ensures that patients are notstigmatized and discriminated against. This group ensuresthat no patient is denied requisite care because of failure toparticipate in any of our studies including this study.

3. Results

During the three months study period, one hundred and sev-enty eligible HIV-positive pregnant women consented andwere interviewed. One hundred and thirty seven (80.6%) ofthe interviewed women reported achieving adherence levelof greater or equal to 95% using 3 day recall method, with anonadherent rate of 19.4%.

The sociodemographic characteristics of the respondentsare shown in Table 1. Majority of the women were aged30–34 years (52.4%), married (77.1%), have at least oneliving child (70.0%), employed/working (76.5%), had at leastsecondary education (81.1%), and from the three majorethnic tribes of Yoruba, Igbo, and Hausa (84.1%).

Pregnancy and HIV-related characteristics are summa-rized in Table 2. One hundred and forty seven (86.5%)women had disclosed their HIV status and their disclosurewas to their partners in most of the cases (97.3%). Themajority of the women were on first line HAART regimen(58.2%), while the rest of the women were either on non-HAART prophylactic regimen (24.7%) or 2nd line HAARTregimen (17.1%). Only 20.0% and 22.9% of the womenhave had previous PMTCT experience and were within thefirst trimester, respectively. Their knowledge of HIV andantiretroviral therapy were very good as over 85% of therespondent had very good knowledge. The use of treatmentsupport was relatively common as greater than half of thewomen had a treatment supporter (55.9%) who is thehusband in 89.5% of the cases.

The outcome of subanalysis of reasons given for adheringto the antiretroviral drugs as prescribed, among the womenthat had adherence level greater 95%, is shown in Table 3.The desire to ensure that the unborn child is protected fromHIV infection was the greatest motivation (51.8%). Some ofthe written expression of the women, expressing the desire toprotect the child as the motivation for adherence includes “Iwant to protect my baby,” “I do not want to live with the guilt ofinfecting my baby,” and “I learnt my baby will not be infectedif I take my drugs religiously.” The desire to remain healthyand alive was the other motivator for adhering to the ARVdrugs (21.2%). These are some of the respondents’ writtenexpression “I want to live long train my children and fulfill my

Table 1: The sociodemographic characteristics of the women en-rolled in the study.

CharacteristicsNumber of women (%)

N = 170

Age (years)

(i) <20 2 (1.2)

(ii) 20–24 20 (11.8)

(iii) 25–29 43 (25.3)

(iv) 30–34 89 (52.4)

(v) 35–39 13 (7.6)

(vi) ≥40 3 (1.8)

Number of living children

(i) 0 51 (30.0)

(ii) 1–2 67 (39.4)

(iii) >2 52 (30.6)

Marital status

(i) Married 131 (77.1)

(ii) Not married 39 (22.9)

Occupation

(i) Housewife 35 (20.6)

(ii) Unemployed 5 (2.9)

(iii) Employed/Working 130 (76.5)

Educational level completed

(i) No formal education 3 (1.8)

(ii) primary 29 (17.1)

(iii) Secondary 99 (58.2)

(iv) Tertiary 39 (22.9)

Ethnic group

(i) Yoruba 78 (45.9)

(ii) Igbo 52 (30.6)

(iii) Hausa 13 (7.6)

(iv) Other tribes 27 (15.9)

life dreams,” “To keep me healthy, as I was always on admissionin the hospital before I started treatment. I am now okay sinceI started the ARV drugs.”

The reasons given by the nonadherent respondents formissing or skipping their drugs are shown in Table 4.Forgetfulness (57.6%), tight work schedule (39.4%), andfear of being identified as HIV positive (63.6%) were thecommon reasons for skipping or missing drugs among thenonadherent women. The respondents expressed this inwritten form, for example, “I get stuck in traffic sometimes onmy way home from work and am not comfortable carrying mydrugs around or taking them in the public.”

Table 5 shows the bivariate analysis of some possiblefactors associated with good drug adherence. Of the ninevariables, only four variables which are marital status (P =0.023), disclosure of HIV status (P = 0.000), good knowl-edge of HIV, and ART (P = 0.001 and having a treatmentsupporter (0.002) were found to be significantly associatedwith good adherence. However, after subjecting the variablesfound to be significantly associated with good adherenceto multiple logistic regressions while controlling for other

Page 13: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

4 Journal of Pregnancy

Table 2: The distribution of pregnancy and HIV related character-istics of the women enrolled in the study.

CharacteristicsNumber of women (%)

N = 170

HIV status disclosure

(i) Disclosed 147 (86.5)

(ii) Not disclosed 23 (13.5)

ARV drug regimen

(i) Mono and Dual therapy 42 (24.7)

(ii) First line HAART 99 (58.2)

(iii) 2nd line HAART 29 (17.1)

Gestational age (weeks)

(i) Less than 13 39 (22.9)

(ii) 13–28 52 (30.6)

(iii) Greater or equal to 29 79 (46.5)

Previous PMTCT experience

(i) Yes 34 (20.0)

(ii) No 136 (80.0)

Knowledge of HIV and ART

(i) Good 146 (85.9)

(ii) Poor 24 (14.1)

Has treatment supporter

Yes 95 (55.9)

No 75 (44.1)

Table 3: Reason for adherence to ARV drugs among respondentswith over 95% drug adherence.

Reason for Adherence No of respondents (%)∗

To protect my unborn child 71 (51.8)

To stay healthy and alive 29 (21.2)

As I was told by counselors 27 (19.7)

Informed by my previousPMTCT experience

16 (11.7)

∗No of respondents greater than 137 because of multiple responses.

Table 4: Reasons for missing drugs among the 33 respondents thathad less than 95% adherence.

Reason for nonadherence No of respondents (%)∗

Forgetfulness 19 (57.6)

Slept off 11 (33.3)

Work schedule 13 (39.4)

Religious activity 9 (27.3)

Food requirement of the drug 7 (21.2)

Afraid of someone identifyingmy drug as HIV drugs

21 (63.6)

∗No of respondents greater than 33 because of multiple responses.

potential confounders, only HIV status disclosure (oddsratios: 6.1; CI: 2.8–11.6) and having a treatment partner(odds ratios: 2.5; CI: 1.3–6.7) retained their association withgood adherence.

4. Discussion

Effective strategies to reduce mother-to-child transmissionof HIV infection are well known and well established [1, 3].These include the use of ARV drugs, avoiding unplanned andunwanted pregnancy in HIV-positive women, safe deliveryand infant feeding options, reduction of unwarranted andunnecessary surgical intervention during pregnancy andlabor, prevention of prolonged rupture of membrane, andso forth [1, 3]. Research has also shown that at undetectableviral load, it is possible to achieve zero mother to childtransmission of HIV infection [1]. Antiretroviral drugs canonly achieve the required effect at adherence level of at least95% [14]. Poor adherence to antiretroviral drugs has beenreported to be the major challenge to achieving the goal ofantiretroviral therapy [15]. Many factors have been cited asreasons for nonadherence in HIV-positive adults [10–12, 15–18], but relatively few adherence studies have been done inHIV positive pregnant women [4, 19]. Apart from factorspreventing adherence in nonpregnant adults, the nausea andvomiting of pregnancy and possible effect of the drugs onthe fetus are additional factors that makes adherence inpregnancy a challenge.

We conducted this study not only to provide this scarceinformation on adherence in pregnancy but also to generateinformation that will assist during adherence counselingfor pregnant HIV positive women accessing PMTCT ser-vices.

The nonadherence rate of 19.4% among our cohort,though comparable to 21.7% reported by Igwegbe et al.in South Eastern Nigeria is much lower than 37.1% and37.4% reported by Olowookere et al. [10] and Shaahu et al.[11], respectively, from Southwestern Nigeria where presentstudy was conducted. Considering that our study and thatof Igwegbe et al. [4] was among pregnant women unlikethe two studies by Olowookere et al. [10] and Shaahu etal. [11], it seems that adherence is better in pregnancy. Itis not surprising that women are willing and ready to doanything, to ensure the wellbeing of their offspring. Theabove statement was confirmed in this study in which 51.8%of the adherent women gave the reason of protecting theirunborn child from HIV infection as the major motivatorfor taking their drug as prescribed. The effect of vomitingand nausea of pregnancy was not noticed in this study,as majority of the women are referred patients after HIVdiagnosis. At the time, these women present to NIMR HIVtreatment centre, the nausea and vomiting of pregnancywould have subsided.

The free services offered at NIMR, compared to othercentres, may have played a great role, as cost of drugs andlaboratory services have been shown to be a major barrierto antiretroviral drug adherence [18, 20]. The adherencerate of 99% reported by Jones and Barthlomew [19] among194 pregnant women where services were completely freegives credence to this assertion. The adherence level of 99%reported above also showed that we need to do a lot of workon issues of adherence. Presently, all patients are counseledprior to the initiation of antiretroviral therapy, but thereis need to reevaluate counseling methods and techniques

Page 14: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Journal of Pregnancy 5

Table 5: Factors associated with good adherence among the respondents.

FactorsAdherent respondents (%)

N = 137Nonadherent respondents

(%) N = 33OR (95% CI) P value

Age (years)

(i) <20 2 (1.5) 0 1.00

(ii) 20–35 123 (89.8) 28 (84.9) 1.00

(iii) ≥35 12 (8.8) 5 (15.1) 2.56 (0.85–7.85) 0.08

Number of living children

(i) 0 42 (30.7) 9 (27.3) 0.82 (0.28–2.44) 0.88

(ii) 1–2 57 (41.6) 10 (30.3) 1.00

(iii) >2 38 (27.7) 14 (42.4) 2.10 (0.78–5.75) 0.17

Marital status

(i) Married 119 (86.9) 12 (36.4) (1.13–6.79) 0.23

(ii) Not married 18 (13.1) 21 (63.6) 1.00

Occupation

(i) Unemployed 28 (20.4) 12 (36.4) 1.00

(ii) Employed/Working 109 (79.6) 21 (63.6) 0.45 (0.18–1.11) 0.09

Educational levelcompleted

(i) Less than secondary 26 (19.0) 6 (18.2) 1.00

(ii) At least Secondary 111 (81.0) 27 (81.8) 1.05 (0.36–3.18) 0.89

HIV status disclosure

(i) Disclosed 131 (95.6) 16 (48.5) (3.06–23.6) 0.000

(ii) Not disclosed 6 (4.4) 17 (51.5) 1.00

Previous PMTCTexperience

(i) Yes 31 (22.6) 3 (9.0) 2.92 (0.78–12.92) 0.13

(ii) No 106 (77.4) 30 (91.0) 1.00

Knowledge of HIV andART

(i) Good 124 (90.5) 22 (66.7) 4.77 (1.73–13.22) 0.001

(ii) Poor 13 (9.5) 11 (33.3) 1.00

Has treatment supporter

Yes 85 (62.0) 10 (30.3) 3.76 (1.55–9.26) 0.002

No 52 (38.0) 23 (69.7) 1.00

where necessary, aiming to achieve the 99% adherence levelreported by Jones and Barthlomew [19].

Encouraging the women to come along with a treatmentsupporter for the counseling sessions prior to initiation ofantiretroviral therapy preferably the partners would help ineducating the partners appropriately and improving adher-ence in the long run.

Apart from reviewing the counseling techniques, theknowledge of the adherence counselors also needs to be eval-uated as their beliefs and attitudes are central to effectivecounseling.

Among the nonadherent women, similar reasons report-ed in other studies for missing drugs were found [4, 10,11, 18]. It, therefore, shows that pregnancy related factorsare not the reasons for missing antiretroviral drugs duringpregnancy, but as a result of other personal and socioculturalfactors. Stigma and discrimination remain an important

factor militating against quality HIV care, as 63.6% ofwomen in this study expressed the reason for missing theirdrugs as afraid of being identified as HIV positive. Thereis, therefore, the need for continued campaign against stig-ma and discrimination if we must improve adherence toantiretroviral drugs and uptake of other HIV-related services.

The findings of HIV status disclosure and having atreatment support as factors associated with good adherenceafter controlling for potential confounders is in agreementwith previous studies [8]. With the disclosure of HIV statusto partners, who in most cases is the husband, he will notonly provide support but will act as treatment partner forthe spouse. We, therefore, need to encourage these women, todisclose their status to get the maximal benefit of disclosure.It is important, however, to note that women should not beforced to disclose their status, as HIV status disclosure hasbeen reported to be accompanied by partner violence [21].

Page 15: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

6 Journal of Pregnancy

Instead women who decline to disclose should be counseledand encouraged until they feel safe to disclose.

It is important to state that self-report used to measureadherence is not the gold standard for adherence measure-ment, but it has been reported to be sufficient enough toassess adherence when pill count is not possible and elec-tronic devices and blood ARV blood measurement are notfeasible [15].

5. Conclusion

In conclusion, this study has shown that good adherence isachievable during pregnancy, however, more has to be doneto achieve adherence rate of 99% reported elsewhere. Theidentified reasons for nonadherence and the effects of statusdisclosure and having a treatment partner will be used toimprove adherence counseling process.

References

[1] J. I. Ikechebelu, J. O. Ugboaja, S. O. Kalu, and D. O. Igboelina,“Outcome of prevention of mother to child transmission(PMTCT) of HIV infection programme in Nnewi, SoutheastNigeria,” in Proceedings of the 43rd Annual General Meetingand Scientific Conference of the Society of Gynaecology andObstetrics of Nigeria held in Kano, Novmber 2009.

[2] N. Abdulsalami and O. Tekena, “The epidemiology ofHIV/AIDS in Nigeria,” in AIDS in Nigeria: A Nation on theThreshold, O. Adeyi, P. Kanki, O. Odutolu, and J. Idoko,Eds., Havard Center for Population and Development Studies,Cambridge, Mass, USA, 2006.

[3] J. Homsy, D. Moore, A. Barasa et al., “Mother-to child HIVtransmission and infant mortality among women receivinghighly active antiretroviral therapy (HAART) in rural Uganda(abstract TUPE0354),” in Proceedings of the XVI InternationalAIDS Conference, Toronto, Canada, August 2006.

[4] A. O. Igwegbe, J. O. Ugboaja, and L. A. Nwajiaku, “Prevalenceand determinants of non-adherence to antiretroviral therapyamong HIV- positive pregnant women in Nnewi, Nigeria,”International Journal of Medicine and Medical Sciences, vol. 2,no. 8, pp. 238–245, 2010.

[5] L. Hayman, “Adherence to antiretrovirals among us womenduring and after pregnancy,” MCN The American Journal ofMaternal/Child Nursing, vol. 34, no. 1, p. 69, 2009.

[6] D. A. Murphy, K. J. Roberts, D. J. Martin, W. Marelich, and D.Hoffman, “Barriers to antiretroviral adherence among HIV-infected adults,” AIDS Patient Care and STDs, vol. 14, no. 1,pp. 47–58, 2000.

[7] N. B. Palmer, “Special group—drug use, mental illness,pregnancy,” in Medication Adherence in HIV/AIDS, L. Jeffrey,Ed., Liebert, Larchmont, NY, USA, 2004.

[8] J. B. Nachega, A. R. Knowlton, A. Deluca et al., “Treatmentsupporter to improve adherence to antiretroviral therapy inHIV-infected South African adults: a qualitative study,” Jour-nal of Acquired Immune Deficiency Syndromes, vol. 43, supple-ment 1, pp. S127–S133, 2006.

[9] W. L. Holzemer, I. B. Corless, K. M. Nokes et al., “Predictorsof self-reported adherence in persons living with HIV disease,”AIDS Patient Care and STDs, vol. 13, no. 3, pp. 185–197, 1999.

[10] S. A. Olowookere, A. A. Fatiregun, J. O. Akinyemi, A. E. Bamg-boye, and G. K. Osagbemi, “Prevalence and determinants ofnon adherence to highly active antiretroviral therapy amongpeople living with HIV/AIDS in Ibadan, Nigeria,” The Journal

of Infection in Developing Countries, vol. 2, no. 5, pp. 369–372,2008.

[11] V. N. Shaahu, T. O. Lawoyin, and A. O. Sangowawa, “Adher-ence to highly active antiretroviral therapy (HAAT) at a federalmedical centre,” African Journal of Medicine and MedicalSciences, vol. 37, no. 1, pp. 29–36, 2008.

[12] Z. Iliyasu, M. Kabir, I. S. Abubakar, M. Babashani, and Z. A.Zubair, “Compliance to antiretroviral therapy among AIDSpatients in aminu kano teaching hospital, Kano, Nigeria,”Nigerian Journal of Medicine, vol. 14, no. 3, pp. 290–294, 2005.

[13] Federal ministry of Health, Nigeria, National guideline forHIV and AIDS treatment and care in adolescents and adults,2007.

[14] E. O. Idigbe, T. A. Adewole, G. Eisen et al., “Manage-ment of HIV-1 infection with a combination of nevirapine,stavudine, and lamivudine: a preliminary report on theNigerian antiretroviral program,” Journal of Acquired ImmuneDeficiency Syndromes, vol. 40, no. 1, pp. 65–69, 2005.

[15] O. Ezechi, D. Onwujekwe, N. Odunnukwe et al., “Self-reported adherence to HAART is a reliable and cost effectivetool in low resource setting,” in Proceedings of the 3rd Interna-tional AIDS Society Conference on HIV Pathogensis, Treatmentand Prevention, July 2005, Abstract No. WePe12.7C35.

[16] J. A. Bartlett, “Addressing the challenges of adherence,” Journalof Acquired Immune Deficiency Syndromes, vol. 29, supplement1, pp. S2–S10, 2002.

[17] K. A. Gebo, J. Keruly, and R. D. Moore, “Association of socialstress, illicit drug use, and health beliefs with nonadherenceto antiretroviral therapy,” Journal of General Internal Medicine,vol. 18, no. 2, pp. 104–111, 2003.

[18] E. Markos, A. Worku, and G. Davey, “Adherence to ART inPLWHA at Yirgalem Hospital, South Ethiopia. Ethiop,” Jour-nal of Health & Development, vol. 22, no. 2, pp. 174–179, 2008.

[19] A. Jones and C. Barthlomew, “Adherence to antiretroviraltherapy in pregnant woment at an HIV treatment clinic,” inProceedings of the 3rd International AIDS Society Conferenceon HIV Pathogensis, Treatment and Prevention, Port of SpainTrinidad, July 2005, Abstract No. TuPe 5.5p18.

[20] J Day, N. Godoka, P. Nyamafeni et al., “Adherence to ARTin clinical trial settings in Zimbabwe and Uganda: lessonslearned,” in Proceedings of the International AIDS Conference,Bangkok, Thailand, July 2002.

[21] O. C. Ezechi, C. Gab-Okafor, D. I. Onwujekwe, R. A. Adu,E. Amadi, and E. Herbertson, “Intimate partner violence andcorrelates in pregnant HIV positive Nigerians,” Archives ofGynecology and Obstetrics, vol. 280, no. 5, pp. 745–752, 2009.

Page 16: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Hindawi Publishing CorporationJournal of PregnancyVolume 2012, Article ID 379271, 7 pagesdoi:10.1155/2012/379271

Review Article

Tuberculosis in Pregnancy: A Review

Olabisi M. Loto and Ibraheem Awowole

Department of Obstetrics and Gynaecology, Obafemi Awolowo University Teaching Hospital, P.M.B. 5538, Ile-Ife, Osun State, Nigeria

Correspondence should be addressed to Olabisi M. Loto, [email protected] and Ibraheem Awowole, [email protected]

Received 9 May 2011; Accepted 1 September 2011

Academic Editor: Oliver Ezechi

Copyright © 2012 O. M. Loto and I. Awowole. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Tuberculosis (TB) was declared a public health emergency by WHO in 2005. The disease is a significant contributor to maternalmortality and is among the three leading causes of death among women aged 15–45 years in high burden areas. The exact incidenceof tuberculosis in pregnancy, though not readily available, is expected to be as high as in the general population. Diagnosis oftuberculosis in pregnancy may be challenging, as the symptoms may initially be ascribed to the pregnancy, and the normal weightgain in pregnancy may temporarily mask the associated weight loss. Obstetric complications of TB include spontaneous abortion,small for date uterus, preterm labour, low birth weight, and increased neonatal mortality. Congenital TB though rare, is associatedwith high perinatal mortality. Rifampicin, INH and Ethambutol are the first line drugs while Pyrazinamide use in pregnancy isgaining popularity. Isoniazid preventive therapy is a WHO innovation aimed at reducing the infection in HIV positive pregnantwomen. Babies born to this mother should be commenced on INH prophylaxis for six months, after which they are vaccinated withBCG if they test negative. Successful control of TB demands improved living conditions, public enlightenment, primary preventionof HIV/AIDS and BCG vaccination.

1. Introduction

Tuberculosis (TB) is believed to be nearly as old as humanhistory. Traces of it in Egyptian mummies date back to about7000 years ago, when it was described as phthisis by Hippo-crates [1]. It was declared a public health emergency in theAfrican Region in 2005 [1] and has since continued to be amajor cause of disability and death. About 9.4 million newcases of tuberculosis were diagnosed in 2009 alone and 1.7million people reportedly died from the disease in the sameyear, translating to about 4700 deaths per day [2].

About one-third of the world’s population (estimated tobe about 1.75 billion) is infected with the tubercule bacillus[3]. As much as 75% of individuals with TB are within theeconomically productive age group of 15 to 54 years. This sig-nificantly impairs socioeconomic development, thereby per-petuating the poverty cycle [4].

Tuberculosis has been on the rise in tandem with HIV/AIDS. This is because people with HIV/AIDS, whose im-mune systems are weakened have with a 20–37 times therisk of developing a progressive disease compared with HIV-negative individuals [4].

2. Microbiology of Tuberculosis

Mycobacterium tuberculosis, an aerobic, non-spore-forming,nonmotile bacillus, is one of five members of the Mycobac-terium tuberculosis complex, others being M. bovis, M. ulcer-ans, M. Africanum, and M. microti, though M. tuberculosis isthe major human pathogen. It belongs to the family Myco-bacteriaceae. Other Mycobacterium species that may infecthumans include Mycobacterium leprae, M. avium, M. Intra-cellulare, and M. scrofulaceum.

3. Pathophysiology

Tuberculosis affects almost every organ in the body, but theusual site of the disease is the lungs, accounting for more than80 percent of tuberculosis cases [5]. The pattern of the infec-tion in HIV positive patients may, however, be different, withincreasing trends towards extrapulmonary spread [6].

Almost all tuberculosis infections are caused by inhala-tion of infectious particles aerosolized by coughing, sneezing,talking, or manipulation of infected tissues. Other modalities

Page 17: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

2 Journal of Pregnancy

of transmission may, however, include ingestion of unpas-teurised milk and direct implantation through skin abrasionor the conjunctiva. Aerosolized tuberculosis particles withsizes ranging between 1 and 5 um are carried to the terminalair spaces of high-airflow areas, where multiplication of thetubercle occurs. Following phagocytosis by pulmonary ma-crophages, a granulomatous reaction may be initiated, inconjunction with the regional lymph nodes, thereby formingthe Ghon’s focus. The bacilli remain in a state of dormancywithin the Ghon’s focus, from where they may later becomereactivated.

4. Tuberculosis in Pregnancy

The wide array of opinion of Medical practitioners on tuber-culosis in pregnancy simply reflects the Public Health signif-icance of the condition. It is best described as a doubled-edged sword, one blade being the effect of tuberculosis onpregnancy and the pattern of growth of the newborn, whilethe other is the effect of pregnancy on the progression oftuberculosis.

Tuberculosis not only accounts for a significant propor-tion of the global burden of disease, it is also a significantcontributor to maternal mortality, with the disease beingamong the three leading causes of death among women aged15–45 years [2].

The exact incidence of tuberculosis in pregnancy is notreadily available in many countries due to a lot of confound-ing factors. It is, however, expected that the incidence of tu-berculosis among pregnant women would be as high as inthe general population, with possibly higher incidence indeveloping countries.

Earlier study by Schaefer reported a new case rate of18–29/100,000 in pregnancy, which was similar to the 19–39/100,000 reported for the city of New York [7]. A recentUnited Kingdom study, however, quoted an incidence of 4.2per 100,000 maternities [8], which may be a reflection thecurrent global fall in the incidence of the disease [2].

5. Effects of Pregnancy on Tuberculosis

Researchers from the days of Hippocrates have expressedtheir worries about the untoward effects that pregnancy mayhave on preexisting tuberculosis. Pulmonary cavities result-ing from tuberculosis were believed to collapse as a resultof the increased intra-abdominal pressure associated withpregnancy. This belief was widely held till the beginning ofthe fourteenth century! Indeed, a German physician rec-ommended that young women with TB should get marriedand become pregnant to slow the progression of the disease.This was practiced in many areas till the 19th century [9],while in the early 20th century, induced abortion was rec-ommended for these women [10, 11]. Researchers likeHedvall [12] and Schaefer [7], however, demonstrated no netbenefit or adverse effect of pregnancy on the progression ofTB. Frequent, consecutive pregnancies may, however, have anegative effect, as they may promote recrudescence or reacti-vation of latent tuberculosis.

It is, however, important to note that the diagnosis of tu-berculosis in pregnancy may be more challenging, as thesymptoms may initially be ascribed to the pregnancy. Theweight loss associated with the disease may also be temporar-ily masked by the normal weight gain in pregnancy.

6. Effects of Tuberculosis on Pregnancy

The effects of TB on pregnancy may be influenced by manyfactors, including the severity of the disease, how advancedthe pregnancy has gone at the time of diagnosis, the presenceof extrapulmonary spread, and HIV coinfection and thetreatment instituted.

The worst prognosis is recorded in women in whom a di-agnosis of advanced disease is made in the puerperium aswell as those with HIV coinfection. Failure to comply withtreatment also worsens the prognosis [13].

Other obstetric complications that have been reported inthese women include a higher rate of spontaneous abortion,small for date uterus, and suboptimal weight gain in preg-nancy [14, 15]. Others include preterm labour, low birthweight and increased neonatal mortality [13]. Late diagnosisis an independent factor, which may increase obstetric mor-bidity about fourfolds, while the risk of preterm labour maybe increased ninefolds [15–18].

7. Tuberculosis and the Newborn

Congenital tuberculosis is a rare complication of in utero tu-berculosis infection [19] while the risk of postnatal trans-mission is significantly higher [20]. Congenital tuberculosismay be as a result of haematogenous spread through the um-bilical vein to the foetal liver or by ingestion and aspiration ofinfected amniotic fluid [21]. A primary focus subsequentlydevelops in the liver, with involvement of the peri-portallymph nodes. The tubercle bacilli infect the lungs secondar-ily, unlike in adults where over 80% of the primary infectionsoccur in the lungs [5].

Congenital tuberculosis may be difficult to distinguishfrom other neonatal or congenital infections from whichsimilar symptoms may arise in the second to the third weekof life. These symptoms include hepato-splenomegaly, res-piratory distress, fever, and lymphadenopathy. Radiographicabnormalities may also be present but these generally appearlater [13]. The diagnosis of neonatal tuberculosis may, how-ever, be facilitated by employing a set of diagnostic criteriadeveloped by Cantwell et al. [22], including the demonstra-tion of primary hepatic complex/caseating granuloma onpercutaneous liver biopsy at birth, tuberculous infection ofthe placenta, or maternal genital tract tuberculosis, and thedemonstration of lesions during the first week of life. Thepossibility of postnatal transmission must be excluded bya thorough investigation of all contacts, including hospitalstaffs and attendants.

As much as half of the neonates delivered with congenitaltuberculosis may eventually die, especially in the absence oftreatment [7, 23–26].

Page 18: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Journal of Pregnancy 3

8. Diagnosis of Tuberculosis in Pregnancy

To diagnose this condition, history of exposure to individualswith chronic cough or recent visit to areas endemic with tu-berculosis should be obtained. History of symptoms, whichis likely to be the same as in nonpregnant women, is alsoessential. Caution must, however, be exercised, as thesesymptoms may be nonspecific in pregnancy [27, 28]. Thesesymptoms include night sweat, evening pyrexia, haemopt-ysis, progressive weight loss, and chronic cough of over 3weeks duration. There may also be a history of ineffectiveattempts at antibiotics therapy [27, 29].

In pregnant women with suggestive symptoms and signsof TB, a tuberculin skin test should be carried out. This hassince been accepted to be safe in pregnancy [21, 30]. Thedebate, however, is about the sensitivity of tuberculin testduring pregnancy. Earlier reports suggested diminished tu-berculin sensitivity in pregnancy [31], while recent studiesrevealed no significant differences in the pregnant and non-pregnant populations [27, 32–35].

The two types of tuberculin skin tests are discussedbelow.

8.1. Tine Test. This test utilises an instrument with multipleneedles that are dipped in a purified form of the TB bacteriacalled old tuberculin (OT). The skin is pricked with theseneedles and the reaction is analysed 48–72 hours later. Itis, however, no longer popular except in large populationscreening.

8.2. Mantoux Test. A single-needle intradermal injection of0.1 mL of purified protein derivative (5 Tuberculin units) isadministered, and the skin reaction is analysed 48–72 hourslater, based on the largest diameter of the indurations de-veloped. It is a more accurate and reproducible test than theTines test.

False-positive results may be obtained in individuals whohad previously been vaccinated with the BCG vaccine, thosewith previously treated tuberculosis, as well as in people withinfection from other Mycobacterium species. False negativeson the other hand are commonly due to a compromised im-mune system and technical errors [36].

A chest radiograph with abdominal lead shield maybe done after the tuberculin skin testing, though pregnantwomen are more likely to experience a delay in obtaining achest X-ray due to concerns about fetal health [27].

Microscopic examination of sputum or other specimenfor Acid-fast bacilli (AFB) remains the cornerstone of labo-ratory diagnosis of TB in pregnancy. Three samples of spu-tum should be submitted for smear, culture, and drug-sus-ceptibility testing. Staining for AFB is also done, using theZiehl-Neelsen, fluorescent, Auramine-Rhodamine, and theKinyoun techniques [37]. Light-emitting diode (LED) fluo-rescent microscopy has recently been introduced to improvediagnosis [37]. According to the WHO’s 2009 report onglobal TB control, the percentage of new cases of smear-posi-tive TB detected ranged between 56 and 68%. The stainingtechniques may, therefore, not sufficient for the diagnosis ofTB, as smear-negative cases will be missed [37].

9. Culture

The traditional culture on Lowenstein-Jensen’s medium maytake 4–6 weeks to obtain a result. This may, however, still beuseful in cases of diagnostic doubts and management ofsuspected drug-resistant tuberculosis [38]. Newer diagnostictools are now available to facilitate diagnosis, including theliquid Bactec culture medium, which has been endorsed byWHO. Other culture media that could be used include themodified Lowenstein’s medium, Petragnani medium, Tru-deau Committee medium, Peizer’s medium, Dubos Middle-brook media, Tarshis blood agar, Middlebrook’s 7-H3, Mid-dlebrook’s 7-H9, and Middlebrook’s 7H-10 media [38]. Liq-uidisation and decontamination with N-Acetytl-L-Cysteinein 1% Sodium Hydroxide solution before inoculation mayenhance sensitivity [38].

M. tuberculosis produces niacin and heat-sensitive cata-lase and it lacks pigment. It may, therefore, be differentiatedfrom other mycobacterium species using these features.Others include reduction of nitrates and its isoniazide sen-sitivity, which may, however, not be reliable in cases of INHresistance.

Molecular Line Probe Assay (LPA) as well as the use ofpolymerase chain reaction (PCR) are presently facilitatingthe specific identification of the tubercle bacilli [37].

10. Treatment of Tuberculosis

“Untreated tuberculosis represents a far greater hazard to apregnant woman and her fetus than does treatment of thedisease” [39].

The management of tuberculosis in pregnancy is a mul-tidisciplinary approach, with the team comprising the obste-trician, communicable disease specialty personnel, neonatol-ogists, counselling unit, and public health officials.

Treatment is achieved through the use of Directly Ob-served Therapy, Short Course (DOTS). This therapy entailsthe use of combination therapy for at least 6 months, de-pending on the combination of antituberculous agents thatare available. This combination includes isoniazide and ri-fampicin compulsorily, supported by ethambutol and pyraz-inamide [40–44].

For patients with drug-susceptible TB and good drugadherence, these regimens will cure around 90% of TB cases.Treatment is done on out-patient basis, unless otherwise in-dicated [37].

The use of these first-line antituberculous drugs in preg-nancy are considered safe for the mother and the baby byThe British Thoracic Society, International Union AgainstTuberculosis and Lung Disease, and the World Health Or-ganisation [16, 45].

10.1. Isoniazide. INH is safe during pregnancy even in thefirst trimester, though it can cross the placenta [11]. Thewomen must, however, be followed up because of the possi-bility of INH-induced hepatotoxicity. Pyridoxine supple-mentation is recommended for all pregnant women takingINH at a dose of 50 mg daily [39, 46].

Page 19: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

4 Journal of Pregnancy

10.2. Rifampicin. This is also believed to be safe in pregnan-cy, though in an unknown proportion of cases, there maybe an increased risk of haemorrhagic disorders in the new-born (some authorities prescribe supplemental vitamin K(10 mg/day) for the last four to eight weeks of pregnancy.)while some other researchers reported the possibility of limbdeformity but none of these are in excess of what is obtainedin the normal population.

10.3. Ethambutol. The retrobulbar neuritis that may compli-cate the use of this drug in adults generated the fear thatit may interfere with ophthalmological development whenused in pregnancy but this has not been demonstrated whenthe standard dose is used. This was also confirmed in experi-mental studies on some abortuses [47].

10.4. Pyrazinamide. The use of pyrazinamide in pregnancywas avoided by many physicians for a long time due to un-availability of adequate data on its teratogenicity. Presently,many international organizations now recommend its use,including the International Union Against Tuberculosis AndLung diseases (IUATLD), British Thoracic Society, AmericanThoracic Society, the World Health Organisation as well asthe Revised National Tuberculosis Control Programme ofIndia. There are no reports of significant adverse events fromthe use of this drug in the treatment of TB in pregnantwomen despite its use as part of the standard regimen inmany countries [48].

Its use is particularly indicated in women with tubercu-lous meningitis in pregnancy, HIV coinfection, and sus-pected INH resistance [49–52]. Breastfed infants of motherson antituberculous therapy should, however, be monitoredfor jaundice, which may suggest drug-induced hepatitis, aswell as joint pains resulting from drug-induced hyperuri-caemia.

10.5. Streptomycin. The drug has been proven to be poten-tially teratogenic throughout pregnancy. It causes fetal mal-formations and eighth-nerve paralysis, with deficits rangingfrom mild hearing loss to bilateral deafness. Many centres areagainst the use of this drug in pregnancy [49, 53, 54].

11. Multidrug-Resistant Tuberculosisin Pregnancy (MDR-TB)

Pregnant women with MDR-TB have a less favourable prog-nosis [55]. They may sometimes require treatment withsecond-line drugs, including cycloserine, ofloxacin, ami-kacin, kanamycin, capreomycin, and ethionamide. The safetyof these drugs is unfortunately not well-established in preg-nancy [49].

Para-amino salicylic acid had been used as combinationtherapy with INH in pregnancy in the past without any sig-nificant teratogenic side effects, though maternal gastroin-testinal side effects may be pronounced.

Ethionamide is associated with growth retardation, cen-tral nervous system and skeletal abnormalities in animalstudies involving rats and rabbits [56, 57]. Human studies

also demonstrated increased central nervous system defectsfollowing its use in early pregnancy [58]. Its use is, therefore,not recommended in pregnancy.

Therapeutic abortion has been proposed as an option ofmanagement for these women [59], as MDR-TB poses morerisk to the woman and the society at large. Another optionis to delay initiating treatment to the second trimester wherepossible [10]. Individualised Treatment Regimen (ITR) usingvarious combinations of the 2nd line antituberculous agentsbased on their susceptibility profile had, however, been triedin some pregnant women with no adverse obstetric outcome[60].

The outlook for those patients is expected to improve asexperience and knowledge in the management of the condi-tion increases.

12. Treatment of TB in Lactating Women

Breastfeeding is simply the cheapest and healthiest way tofeed a baby. The final decision on breastfeed must, therefore,be taken with necessary input from the neonatologists, ob-stetricians, and pharmacologists. The American Academy ofPediatrics recommends that women with tuberculosis whohave been treated appropriately for two weeks or more andwho are not considered contagious may breastfeed [61],while the RNTCP recommends breast-feeding of neonatesregardless of the mother’s TB status [62].

Antituberculous drugs are excreted into breast milk,though the dose is less compared with the therapeutic dosefor infants. Breastfed infants may receive as much as 20% ofthe therapeutic dose of INH for infants, while other anti-tuberculous drugs are less excreted. No toxicity has beenreported from this small concentration in breast milk [49].Caution must, however, be exercised as the breast milk dosemay contribute to the development of abnormally high plas-ma levels in newborns who are on antituberculous medica-tions. To minimise this possibility, the mother may take hermedications immediately after a feed and substitute a bottlefor the next feed. She may then return to her usual pattern offeeding [49, 63].

Pyridoxine deficiency may cause seizures in the newborn.Supplemental pyridoxine should, therefore, be administeredto infants on INH or whose mother is taking the drug.

Breastfeeding may be discouraged in women who are yetto commence treatment at the time of delivery and those whoare still actively excreting the bacillus while coughing. It mayalso be discouraged as part of a prevention of mother to childtransmission in HIV coinfection and women with tuberculo-sis of the lactiferous ducts or glands.

In the absence of evidence of congenital tuberculosis,isoniazide (10 mg/kg/day) should be commenced at birthand continued for six months. Clinical or radiological fea-tures of active tuberculosis and a positive tuberculin skin testare indications for a full course of anti-tuberculous treat-ment. The tuberculin skin test and chest X-rays are done at 6weeks, 12 weeks, and 6 months. The baby is vaccinated withBCG at 6 months if these tests are negative. The baby is, how-ever, changed to multiple drug therapy if any of these teststurn positive during the period of monitoring.

Page 20: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Journal of Pregnancy 5

13. HIV and TB Coinfection in Pregnancy

HIV and TB are inextricably linked. Their effect is even moredeadly in pregnancy, when they may contribute significantlyto maternal morbidity and mortality. Over 50% of the mater-nal mortality occurring in mothers with TB in pregnancy isdue to coinfection with HIV [64]. Moreover, treatment iscomplicated by the challenges of adherence, polypharmacyand the overlapping side effect profiles of antituberculosisand antiretroviral drugs [65–67].

The key concern is about the interactions between therifamycins and antituberculous drugs. The suboptimal out-comes of therapeutic trials without a rifamycin has made theuse of the drug mandatory, even in the face of drug inter-actions [68, 69].

The spectrum of antiretroviral drugs available for use inpregnancy is limited. Efavirenz is contraindicated before thethirteenth week of gestation, while the risk of toxicity fromthe use of didanosine and stavudine is significantly increasedin pregnancy. Rifampicin may cause a reduction in the serumconcentration of efavirenz, though, increasing the dose ofefavirenz does not result in any significant outcome [70].

Nevirapine, which is an alternative to the use of efavirenz,also exhibits some drug interaction with rifampicin. Rifam-picin may lead to the reduction of serum concentration ofnevirapine by as much as 50%. To circumvent this problem,rifabutin, another rifamycin that is as effective as rifampicinin the treatment of tuberculosis may be used, as the drug hasless effect on the CYP3A system that metabolizes nevirapine[71].

Generally, there is a dearth of studies and data on howpregnancy may affect the aforementioned interactions. Cau-tion is, therefore, of great importance when managing preg-nant women with this cruel duo.

14. Prevention of Tuberculosis

The BCG vaccine has been incorporated into the Nationalimmunization policy of many countries, especially the highburden countries, thereby conferring active immunity fromchildhood. Nonimmune women travelling to tuberculosisendemic countries should also be vaccinated. It must, how-ever, be noted that the vaccine is contraindicated in preg-nancy [72].

The prevention, however, goes beyond this as it is es-sentially a disease of poverty. Improved living condition is,therefore, encouraged with good ventilation, while over-crowding should be avoided. Improvement in nutritional sta-tus is another important aspect of the prevention.

Pregnant women living with HIV are at higher risk forTB, which can adversely influence maternal and perinataloutcomes [73]. As much as 1.1 million people were diag-nosed with the co-infection in 2009 alone [2]. Primary pre-vention of HIV/AIDS is, therefore, another major step in theprevention of tuberculosis in pregnancy. Screening of allpregnant women living with HIV for active tuberculosis isrecommended even in the absence of overt clinical signs ofthe disease.

Isoniazid preventive therapy (IPT) is another innovationof the World Health Organisation that is aimed at reducingthe infection in HIV positive pregnant women based on ev-idence and experience and it has been concluded that preg-nancy should not be a contraindication to receiving IPT.However, patient’s individualisation and rational clinicaljudgement is required for decisions such as the best time toprovide IPT to pregnant women [73].

Most importantly, governments commitments are highlyencouraged so that the World Health Organisation and allother international bodies involved in fighting tuberculosismay succeed in chasing this monster out of all communities.

References

[1] “Facts about health in African Subregion,” Fact sheet N◦314World Health Organisation, 2011.

[2] “2010/2011 tuberculosis global fact; World Health Organ-isation,” http://www.who.int/tb/country/en/index.html, Nov.2010.

[3] “Global tuberculosis control: a short update to the 2009report,” Tech. Rep., World Health Organization, Geneva,Switzerland, (WHO/HTM/TB/2009.426), 2009.

[4] “Global tuberculosis control 2010,” Tech. Rep., World HealthOrganization, Geneva, Switzerland, (WHO/HTM/TB/2010.),2010.

[5] National Tuberculosis and Leprosy Control Programme,Workers Manual, Federal Ministry of Health of Nigeria, De-partment of Public Health, Nigeria, Abuja, 4th edition, 2004.

[6] P. G. Marjorie and R. V. Holenarasipur, “Extrapulmonary tu-berculosis: an overview,” American Family Physician, vol. 72,no. 9, pp. 1761–1768, 2005.

[7] G. Schaefer, I. A. Zervoudakis, F. F. Fuchs, and S. David,“Pregnancy and pulmonary tuberculosis,” Obstetrics and Gy-necology, vol. 46, no. 6, pp. 706–715, 1975.

[8] M. Knight, J. J. Kurinczuk, and C. Nelson-Piercy, “Tuberculo-sis in pregnancy in the UK,” BJOG, vol. 116, no. 4, pp. 584–588, 2009.

[9] D. Snider Jr., “Pregnancy and tuberculosis,” Chest, vol. 86, no.3, 1984.

[10] J. G. Vallejo and J. R. Starke, “Tuberculosis and pregnancy,”Clinics in Chest Medicine, vol. 13, no. 4, pp. 693–707, 1992.

[11] D. E. Snider, P. M. Layde, M. W. Johnson, and M. A.Lyle, “Treatment of tuberculosis during pregnancy,” AmericanReview of Respiratory Disease, vol. 122, no. 1, pp. 65–79, 1980.

[12] E. Hedvall, “Pregnancy and tuberculosis,” Acta Medica Scandi-navica, vol. 147, supplement 1286, pp. 1–101, 1953.

[13] P. Ormerod, “Tuberculosis in pregnancy and the puerperium,”Thorax, vol. 56, no. 6, pp. 494–499, 2001.

[14] N. K. Jain, “Safety of anti-tuberculosis drugs in pregnancy,” inProceedings of the National Conference on Pulmonary Diseases(NAPCON ’01), vol. 33, Mumbai, Maharashtra, 2001.

[15] J. Kishan, Sailaja, and S. Kaur, “Tuberculosis and pregnancy,”in Proceedings of the National Conference on PulmonaryDiseases (NAPCON ’01), Mumbai, Maharashtra, Nov 2001.

[16] Health Protection Agency, Pregnancy and Tuberculosis: Guid-ance for Clinicians, Health Protection Agency, London, UK,2006.

[17] T. Pillay, M. Khan, J. Moodley, M. Adhikari, and H. Coovadia,“Perinatal tuberculosis and HIV-1: considerations for re-source-limited settings,” Lancet Infectious Diseases, vol. 4, no.3, pp. 155–165, 2004.

Page 21: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

6 Journal of Pregnancy

[18] N. Jana, K. Vasishta, S. K. Jindal et al., “Perinatal outcome inpregnancies complicated by pulmonary tuberculosis,” Interna-tional Journal of Gynecology and Obstetrics, vol. 44, no. 2, pp.119–124, 1994.

[19] M. Cantwell, D. E. Snider Jr., G. M. Cauthen, and I. M.Onorato, “Epidemiology of tuberculosis in the United States,1985 through 1992,” JAMA, vol. 272, no. 7, pp. 535–539, 1994.

[20] J. R. Starke, “Tuberculosis: an old disease but a new threat tothe mother, fetus, and neonate,” Clinics in Perinatology, vol. 24,no. 1, pp. 107–127, 1997.

[21] M. A. Hamadeh and J. Glassroth, “Tuberculosis and preg-nancy,” Chest, vol. 101, no. 4, pp. 1114–1120, 1992.

[22] M. R. Cantwell, Z. M. Shehab, A. M. Costello et al., “Briefreport: congenital tuberculosis,” New England Journal of Med-icine, vol. 330, no. 15, pp. 1051–1054, 1994.

[23] R. Nemir and D. O’Hare, “Congenital tuberculosis,” AmericanJournal of Diseases of Children, vol. 139, no. 3, pp. 284–287,1985.

[24] B. L. Varudkar, “Short course chemotherapy for tuberculosisin children,” The Indian Journal of Pediatrics, vol. 52, no. 419,pp. 593–597, 1985.

[25] Medical Research Council Tuberculosis and Chest DiseaseUnit, “Management and outcome of chemotherapy for child-hood tuberculosis,” Archives of Disease in Childhood, vol. 64,no. 7, pp. 1004–1012, 1989.

[26] J. Biddulph, “Short course chemotherapy for childhood tuber-culosis,” Pediatric Infectious Disease Journal, vol. 9, no. 11, pp.793–801, 1990.

[27] R. F. C. Doveren and R. Block, “Tuberculosis and pregnan-cy: a provincial study (1990-1996),” Netherlands Journal ofMedicine, vol. 52, no. 3, pp. 100–106, 1998.

[28] J. T. Good, M. D. Iseman, and P. T. Davidson, “Tuberculosisin association with pregnancy,” American Journal of Obstetricsand Gynecology, vol. 140, no. 5, pp. 492–498, 1981.

[29] E. J. Carter and S. Mates, “Tuberculosis during pregnancy: theRhode Island experience, 1987 to 1991,” Chest, vol. 106, no. 5,pp. 1466–1470, 1994.

[30] M. T. Medchill and M. Gillum, “Diagnosis and management oftuberculosis during pregnancy,” Obstetrical and GynecologicalSurvey, vol. 44, no. 2, pp. 81–84, 1989.

[31] R. Finn, C. A. St Hill, A. Govan, I. G. Ralfs, F. J. Gurney, and V.Denye, “Immunological responses in pregnancy and survivalof fetal homograft,” British medical journal, vol. 3, no. 819, pp.150–152, 1972.

[32] P. Present and G. W. Comstock, “Tuberculin sensitivity inpregnancy,” American Review of Respiratory Disease, vol. 112,no. 3, pp. 413–416, 1975.

[33] N. L. Eriksen and A. W. Helfgott, “Cutaneous anergy inpregnant and non-pregnant women with human immunode-ficiency virus,” Infectious Diseases in Obstetrics and Gynecology,vol. 6, pp. 13–17, 1998.

[34] L. M. Mofenson, E. M. Rodriguez, R. Hershow et al., “Myco-bacterium tuberculosis infection in pregnant and nonpreg-nant women infected with HIV in the women and infantstransmission study,” Archives of Internal Medicine, vol. 155, no.10, pp. 1066–1071, 1995.

[35] T. E. Nolan, T. L. Espinosa, and J. G. Pastorek, “Tuberculosisskin testing in pregnancy: trends in a population,” Journal ofPerinatology, vol. 17, no. 3, pp. 199–201, 1997.

[36] J. B. Bass Jr., L. S. Farer, P. C. Hopewell et al., “Treatment oftuberculosis and tuberculosis infection in adults and children,”American Journal of Respiratory and Critical Care Medicine,vol. 149, no. 5, pp. 1359–1374, 1994.

[37] The Global Plan to Stop Tb 2011-2015: Transforming the FightTowards Elimination of Tuberculosis, World Health Organiza-tion, Geneva, Switzerland, 2010.

[38] Pathways to Better Diagnostics for Tuberculosis; A Blueprint forDevelopment of TB Diagnostics, World Health Organization,Geneva, Switzerland, 2009.

[39] Centre for Disease Control, “Treatment of tuberculosis,”MMWR, vol. 52, no. RR-11, pp. 1–77, 2003.

[40] Joint Tuberculosis Committee of the British Thoracic Soci-ety, “Chemotherapy and management of tuberculosis: recom-mendations 1998,” Thorax, vol. 53, pp. 536–548, 1998.

[41] Singapore Tuberculosis Service and British Medical ResearchCouncil, “Clinical trial of six-month and four-month chem-otherapy in the treatment of pulmonary tuberculosis. Theresults up to 30 months ,” Tubercle, vol. 62, pp. 95–102, 1981.

[42] East and Central African and British Medical Research Coun-cil, “Clinical controlled trial of four short course regimens ofchemotherapy (three six-months and one eight months) forpulmonary tuberculosis,” Tubercle, vol. 64, pp. 153–166, 1983.

[43] British Thoracic Society, “A controlled trial of 6 month’schemotherapy in pulmonary tuberculosis. Final report: resultsduring the 36 months after the end of chemotherapy and be-yond,” British Journal of Diseases of the Chest, vol. 78, no. 4, pp.330–336, 1984.

[44] L. P. Ormerod, “Chemotherapy of tuberculosis,” European Re-spiratory Monograph, vol. 2, supplement 4, pp. 273–297, 1997.

[45] J. H. Tran and P. Montakantikul, “The safety of antitubercu-losis medications during breastfeeding,” Journal of HumanLactation, vol. 14, no. 4, pp. 337–340, 1998.

[46] G. Bothamley, “Drug treatment for tuberculosis during preg-nancy: safety considerations,” Drug Safety, vol. 24, no. 7, pp.553–565, 2001.

[47] T. Lewitt, L. Nebel, S. Terracina, and S. Karman, “Ethambutolin pregnancy: observations on embryogenesis,” Chest, vol. 66,no. 1, pp. 25–26, 1974.

[48] G. D. Anderson, “Tuberculosis in pregnancy,” Seminars in Per-inatology, vol. 21, no. 4, pp. 328–335, 1997.

[49] Management, Control and Prevention of Tuberculosis; Guide-lines for Health Care Providers (2002–2005), Department ofHuman Services, Victoria, Australia, 2002.

[50] H. M. Blumberg, W. J. Burman, R. E. Chaisson et al.,“American Thoracic Society/Centers for Disease Control andPrevention/Infectious Diseases Society of America: treatmentof tuberculosis,” American Journal of Respiratory and CriticalCare Medicine, vol. 167, no. 4, pp. 603–662, 2003.

[51] Anon, “Treatment of tuberculosis,” MMWR Recommendationsand Reports, vol. 52, pp. 1–77, 2003.

[52] J. G. Bartlett, “Guidelines section,” Infectious Diseases in Clini-cal Practice, vol. 11, pp. 467–471, 2002.

[53] D. Heilman, F. Heilman, H. Hinshaw et al., “Streptomycin:absorption, diffusion, excretion and toxicity,” The AmericanJournal of the Medical Sciences, vol. 210, pp. 576–584, 1945.

[54] J. Woltz and M. Wiley, “Transmission of streptomycin frommaternal blood to the fetal circulation and amniotic fluid,”Proceedings of the Society for Experimental Biology and Medi-cine, vol. 60, pp. 106–107, 1945.

[55] J. T. Good, M. D. Iseman, and P. T. Davidson, “Tuberculosisin association with pregnancy,” American Journal of Obstetricsand Gynecology, vol. 140, no. 5, pp. 492–498, 1981.

[56] I. Khan and A. Azam, “Study of teratogenic activity of tri-fluoperazine,amitriptyline, ethionamide and thalidomide inpregnant rabbits and mice,” Proceedings of the European Societyfor the Study of Drug Toxicity, vol. 10, pp. 235–242, 1969.

Page 22: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Journal of Pregnancy 7

[57] S. Takekoschi, “Effects of hydroxymethylpyrimidine on isoni-azid and ethionamide induced teratosis,” Gunma Journal ofMedical Sciences, vol. 14, pp. 233–244, 1965.

[58] J. L. Schardein, Chemically Induced Birth Defects, MarcelDekker, New York, NY, USA, 3rd edition, 2000.

[59] J. T. Good, M. D. Iseman, and P. T. Davidson, “Tuberculosis inassociation with pregnancy,” American Journal of Obstetricsand Gynecology, vol. 140, no. 5, pp. 492–498, 1981.

[60] S. Sonya, D. Guerra, M. Rich et al., “Treatment of multidrug-resistant tuberculosis during pregnancy: a report of 7 cases,”Clinical Infectious Diseases, vol. 36, no. 8, pp. 996–1003, 2003.

[61] L. K. Pickering, C. J. Baker, S. S. Long, and J. A. McMillan,Red Book: 2006 Report of the Committee on Infectious Diseases,American Academy of Pediatrics, Elk Grove Village, Ill, USA,27th edition, 2006.

[62] Central TB Division, Managing the Revised National Tuber-culosis Control Programme in your area – A training course;Modules 1-4, New Delhi, India, 2001.

[63] Management, Control and Prevention of Tuberculosis. Guide-lines for Health Care Providers (2002–2005), Department ofHuman Services, Victoria, Australia, 2002.

[64] M. Khan, T. Pillay, J. M. Moodley, C. A. Connolly, and theDurban Perinatal TB HIV-1 Study Group, “Maternal mortalityassociated with tuberculosis-HIV-1 co-infection in Durban,South Africa,” AIDS, vol. 15, no. 14, pp. 1857–1863, 2001.

[65] “CDC. Managing Drug Interactions in the Treatment of HIV-Related Tuberculosis [online],” http://www.cdc.gov/tb/TBDrugs/default.htm, 2007.

[66] Departments of Health, Immunisation Against Infectious Dis-ease, HMSO, London ,UK, 1996.

[67] W. J. Burman, “Issues in the management of HIV-related tu-berculosis,” Clinics in Chest Medicine, vol. 26, no. 2, pp. 283–294, 2005.

[68] A. Jindani, A. J. Nunn, and D. A. Enarson, “Two 8-month reg-imens of chemotherapy for treatment of newly diagnosed pul-monary tuberculosis: international multicentre randomisedtrial,” Lancet, vol. 364, no. 9441, pp. 1244–1251, 2004.

[69] A. Okwera, C. Whalen, F. Byekwaso, and The Makerere Uni-versity-Case Western University Research Collaboration,“Randomised trial of thiacetazone and rifampicincontainingregimens for pulmonary tuberculosis in HIV-infected Ugan-dans,” Lancet, vol. 344, no. 8933, pp. 1323–1328, 1994.

[70] L. Sarner and A. Fakoya, “Acute onset lactic acidosis andpancreatitis in the third trimester of pregnancy in HIV-1positive women taking antiretroviral medication,” SexuallyTransmitted Infections, vol. 78, no. 1, pp. 58–59, 2002.

[71] E. Perucca, R. Grimaldi, G. M. Frigo, A. Sardi, H. Monig,and E. E. Ohnhaus, “Comparative effects of rifabutin and ri-fampicin on hepatic microsomal enzyme activity in normalsubjects,” European Journal of Clinical Pharmacology, vol. 34,no. 6, pp. 595–599, 1988.

[72] A. Gupta, U. Nayak, M. Ram et al., “Postpartum tuberculosisincidence and mortality among HIV-infected women andtheir infants in Pune, India, 2002-2005,” Clinical InfectiousDiseases, vol. 45, no. 2, pp. 241–249, 2007.

[73] WHO, Guideline for intensified tuberculosis case finding andisoniazide preventive therapy for people living with HIV, WHO,Geneva, Switzerland, 2011.

Page 23: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Hindawi Publishing CorporationJournal of PregnancyVolume 2012, Article ID 508657, 6 pagesdoi:10.1155/2012/508657

Research Article

The Impact of HIV on Maternal Morbidity inthe Pre-HAART Era in Uganda

Harriet Nuwagaba-Biribonwoha,1 Richard T. Mayon-White,2 Pius Okong,3

Peter Brocklehurst,4 and Lucy M. Carpenter5

1 ICAP, Mailman School of Public Health, Columbia University, P.O. Box 3567, Kampala, Uganda2 Department of Primary Health Care, University of Oxford, Oxford, OX1 2-ET, UK3 St. Raphael of St. Francis Hospital, Nsambya, Kampala, Uganda4 Institute for Women’s Health, University College London, London, WC1E AU, UK5 Department of Public Health, University of Oxford, Oxford, OX3 7LF, UK

Correspondence should be addressed to Harriet Nuwagaba-Biribonwoha, [email protected]

Received 13 June 2011; Accepted 20 July 2011

Academic Editor: Oliver Ezechi

Copyright © 2012 Harriet Nuwagaba-Biribonwoha et al. This is an open access article distributed under the Creative CommonsAttribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work isproperly cited.

Objective. To compare maternal morbidity in HIV-infected and uninfected pregnant women. Methods. Major maternal morbidity(severe febrile illness, illnesses requiring hospital admissions, surgical revisions, or illnesses resulting in death) was measuredprospectively in a cohort of HIV-infected and uninfected women followed from 36 weeks of pregnancy to 6 weeks after delivery.Odds ratios of major morbidity and associated factors were examined using logistic regression. Results. Major morbidity wasobserved in 46/129 (36%) and 104/390 (27%) of the HIV-infected and HIV-uninfected women, respectively, who remained infollowup. In the multivariable analysis, major morbidity was independently associated with HIV infection, adjusted odds ratio(AOR) 1.7 (1.1 to 2.7), nulliparity (AOR 2.0 (1.3 to 3.0)), and lack of, or minimal, formal education (AOR 2.1 (1.1 to 3.8)).Conclusions. HIV was associated with a 70% increase in the odds of major maternal morbidity in these Ugandan mothers.

1. Introduction

Maternal morbidity is defined as illness in a woman who wasor is pregnant from any cause related to the pregnancy, abor-tion, or child birth, excluding incidental or accidental causes[1]. In Uganda, the true level of maternal morbidity is un-known, but national estimates suggest that 5.1 maternaldeaths occur for every 1,000 live births [2]. Although therehas been a decline in maternal mortality since 2000 [3],child bearing is still associated with significant morbidity andmortality in this country. Like most sub-Saharan countries,Uganda has experienced a heavy burden of HIV-relateddisease in the last few decades. In men and non-pregnantwomen, HIV has been associated with at least a 2-foldincrease in morbidity [4, 5] and 10–20-fold increase inmortality [6, 7]. Less is known about the association betweenHIV and maternal morbidity. We undertook a prospec-tive observational study to describe, estimate, and compare

the risk of maternal morbidity in HIV-infected and unin-fected women in Uganda.

2. Methods

2.1. Setting. The study was conducted at St. Raphael of St.Francis Hospital Nsambya, in Kampala, Uganda, betweenNovember 2002 and November 2003. The hospital is a catho-lic missionary hospital located in the capital city (Kampala).During the study year, there were a total of 24,461 antenatalvisits (of whom 6,016 were new bookings) and 10,768 deliv-eries recorded. The hospital provided care to general and pri-vate patients who paid at total of about $10 (general) and $40(private) for antenatal care and a normal vaginal delivery.This hospital was one of the pilot sites for the Ugan-dan national programme to prevent mother-to-child HIVtransmission (PMTCT) and the main source of antenatal,

Page 24: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

2 Journal of Pregnancy

intrapartum and postnatal care for the local population [8].During the study period all women attending the hospitalfor antenatal care were offered free opt-in HIV testing andcounselling. Women were classified as HIV-infected if theyhad two positive rapid tests. HIV-infected women were givenmodified intrapartum obstetric care by delaying the ruptureof membranes and avoiding episiotomies when possible. Themain PMTCT regimen offered to women at the time waszidovudine monotherapy from 36 weeks of pregnancy to1 week after delivery or a single dose of nevirapine givento HIV-infected women at the onset of labour. The use ofhighly active antiretroviral therapy (HAART) in pregnancywas not yet widely implemented. Exclusive breastfeeding wasencouraged for both HIV-infected and uninfected mothers,but infected mothers could also give replacement feeds ifafter assessment this mode of feeding was deemed feasible.

2.2. Study Design. Pregnant women with known HIV statusand willing to participate in the study were recruited intoa prospective cohort at about 36 weeks of pregnancy andfollowed up until six weeks after the delivery. As maternalmorbidity is known to be associated with parity [9], thecohort was restricted to nulliparous (para 0) and uniparous(para 1) women. HIV-infected and uninfected mothers wererecruited in the ratio of 1 to 3. The main exposure of interestwas the woman’s HIV status and the primary outcome wasany observed or self-reported major maternal morbidity. Awoman was classified as having major maternal morbidityif she experienced one or more of the following: illnessleading to hospital admission, severe febrile illness (feverlasting 3 days or longer), surgical revisions of abdominal orperineal wounds due to wound breakdown, and any illnessesthat resulted in death. These maternal conditions have beenclassified as major morbidity by other authors [10–12].

Baseline data including demographic characteristics (na-tionality, residence, and age), socioeconomic confounders(education, marital status, and employment), medical his-tory, and obstetric history were collected by interview inthe antenatal clinic at 36–40 weeks of pregnancy using asemistructured questionnaire. CD4 tests were offered toHIV-infected women at baseline and analysed using rou-tine standard laboratory procedures. Delivery data wereabstracted from hospital records. A second interview washeld six weeks after delivery for women to report ontheir morbidity during the study period. Women who didnot attend the hospital for delivery and/or the 6 weekspostpartum appointment were telephoned and/or visited athome immediately after the missed hospital appointment.

The data were analysed using SPSS 11.5. Characteristicsof HIV-infected and uninfected women were comparedusing either binary measures (percentages or odds ratios)where statistical significance of differences was examinedwith chi-squared test statistics, or means where differenceswere compared using F or t test statistics. In addition, CD4measures were summarised using medians and interquartileranges (IQR). Fisher’s exact test was used when expectedvalues in contingency table cells were less than 5, and Yates’correction was applied to chi-square tests on 2 × 2 tables.

Odds ratios were first estimated crudely, without adjustmentfor potential confounders and then with adjustment forconfounders (adjusted odds ratios, AOR) using logisticregression models.

The study was approved in the UK by the OxfordshireTropical Research Ethics Committee (OXTREC) and inUganda by the National AIDS Research Ethics Committee,the National Council of Science and Technology, and St.Raphael of St. Francis Hospital Nsambya administration.

3. Results

The characteristics of the 132 HIV-infected women and399 HIV-uninfected women recruited into the cohort overthe 12 month study period are summarized in Table 1.There were statistically significant differences between HIV-infected and uninfected women in parity, mean gestationalage at recruitment, education, and infant feeding practices(Table 1). The majority of the HIV-infected women (100/132,76%) were classified as asymptomatic (WHO stage 1). In106 (80%) of HIV-infected women for whom CD4 counts at36–40 weeks of pregnancy were available, the median CD4count was 350/uL, (IQR 221 to 552/uL); 71 (67%) of thewomen had CD4 count >200/uL. PMTCT regimen data wereavailable for 125 HIV infected women and the majority tookmonotherapy with AZT 74 (59%) or Nevirapine 48 (38.4%).Only 3 (2.4%) were on HAART and excluding them from theanalysis had no significant impact on the results.

At 6 weeks after delivery, follow-up data was availablefor 129/132 (97%) HIV-infected and 389/399 (97%) HIVuninfected women. Of these women, 46/129 (36%) whowere HIV-infected, and 104/390 (27%) who were uninfectedreported major morbidity. The crude odds ratio for anymajor morbidity in HIV-infected women compared to unin-fected women was 1.52 (95% confidence interval 0.97 to 2.38,P = 0.05). The conditions that contributed to major morbid-ity are shown in Table 2. Two of the 6 hospital admissionsamong HIV-infected women occurred before delivery andwere due to severe malaria. The remaining four admissionsoccurring after delivery were reported as due to severemalaria (2), wound sepsis and disruption (1) and a high fever(1). Eight of the 11 HIV-uninfected women were admitted tohospital before delivery with severe malaria (2), false labour(2), preeclampsia (1), urinary tract infection (1), diarrhoea(1), and psychosis (1). The three admissions after delivery inthese women were due to severe malaria (1), wound sepsisand disruption (1), and postpartum haemorrhage (1).

Severe febrile illness was the commonest contributor tomajor morbidity (Table 2), and was experienced by 34%(44/129) and 24% (93/390) of the HIV infected and unin-fected women respectively. It occurred more often after deliv-ery in 89% (39/44) of the HIV-infected women and 84%(78/93) of uninfected women. The crude odds ratio for se-vere febrile illness after delivery among HIV-infected womencompared to uninfected women was 1.65 (95% confidenceinterval 1.05–2.60, P = 0.02). Severe febrile illness was lesscommon in women who had Caesarean section deliveries(20/104, 19%) than those delivered vaginally (117/381, 31%)

Page 25: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Journal of Pregnancy 3

Table 1: Characteristics of HIV-infected and uninfected women in the cohort study.

HIV infectedN = 132 (% or SD)

HIV uninfectedN = 399 (% or SD)

P value

Sociodemographic characteristics

Private patients 16 (12) 46 (12) 1.0

Mean maternal age (standard deviation (SD)) 24.9 (3.8) 23.5 (3.7) 0.002

Nulliparous 56 (42) 250 (63) <0.001

4 or more antenatal visits in this pregnancy 94 (71) 273 (68) 0.6

Mean gestational age in weeks at recruitment(standard deviation)

36.1 (1.9) 36.7 (1.3) <0.001

Highest level of formal education

Primary school or no formal education 36 (27) 79 (20)

Secondary school 58 (44) 165 (41) 0.02

Tertiary (higher) education 38 (29) 155 (39)

Women without own income 75 (57) 233 (58) 0.9

Married or cohabiting 102 (77) 341 (86) 0.04

Single 32 (23) 59 (14)

Place of delivery

St. Francis Hospital∗ 115 (87) 344 (86) 1.0

Home or other health care setting 16 (13) 50 (12)

Mode of delivery∗

Vaginal delivery 88 (77) 251 (79)

Elective Caesarean section 6 (5) 7 (2) 0.2

Emergency Caesarean section 20 (18) 65 (19)

Infant feeding practices 6 weeks after delivery∗∗

Exclusive breast feeding 75 (63) 299 (80)

Replacement feeding 39 (32) 8 (2) <0.001

Mixed feeding 6 (5) 68 (18)∗

Mode of delivery data only available for 459 births that occurred at St. Francis Hospital.∗∗Infant feeding practices data only available for 120 and 375 births.

Table 2: Major morbidity in HIV-infected and uninfected women by 6 weeks after delivery.

HIV infectedN = 129 (%)

HIV uninfectedN = 390 (%)

Crude odds ratio (95% CI) P value

Hospital admissions 6 (5) 11 (4) 1.68 (0.54–5.04) 0.31

Severe febrile illness 44 (34) 93 (24) 1.65 (1.05–2.60) 0.02

Surgical revision 3 (2) 7 (2) 1.30 (0.26–5.68) 0.71∗

Illnesses causing death 2 (2) 1 (0) 6.13 (0.43–172.02) 0.15∗

Major morbidity(women with any of the above)

46 (36) 104 (27) 1.52 (0.97–2.38) 0.05

∗Fisher’s exact 2-tailed P value.

P = 0.03, and in women who were not exclusively breastfeed-ing (15/100, 15%) than in those breastfeeding (96/370, 26%),P = 0.03. Three HIV-infected women and 6 uninfectedwomen had surgical revisions of abdominal incisions; oneuninfected woman had an episiotomy resutured.

Three deaths occurred after delivery, two of which werein HIV-infected mothers where the cause appeared AIDSrelated: one with pneumonia and the other with a high fevercombined with vomiting and severe oral ulceration. The onedeath of an HIV-uninfected mother was a primigravidaewoman who died from eclampsia and severe anaemia after

delivering twins. Unadjusted analyses of each potential con-founding factor revealed higher odds for any major mor-bidity among nulliparous women and women with no, orminimal, formal education while women aged 25–29 yearsseemed less likely to have major morbidity (not shown).

When potential confounders (age, parity, private patientsversus general patients, employment status, and education)were adjusted for, HIV status remained significantly associ-ated with major morbidity among women at 6 weeks afterdelivery (Table 3), the adjusted odds ratio was 1.7 (95% CI1.1 to 2.7), P = 0.02. Nulliparity (P = 0.003) and lack of or

Page 26: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

4 Journal of Pregnancy

Table 3: Major morbidity due to HIV infection adjusted for selected demographic factors.

N = 519 Adjusted odds ratio (95% CI) P value

HIV infection 129 1.7 (1.1–2.7) 0.02

Nulliparity 299 2.0 (1.3–3.0) 0.003

Private patients 60 0.9 (0.4–1.7) 0.64

Women without own income 300 0.8 (0.5–1.3) 0.33

Education

Tertiary education 110 Reference —

Secondary education 220 1.9 (1.1–3.1) 0.02

Primary or no education 189 2.1 (1.1–3.8) 0.02

Age

Age <20 years 68 Reference —

Age 20–24 236 1.1 (0.6–2.0) 0.77

Age 25–29 178 0.7 (0.4–1.5) 0.43

Age 30+ 37 0.7 (0.3–2.1) 0.57

minimal, formal education (P = 0.02) remained significantlyassociated with higher odds for major morbidity in theseadjusted analyses.

Baseline CD4 counts were available for 103 (80%) of the129 HIV-infected women who remained in follow-up. HIV-infected women who had major morbidity had significantlylower CD4 counts than those that did not have any majormorbidity. The mean baseline CD4 count for the formergroup was 312/uL (standard deviation 177/uL) while in thelatter group mean CD4 count was 456/ul (standard deviation245/uL), P = 0.002. Major morbidity among HIV-infectedwomen stratified by WHO HIV clinical stage at recruitmentwas 29/97 (30%) among women in WHO clinical stage I,7/15 (58%) among women in WHO clinical stage II, and10/16 (63%) among women in WHO clinical stage III, (chi-square test for linear trend = 7.9, df = 1, P = 0.005).

4. Discussion

This cohort study estimated the odds of any major mor-bidity to be 70% higher among HIV-infected women inUganda. The commonest major morbidity was severe febrileillness, which occurred in 34% of HIV-infected and 24%of uninfected women. These results are in line with whathas been found in other sub-Saharan African countries onthe impact of HIV on major maternal morbidity leading tomortality [13, 14]. A five year audit in South Africa showedthat maternal mortality (meaning women experienced majormorbidity) was 6-times higher among HIV-infected womenthan infected women [15]. In Mozambique and Malawi,HIV-related illness was the leading cause of death amongwomen who did not die of obstetric complications [16, 17].HIV can aggravate other prevalent debilitating conditionsin pregnancy like anaemia, thereby increasing severity ofmaternal morbidity [18, 19]. Similar findings have beenobserved in developed countries with HIV-infected pregnantwomen having a higher risk of hospitalisation, and longerstay during hospitalisation than HIV-uninfected womendespite the use of HAART [20, 21]. Our study quantified thelikely magnitude of the deleterious effect of HIV on maternal

morbidity in Uganda, highlighting the need for interventionsto address maternal morbidity as PMTCT programs scaleup, even after HAART is widely available [21, 22]. In thiscohort study, HIV-infected women delivered by Caesareansection had fewer severe febrile illnesses after delivery thanthose delivered vaginally. This finding is contrary to what hasbeen observed by others [23–25], but in our study womenwho delivered by Caesarean section routinely had antibioticprophylaxis while those who delivered vaginally did not.

The strengths of this study lay in its prospective natureand in having a comparison group of women who were notinfected with HIV. The same set of structured questions wereused for both HIV-infected and uninfected women to limitbiased inquiry about and reporting of maternal morbiditybecause HIV status of the mothers was generally knownto the interviewers. Some limitations of the study shouldbe noted: hospital admissions could overestimate majormorbidity if women are admitted for minor conditions [26],but this was unlikely in our cohort where the financial andsocial costs of inpatient care are generally a deterrent toadmissions in this setting. Laboratory investigations werelimited in this study which meant that the study relied onclinical features and maternal reports. However, the resultsare applicable to resource-limited settings where manypatients are managed similarly. Finally, this study was notdesigned or powered to examine maternal morbidity by ARTstatus or HIV disease stage, but the stratified analysis impliedhigher major maternal morbidity with more advanced HIVdisease. Further studies are needed to provide data on thisas the combination of maternal morbidity and HIV-relatedmorbidity can lead to a significantly negative impact onmaternal health and survival. This study was limited to anurban missionary hospital in the capital city so future studiesshould include rural and government health facilities wherethe impact of HIV on maternal morbidity could be higher.

5. Conclusion

HIV infection increased the odds of major maternal morbid-ity by 70% in this cohort of Ugandan mothers giving birth

Page 27: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Journal of Pregnancy 5

for the first or second time. Maternal morbidity remains asignificant health issue in such resource-limited settings andthere is need for more interventions to address the effects ofHIV infection on maternal morbidity.

Funding

The Rhodes Trust funded the first author’s scholarship atthe University of Oxford, and the Public Health ResearchFund Oxfordshire NHS Charitable Foundation funded thefieldwork.

Acknowledgments

The authors thank The Rhodes Trust which funded the firstauthor’s scholarship at the University of Oxford, the PublicHealth Research Fund Oxfordshire NHS Charitable Founda-tion for funding the fieldwork, the Ugandan research teamA. Kiyingi, B. E. Mwebaze, H. Kalanzi, and M. Mugenyithe study participants and the administration of St. FrancisHospital Nsambya for their cooperation. The staff of theDepartment of Public Health, University of Oxford, OldRoad Campus where the first author was based during thisresearch are also gratefully acknowledged.

References

[1] WHO, “Progress in reproductive health research,” 2001, http://www.who.int/hrp/en/.

[2] Uganda Bureaux of Statistics (UBOS) and O. Macro, UgandaDemographic and Health Survey 2000-2001, Macro Interna-tional, Calverton, Maryland, Md, USA, 2001.

[3] M. C. Hogan, K. J. Foreman, M. Naghavi et al., “Maternalmortality for 181 countries, 1980–2008: a systematic analysisof progress towards Millennium Development Goal 5,” TheLancet, vol. 375, no. 9726, pp. 1609–1623, 2010.

[4] D. Morgan, C. Mahe, B. Mayanja, and J. A. G. Whitworth,“Progression to symptomatic disease in people infected withHIV-1 in rural Uganda: prospective cohort study,” BritishMedical Journal, vol. 324, no. 7331, pp. 193–196, 2002.

[5] H. U. Wagner, A. Kamali, A. J. Nunn, J. F. Kengeya-Kayondo,and D. W. Mulder, “General and HIV-1-associated morbidityin a rural Ugandan community,” AIDS, vol. 7, no. 11, pp.1461–1467, 1993.

[6] N. K. Sewankambo, R. H. Gray, S. Ahmad et al., “Mortalityassociated with HIV infection in rural Rakai District, Uganda,”AIDS, vol. 14, no. 15, pp. 2391–2400, 2000.

[7] A. J. Nunn, D. W. Mulder, A. Kamali, A. Ruberantwari, J. F.Kengeya-Kayondo, and J. Whitworth, “Mortality associatedwith HIV-1 infection over five years in a rural Ugandan pop-ulation: cohort study,” British Medical Journal, vol. 315, no.7111, pp. 767–771, 1997.

[8] Nsambya Hospital, Annual Report, 2001.[9] J. Bai, F. W. S. Wong, A. Bauman, and M. Mohsin, “Parity

and pregnancy outcomes,” American Journal of Obstetrics andGynecology, vol. 186, no. 2, pp. 274–278, 2002.

[10] J. Bhatia and J. Cleland, “The contribution of reproductive ill-health to the overall burden of perceived illness among womenin southern India,” Bulletin of the World Health Organization,vol. 79, no. 11, pp. 1065–1069, 2001.

[11] T. A. Grubert, D. Reindell, R. Kastner, R. Lutz-Friedrich, B.H. Belohradsky, and O. Dathe, “Complications after caesareansection in HIV-1-infected women not taking antiretroviraltreatment,” The Lancet, vol. 354, no. 9190, pp. 1612–1613,1999.

[12] E. J. Rodriguez, C. Spann, D. Jamieson, and M. Lindsay, “Post-operative morbidity associated with cesarean delivery amonghuman immunodeficiency virus-seropositive women,” Amer-ican Journal of Obstetrics and Gynecology, vol. 184, no. 6, pp.1108–1111, 2001.

[13] A. Coutsoudis, K. England, N. Rollins, H. Coovadia, M. L.Newell, and R. Bland, “Women’s morbidity and mortality inthe first 2 years after delivery according to HIV status,” AIDS,vol. 24, no. 18, pp. 2859–2866, 2010.

[14] D. Chilongozi, L. Wang, L. Brown et al., “Morbidity and mor-tality among a cohort of human immunodeficiency virus type1-infected and uninfected pregnant women and their infantsfrom Malawi, Zambia, and Tanzania,” Pediatric InfectiousDisease Journal, vol. 27, no. 9, pp. 808–814, 2008.

[15] V. Black, S. Brooke, and M. F. Chersich, “Effect of human im-munodeficiency virus treatment on maternal mortality at atertiary center in South Africa: a 5-year audit,” Obstetrics andGynecology, vol. 114, no. 2, pp. 292–299, 2009.

[16] C. Menendez, C. Romagosa, M. R. Ismail et al., “An autopsystudy of maternal mortality in Mozambique: the contributionof infectious diseases,” PLoS Medicine, vol. 5, no. 2, article e44,2008.

[17] E. J. Kongnyuy, G. Mlava, and N. van den Broek, “Facility-based maternal death review in three districts in the centralregion of Malawi. An analysis of causes and characteristics ofmaternal deaths,” Women’s Health Issues, vol. 19, no. 1, pp. 14–20, 2009.

[18] C. J. Uneke, D. D. Duhlinska, and E. B. Igbinedion, “Prevalenceand public-health significance of HIV infection and anaemiaamong pregnant women attending antenatal clinics in south-eastern Nigeria,” Journal of Health, Population and Nutrition,vol. 25, no. 3, pp. 328–335, 2007.

[19] D. Naniche, A. Bardajı, M. Lahuerta et al., “Impact of maternalhuman immunodeficiency virus infection on birth outcomesand infant survival in rural Mozambique,” American Journalof Tropical Medicine and Hygiene, vol. 80, no. 5, pp. 870–876,2009.

[20] S. Fiore, “Higher rates of post-partum complications in HIV-infected than in uninfected women irrespective of mode ofdelivery,” AIDS, vol. 18, no. 6, pp. 933–938, 2004.

[21] P. Bansil, D. J. Jamieson, S. F. Posner, and A. P. Kourtis,“Hospitalizations of pregnant HIV-infected women in theUnited States in the era of highly active antiretroviral therapy(HAART),” Journal of Women’s Health, vol. 16, no. 2, pp. 159–162, 2007.

[22] J. Moodley, R. Pattinson, C. Baxter, S. Sibeko, and Q. AbdoolKarim, “Strengthening HIV services for pregnant women: anopportunity to reduce maternal mortality rates in SouthernAfrica/sub-Saharan Africa,” International Journal of Obstetricsand Gynaecology, vol. 118, no. 2, pp. 219–225, 2011.

[23] J. Louis, M. B. Landon, R. J. Gersnoviez et al., “Perioperativemorbidity and mortality among human immunodeficiencyvirus-infected women undergoing cesarean delivery,” Obstet-rics and Gynecology, vol. 110, no. 2, pp. 385–390, 2007.

[24] A. Marcollet, F. Goffinet, G. Firtion et al., “Differences in post-partum morbidity in women who are infected with the humanimmunodeficiency virus after elective cesarean delivery, emer-gency cesarean delivery, or vaginal delivery,” American Journalof Obstetrics and Gynecology, vol. 186, no. 4, pp. 784–789, 2002.

Page 28: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

6 Journal of Pregnancy

[25] J. S. Read, R. Tuomala, E. Kpamegan et al., “Mode of deliveryand postpartum morbidity among HIV-infected women: thewomen and infants transmission study,” Journal of AcquiredImmune Deficiency Syndromes, vol. 26, no. 3, pp. 236–245,2001.

[26] S. E. Geller, D. Rosenberg, S. M. Cox, and S. Kilpatrick, “Defin-ing a conceptual framework for near-miss maternal morbid-ity,” Journal of the American Medical Women’s Association, vol.57, no. 3, pp. 135–139, 2002.

Page 29: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Hindawi Publishing CorporationJournal of PregnancyVolume 2012, Article ID 565049, 5 pagesdoi:10.1155/2012/565049

Review Article

A Model of Tuberculosis Screening for Pregnant Women inResource-Limited Settings Using Xpert MTB/RIF

Eleanor R. Turnbull,1, 2 Nzali G. Kancheya,1 Jennifer B. Harris,1, 2 Stephanie M. Topp,1, 2

German Henostroza,1, 2 and Stewart E. Reid1, 2

1 Tuberculosis Department, Centre for Infectious Disease Research in Zambia, 5977 Benakale Road, P.O. Box 34681, Northmead,Lusaka, Zambia

2 Schools of Medicine and Public Health, University of Alabama at Birmingham, AL 35233, USA

Correspondence should be addressed to Eleanor R. Turnbull, [email protected]

Received 14 June 2011; Accepted 10 August 2011

Academic Editor: Oliver Ezechi

Copyright © 2012 Eleanor R. Turnbull et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Timely diagnosis and treatment of maternal tuberculosis (TB) is important to reduce morbidity and mortality for both themother and child, particularly in women who are coinfected with HIV. The World Health Organization (WHO) recommends theintegration of TB/HIV screening into antenatal services but available diagnostic tools are slow and insensitive, resulting in delaysin treatment initiation. Recently the WHO endorsed Xpert MTB/RIF, a highly sensitive, real-time PCR assay for Mycobacteriumtuberculosis that simultaneously detects rifampicin resistance directly from sputum and provides results within 100 minutes. Wepropose a model for same-day TB screening and diagnosis of all pregnant women at antenatal care using Xpert MTB/RIF. Pilotstudies are urgently required to evaluate strategies for the integration of TB screening into antenatal clinics using new diagnostictechnologies.

1. Introduction

Tuberculosis (TB) is a leading cause of nonobstetric maternaldeath in resource-limited settings, accounting globally forapproximately 700,000 deaths every year, the majority ofwhich are in areas with high HIV prevalence [1, 2]. In SouthAfrica, a screening study found that the prevalence of TBis 10 times greater in HIV-infected than in HIV-uninfectedpregnant women [3] and Gouder et al. found TB prevalenceamong women attending antenatal care was 696/100,000among HIV-infected women compared to 200/100,000among HIV-negative women [4]. If untreated, maternalTB can lead to increased neonatal mortality, lower birthweights, prematurity [5, 6] and increased complicationsof pregnancy, including a four-fold increase in maternalmorbidity through higher rates of abortion, postpartumhemorrhage, labor difficulties, and preeclampsia [7, 8].Furthermore studies have demonstrated that HIV-infectedpregnant women who are coinfected with TB are 2.5 timesmore likely to transmit HIV to their babies than women

without TB [9] and their infants are 24 times more likely tohave neonatal TB [10]. Under current standards in resource-limited settings there is a significant delay from the timeof presentation to the diagnosis of TB, due to the lowsensitivity and long turnaround time of available diagnostictools, the need for multiple visits, and the nonspecificity ofsymptoms in pregnant women, particularly those who areHIV infected. In Mexico, Figueroa-Damian found maternalmorbidity, neonatal mortality, and extreme prematurity allto be significantly higher among pregnant women with TBwho started treatment late in pregnancy (25–36 weeks ofgestation), whilst those treated early had minimal negativeoutcomes [11, 12]. In addition, rapid, early diagnosis andtreatment of TB reduces transmission of TB to familymembers, including newborns, and the wider community[13] and reduces the number of women who are lost tofollowup during the lengthy TB diagnostic process. In light ofthese risks, it is clear that early, rapid diagnosis and treatmentof both HIV and TB is critical to improve both maternal

Page 30: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

2 Journal of Pregnancy

and infant outcomes [14]. The World Health Organization(WHO) recommends the integration of TB/HIV care intoantenatal services and TB screening of all pregnant womenin high HIV prevalence areas [15].

2. TB Diagnosis in Pregnant Women

To date the implementation of WHO antenatal TB screen-ing recommendations has been limited in resource poorcountries due in part to significant financial and logisticalconstraints [16]. Clinicians are reluctant to use chest radio-graphy as part of the TB diagnostic workup process andsputum smear microscopy using Ziehl-Neelsen (Z-N) stainhas been demonstrated to have low sensitivity, especiallyin HIV-infected women. The WHO estimates that Z-Nmicroscopy detects only 58% of pulmonary TB cases inHIV-infected individuals [17]. Sensitivities even lower thanthis have been reported; in a South African study, eightof 370 HIV-infected pregnant women were diagnosed withculture-confirmed TB but all were smear negative [18].Despite being the diagnostic reference, standard TB cultureis not accessible in most resource-limited settings due to itscomplex laboratory requirements. A potential alternative isthe recently recommended Xpert MTB/RIF, a real-time PCRassay for Mycobacterium tuberculosis that simultaneouslydetects rifampicin resistance directly from sputum andprovides results within 100 minutes [19]. Xpert MTB/RIFhas been recommended for use up to the subdistrict level,especially in settings where rapid access to appropriatetreatment and care is required [20]. Results from prospectivedemonstration studies involving 6,648 individuals found thesensitivity of a single, direct Xpert MTB/RIF test in culture-positive cases was 90.3% (99.0% in smear-positive sputa and76.9% in smear-negative sputa), rifampicin resistance wasdetected with 94.4% sensitivity and 98.3% specificity andperformance was not significantly affected by HIV status [21,22]. Operationally Xpert MTB/RIF technology has shownto be robust in various locales and can be used outsidelaboratory settings by lay staff with minimal training [21].These characteristics suggest that Xpert MTB/RIF may havethe potential to play a significant role in the diagnosis of TBin antenatal settings in low-resource settings with generalizedHIV epidemics.

Using the example of Zambia, there are approximately5000 women who present monthly for antenatal care atthe 25 government primary healthcare clinics in the capital,Lusaka. In these health clinics the mean gestational age ofwomen at their first antenatal visit is 22 weeks and almosthalf (48%) come for only one visit [23]. Over 90% ofexpectant women agree to HIV testing during antenatalcare and approximately 22% are HIV infected. Despitewell-developed HIV testing programs in Maternal ChildHealthcare (MCH) departments, there is no provision forsystematic TB screening and limited data available withinZambia to guide best practice. Currently, pregnant womanpresenting to MCH with TB symptoms are referred to theoutpatient department (where there are often long queues)for further consultation, sputum collection, and antibiotictrials. Often multiple visits are required over several days to

weeks prior to receiving a final diagnosis and women may belost to followup before the diagnostic process is complete.

While there is no data on the diagnostic delay betweeninitial presentation and TB diagnosis in pregnant women,an evaluation of the 2007 WHO guidelines found thatmedian time to TB treatment initiation was between 3–17days for smear-negative pulmonary TB [24] and a reviewof records at one Lusaka HIV clinic found the averagetime from first presentation to pulmonary TB diagnosisto be 29 days (personal communication, S. Trollip). Dueto more complicated referral systems, this delay may beeven greater in MCH clinics The combination of insensitivetools and lengthy diagnostic delays highlights the urgentneed to integrate TB screening into antenatal care and topilot faster, more sensitive diagnostic tools. Introductionof the Xpert MTB/RIF technology, which provides same-day TB diagnosis with high sensitivity, could potentiallyallow pregnant women to be screened and started on TBtreatment the same day as their antenatal visit. Based oncurrent practice this would represent a reduction of up to 3weeks in time to TB treatment initiation, critically given thatmost women present late in pregnancy.

3. TB Screening Model in Antenatal Clinics

Screening pregnant women for TB in MCH makes sense bothclinically and logistically as even in low-resource settings amajority of women access health care during pregnancy atleast once [25]. Rather than strengthening the limited andseparate TB diagnostic services at the clinic, integration ofTB diagnostic and antenatal services would leverage existingclinic space and staff, simplify patient flow, and reducewaiting times.

We propose a model (Figure 1) to integrate TB and HIVscreening, diagnosis, and treatment into existing antenatalcare using Xpert MTB/RIF technology. Women present earlyin the morning to the antenatal clinic and while they waitfor individual check-ups, group health education is given bya trained lay worker/peer educator. These health educationtalks provide an opportunity to deliver counseling on theTB and HIV screening services and the clinical benefits toboth mother and child of TB (and HIV) screening duringpregnancy. In this model, lay healthcare workers administer asimple TB symptom questionnaire that evaluates patients forkey TB symptoms (any cough, fever, night sweats, or weightloss). Asymptomatic women are referred for provider-initiated counseling and testing (PITC) and continue withthe standard antenatal visit where they may be consideredfor isoniazid prophylaxis therapy. Symptomatic (presenceof any symptom) women are escorted to an isolated andwell-ventilated area for sputum collection and afterwardsfor PITC. Utilizing Xpert MTB/RIF technology, sputumis analyzed in the clinic laboratory or within the MCHdepartment, depending on space availability, security, andstable electricity supply. The clinician provides a formalassessment once TB and HIV test results are available.Women with positive TB and/or HIV test results are referredfor further evaluation and, optimally, TB treatment will beinitiated the same day.

Page 31: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Journal of Pregnancy 3

Screened for TB with XpertMTB/RIF

Xpert MTB positive,RIF resistant

Xpert MTB positive,RIF sensitive

Xpert MTBnegative

Refer to nationalMDR TB treatment

program

Women present at MCH forantenatal care

Sensitization talk given byLay-workers on importance of

HIV/TB screening and treatment

Woman has no TBsymptoms

HIV PITC conducted whilewomen wait for Xpert

MTB/RIF result

Women sees clinician forTB/HIV test results

Registration, triage and TBsymptom screening

questionnaire by Lay-workers

Woman has one or more TBsymptoms (any cough,fever, night sweats or

weight loss)

Refer to TB clinic.Refer for ART care

if HIV positive

Continue standardantenatal care atMCH. Refer forART care if HIV

positive

HIV PITC conducted

Continue standardantenatal care atMCH. Refer forART care if HIV

positive

Figure 1: Proposed model for same-day TB screening and diagnosis of pregnant women at antenatal care using Xpert MTB/RIF (∗MCH:Maternal Child Healthcare Clinic; (M)TB: (Mycobacterium) Tuberculosis; RIF: Rifampicin; PITC: Provider Initiated Testing and Counseling;ART: Antiretroviral therapy; MDR: Multidrug resistant).

4. Discussion

We have highlighted the need for TB screening in pregnantwomen and proposed a model for a point of care screeningalgorithm with rapid turnaround time. Variations of thismodel may be appropriate for different settings dependingon patient flow, space, and staffing. Faster diagnosis ofTB leads to earlier cotreatment, reduced transmission tocommunity and family, and potentially improved maternaland neonatal clinical outcomes. The proposed use of Xpert

MTB/RIF technology has substantial advantages in termsof sensitivity, turnaround time, and simplicity comparedto existing technologies however implementation challengesexist; the instrument and storage of cartridges requiressignificant space, waste generated is considerably more thanthat with microscopy, and instruments require an unin-terrupted power supply and annual validation. At present,Xpert MTB/RIF operation requires substantial expenditure,both in capital costs and consumables. In 2011, the Founda-tion for New Innovative Diagnostics (FIND) negotiated price

Page 32: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

4 Journal of Pregnancy

for a 4-module unit (capable of running 16–20 specimens perday) is 17,000 USD and the price per cartridge is 16.86 USD(with one cartridge required per sputum tested). On the longterm, prices for Xpert MTB/RIF are expected to decrease asglobal testing volumes increase [20]. Further work is requiredto examine its cost-effectiveness in various resource-limitedsettings. In addition, Xpert MTB/RIF technology relieson the ability of a patient to produce sputum which isproblematic for those without pulmonary symptoms. Futurepilots could also consider other less expensive nonsputum-based diagnostic tools that are currently under evaluation,such as a urine lateral flow test for simple and rapid lipoara-binomannan (LAM) detection [26] or simplified nucleicacid amplification tests like the loop-mediated isothermalamplification (LAMP) assay [27], after their endorsement bythe WHO.

The possible clinical benefits of an integrated systemare apparent and acceptance of screening is potentiallyhigh as pregnant women generally have increased healthseeking behavior to protect their unborn child; however itis important to recognize potential patient-related barriersto implementation. For example, Kali et al. found that44% of HIV-infected pregnant women were unwilling togo through TB screening [18]. However, TB screening wasoffered after PITC, suggesting that women may have beenreluctant to face another potential diagnosis immediatelyafter learning that they were HIV infected and may beunwilling to wait the additional 2 hours required to obtain anXpert MTB/RIF result. Women needing treatment for HIVmay be especially reluctant to engage in TB screening for fearof the heavy pill burden associated with TB/HIV treatment,during pregnancy. Screening all women for both diseaseson arrival, as proposed in this model, may increase uptakeand strengthen acceptance of the need for routine HIVand TB screening in high-prevalence settings. Furthermorestrong linkages to care and comprehensive sensitization atboth clinic and community level is essential to “normalize”TB screening in the pregnant population and overcomereluctance to screen.

Integration of HIV and TB screening services into MCHusing similar models has been shown to be feasible andrecommended in high HIV/TB prevalence settings [14, 16].A simple screening questionnaire that assesses patients forany of the four key TB symptoms has been proven toaccurately identify people in need of further diagnosticassessment in resource-constrained settings [28]. In additionthe use of lay workers to conduct a symptom screening hasbeen successfully piloted in South Africa where it addedonly 4–7 minutes to the visit time [18]. In Zambia layworkers/peer educators have been instrumental to the rapidscale-up of both HIV programs (conducting testing, adher-ence counseling, and followup of patients on treatment)[29, 30]; TB programs (conducting PITC, sputum collection,directly observed treatment strategy and facilitating referrallinkage). As Xpert MTB/RIF instruments are reported tobe simple enough to be run by lay workers this greatlyincreases the feasibility of using them in countries that havesignificant shortages of professional health care staff [21].Despite this, implementation of a screening program, as we

have proposed, will have a significant impact on the logisticsof MCH visits; additional staff and appropriate space will berequired for optimal implementation.

There remain many unanswered questions about TBscreening in the antenatal setting that need to be addressed byoperations research. These include optimal linkage betweenMCH and TB care; adherence barriers and toxicity facingcoinfected pregnant mothers in receiving TB and HIVcotreatment; effects of early TB/ART initiation on maternaland neonatal outcomes and neonatal HIV transmission; timeto ART and TB initiation in coinfected pregnant women;feasibility and cost-effectiveness of Xpert MTB/RIF imple-mentation in antenatal clinics; sensitivity and specificityof Xpert MTB/RIF in pregnant women; the role of layworkers/peer educators in integrated screening programs.Pilot programs are urgently needed to evaluate the impactof integrating TB screening strategies into antenatal clinicsusing new diagnostic technologies in order to reduce theburden of TB and HIV in mothers and their children.

Acknowledgment

The authors acknowledge the Zambian Ministry of Healthfor consistent and high-level support of operations researchin the context of TB. All authors have no conflicting interests.

References

[1] J. Grange, M. Adhikari, Y. Ahmed et al., “Tuberculosis in asso-ciation with HIV/AIDS emerges as a major nonobstetric causeof maternal mortality in Sub-Saharan Africa,” InternationalJournal of Gynecology and Obstetrics, vol. 108, no. 3, pp. 181–183, 2010.

[2] Y. Ahmed, P. Mwaba, C. Chintu, J. M. Grange, A. Ustianowski,and A. Zumla, “A study of maternal mortality at the UniversityTeaching Hospital, Lusaka, Zambia: the emergence of tuber-culosis as a major non-obstetric cause of maternal death,”International Journal of Tuberculosis and Lung Disease, vol. 3,no. 8, pp. 675–680, 1999.

[3] T. Pillay, M. Khan, J. Moodley et al., “The increasing burden oftuberculosis in pregnant women, newborns and infants under6 months of age in Durban, KwaZulu-Natal,” South AfricanMedical Journal, vol. 91, no. 11, pp. 983–987, 2001.

[4] C. R. Gounder, N. I. Wada, C. Kensler et al., “Active tuber-culosis case-finding among pregnant women presenting toantenatal clinics in soweto, South Africa,” Journal of AcquiredImmune Deficiency Syndromes, vol. 57, no. 4, pp. e77–e84,2011.

[5] H. C. Lin and S. F. Chen, “Increased risk of low birthweightand small for gestational age infants among women withtuberculosis,” International Journal of Obstetrics and Gynaecol-ogy, vol. 117, no. 5, pp. 585–590, 2010.

[6] T. Pillay, A. W. Sturm, M. Khan et al., “Vertical transmissionof Mycobacterium tuberculosis in KwaZulu Natal: impact ofHIV-1 co-infection,” International Journal of Tuberculosis andLung Disease, vol. 8, no. 1, pp. 59–69, 2004.

[7] T. Bjerkedal, S. L. Bahna, and E. H. Lehmann, “Course andoutcome of pregnancy in women with pulmonary tuberculo-sis,” Scandinavian Journal of Respiratory Diseases, vol. 56, no.5, pp. 245–250, 1975.

Page 33: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Journal of Pregnancy 5

[8] F. Margono, J. Mroueh, A. Garely, D. White, A. Duerr, andH. L. Minkoff, “Resurgence of active tuberculosis amongpregnant women,” Obstetrics and Gynecology, vol. 83, no. 6,pp. 911–914, 1994.

[9] A. Gupta, R. Bhosale, A. Kinikar et al., “Maternal tuberculosis:a risk factor for mother-to-child transmission of humanimmunodeficiency virus,” Journal of Infectious Diseases, vol.203, no. 3, pp. 358–363, 2011.

[10] A. C. Hesseling, M. F. Cotton, T. Jennings et al., “High inci-dence of tuberculosis among HIV-infected infants: evidencefrom a South African population-based study highlights theneed for improved tuberculosis control strategies,” ClinicalInfectious Diseases, vol. 48, no. 1, pp. 108–114, 2009.

[11] S. N. Tripathy, “Tuberculosis and pregnancy,” InternationalJournal of Gynecology and Obstetrics, vol. 80, no. 3, pp. 247–253, 2003.

[12] R. Figueroa-Damian and J. L. Arredondo-Garcia, “Pregnancyand tuberculosis: influence of treatment on perinatal out-come,” American Journal of Perinatology, vol. 15, no. 5, pp.303–306, 1998.

[13] J. E. Golub, C. I. Mohan, G. W. Comstock, and R. E. Chaisson,“Active case finding of tuberculosis: historical perspective andfuture prospects,” International Journal of Tuberculosis andLung Disease, vol. 9, no. 11, pp. 1183–1203, 2005.

[14] C. N. Mnyani and J. A. McIntyre, “Tuberculosis in pregnancy,”International Journal of Obstetrics and Gynaecology, vol. 118,no. 2, pp. 226–231, 2011.

[15] “World Health Organization (WHO) Policy on collaborativeTB/HIV activities,” Tech. Rep. WHO/HTM/TB/2011, Stop TBDepartment and Department of HIV/AIDS, World HealthOrganisation, Geneva, Switzerland, 2011.

[16] A. Deluca, R. E. Chaisson, and N. A. Martinson, “Intensifiedcase finding for tuberculosis in prevention of mother-to-childtransmission programs: a simple and potentially vital additionfor maternal and child health,” Journal of Acquired ImmuneDeficiency Syndromes, vol. 50, no. 2, pp. 196–199, 2009.

[17] “World Health Organization (WHO) report 2009 Global tu-berculosis control—epidemiology, strategy, financing,” WHOReport WHO/HTM/TB/2009411, World Health Organiza-tion, Geneva, Switzerland, 2009.

[18] P. B. N. Kali, G. E. Gray, A. Violari, R. E. Chaisson, J.A. McIntyre, and N. A. Martinson, “Combining PMTCTwith active case finding for tuberculosis,” Journal of AcquiredImmune Deficiency Syndromes, vol. 42, no. 3, pp. 379–381,2006.

[19] C. C. Boehme, P. Nabeta, D. Hillemann et al., “Rapidmolecular detection of tuberculosis and rifampin resistance,”New England Journal of Medicine, vol. 363, no. 11, pp. 1005–1015, 2010.

[20] World Health Organization (WHO), “Rapid implementationof the Xpert MTB/RIF diagnostic test,” in Technical andOperational ”How-to” Practical Considerations, World HealthOrganization, Geneva, Switzerland, 2011.

[21] C. C. Boehme, M. P. Nicol, P. Nabeta et al., “Feasibility,diagnostic accuracy, and effectiveness of decentralised use ofthe Xpert MTB/RIF test for diagnosis of tuberculosis andmultidrug resistance: a multicentre implementation study,”The Lancet, vol. 377, no. 9776, pp. 1495–1505, 2011.

[22] G. Theron, J. Peter, R. van Zyl-Smit et al., “Evaluation ofthe Xpert MTB/RIF assay for the diagnosis of pulmonarytuberculosis in a high HIV prevalence setting,” AmericanJournal of Respiratory and Critical Care Medicine, vol. 184, no.1, pp. 132–140, 2011.

[23] W. P. Killam, B. C. Tambatamba, N. Chintu et al., “Antiretrovi-ral therapy in antenatal care to increase treatment initiation inHIV-infected pregnant women: a stepped-wedge evaluation,”AIDS, vol. 24, no. 1, pp. 85–91, 2010.

[24] O. Koole, S. Thai, K. E. Khun et al., “Evaluation of the 2007WHO guideline to improve the diagnosis of tuberculosis inambulatory HIV-positive adults,” PLoS ONE, vol. 6, no. 4,Article ID e18502, 2011.

[25] J. Nachega, J. Coetzee, T. Adendorff et al., “Tuberculosis activecase-finding in a mother-tochild HIV transmission preventionprogramme in Soweto, South Africa,” AIDS, vol. 17, no. 9, pp.1398–1400, 2003.

[26] R. Mutetwa, C. Boehme, M. Dimairo et al., “Diagnostic ac-curacy of commercial urinary lipoarabinomannan detectionin African tuberculosis suspects and patients,” InternationalJournal of Tuberculosis and Lung Disease, vol. 13, no. 10, pp.1253–1259, 2009.

[27] G. Geojith, S. Dhanasekaran, S. P. Chandran, and J. Kenneth,“Efficacy of loop mediated isothermal amplification (LAMP)assay for the laboratory identification of Mycobacteriumtuberculosis isolates in a resource limited setting,” Journal ofMicrobiological Methods, vol. 84, no. 1, pp. 71–73, 2011.

[28] H. Getahun, W. Kittikraisak, C. M. Heilig et al., “Developmentof a standardized screening rule for tuberculosis in peopleliving with HIV in resource-constrained settings: individualparticipant data meta-analysis of observational studies,” PLoSMedicine, vol. 8, no. 1, Article ID e1000391, 2011.

[29] M. B. Morris, B. T. Chapula, B. H. Chi et al., “Use of task-shifting to rapidly scale-up HIV treatment services: experi-ences from Lusaka, Zambia,” BMC Health Services Research,vol. 9, article no. 5, 2009.

[30] P. Sanjana, K. Torpey, A. Schwarzwalder et al., “Task-shiftingHIV counselling and testing services in Zambia: the role of laycounsellors,” Human Resources for Health, vol. 7, article no. 44,2009.

Page 34: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Hindawi Publishing CorporationJournal of PregnancyVolume 2012, Article ID 958262, 8 pagesdoi:10.1155/2012/958262

Review Article

Understanding Methods for Estimating HIV-AssociatedMaternal Mortality

James E. Rosen,1 Isabelle de Zoysa,2 Karl Dehne,3 Viviana Mangiaterra,4

and Quarraisha Abdool-Karim5

1 Futures Institute, New London Turnpike, Glastonbury, CT 06033, USA2 Lungarno Guicciardini, 17, 50125 Firenze, Italy3 UNAIDS, Avenue Appia, Geneva 1211, Switzerland4 World Health Organization, Avenue Appia, Geneva 1211, Switzerland5 CAPRISA, Nelson R Mandela School of Medicine, University of Kwazulu-Natal, Congella 4013, South Africa

Correspondence should be addressed to Isabelle de Zoysa, [email protected]

Received 13 May 2011; Accepted 7 July 2011

Academic Editor: Oliver Ezechi

Copyright © 2012 James E. Rosen et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The impact of HIV on maternal mortality and more broadly on the health of women, remains poorly documented and understood.Two recent reports attempt to address the conceptual and methodological challenges that arise in estimating HIV-related maternalmortality and trends. This paper presents and compares the methods and discusses how they affect estimates at global and regionallevels. Country examples of likely patterns of mortality among women of reproductive age are provided to illustrate the criticalinteractions between HIV and complications of pregnancy in high-HIV-burden countries. The implications for collaborationbetween HIV and reproductive health programmes are discussed, in support of accelerated action to reach the MillenniumDevelopment Goals and improve the health of women.

1. Introduction

While recent reports indicate declining trends in maternalmortality [1, 2], at the current rate of progress, most coun-tries remain unlikely to reach the Millennium DevelopmentGoal 5 and its target of reducing the maternal mortality ratio(MMR) by 75% between 1990 and 2015. The adverse effectof HIV on women’s health in sub-Saharan Africa appears tobe an important reason for poor progress [1, 3].

The contribution of HIV to maternal mortality has beenrecognized for over a decade [4, 5], but it remains poorlydocumented and understood. The number and proportionof maternal deaths associated with HIV have been difficult todetermine with precision because of various conceptual andmeasurement challenges [1, 2]. This information is criticalto plan services for women in need, including HIV-infectedpregnant women.

This paper aims to enhance the understanding of themethods used to estimate HIV-associated maternal deathsand how they affect global, regional, and country estimates.

Country examples of likely patterns of mortality amongwomen of reproductive age are provided to illustrate thecritical interactions between HIV and maternal mortality inhigh-HIV-burden countries.

2. Methods for Estimating HIV-AssociatedMaternal Deaths

2.1. Definitions. In the International Statistical Classificationof Diseases and Related Health Problems, Tenth Revision,1992 (ICD-10), WHO defines maternal death as the death ofa woman while pregnant or within 42 days of terminationof pregnancy, irrespective of the duration and site of thepregnancy, from any cause related to or aggravated bythe pregnancy or its management but not from accidentalor incidental causes [6]. Maternal deaths are either directobstetric deaths resulting from obstetric complications ofthe pregnant state (pregnancy, delivery, and postpartum),from interventions, omissions, incorrect treatment, or from

Page 35: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

2 Journal of Pregnancy

a chain of events resulting from any of the above, orindirect deaths resulting from previous existing disease ordisease that developed during pregnancy, such as anaemiaand malaria, which was aggravated by physiological effectsof pregnancy. In some instances, deaths during pregnancyor the postpartum period might be due to causes that areincidental to pregnancy, such as accidents and violence.These are defined in ICD as pregnancy-related deaths, but arenot included as part of maternal deaths. Pregnancy-relateddeaths comprise maternal deaths due to direct and indirectcauses and also deaths due to accidental or incidental eventsthat occur when a woman is pregnant.

Some of the deaths of pregnant women with HIV areincidental deaths due to an AIDS-defining condition, with-out any obvious association with pregnancy. These deathsare therefore classified as pregnancy-related deaths, but notmaternal deaths. In practice, the distinction between in-cidental and indirect causes of death is difficult to make,especially in the case of HIV. The effect of pregnancy on HIVdisease progression is uncertain [7–9], but HIV would notlikely be an indirect cause of death except in advanced stagesof the disease. It is also not clear to what extent HIV infectionhas an effect on the maternal outcomes of pregnancy [7, 8].However, there is some evidence that HIV can aggravateobstetric conditions that can lead to death, such as sepsis,haemorrhage, and septic abortion [9, 10].

Figure 1 is a graphical representation of the intersectionof HIV deaths and maternal deaths among all deaths ofwomen of reproductive age (15–49 years). In this paper, weuse the term “HIV-associated maternal death” to describethe death of a pregnant woman with HIV in which the HIVinfection was present at the time of a death from directobstetric causes or else was an indirect cause of maternaldeath. The figure shows that HIV-associated maternal deathsare a subset of what we call “pregnancy-related HIV deaths,”which include all deaths in HIV-positive women that takeplace during pregnancy, childbirth, and 42 days postpartum.These include deaths from HIV disease that were incidentalto the pregnancy. This figure does not attempt to show therelative distribution of deaths in the various categories. Fororders of magnitude, it is useful to recall here that HIV/AIDSand complications of pregnancy and childbirth are the twoleading causes of death in young adult women globally,accounting for 19% and 15% of all deaths in women aged15–44 years, respectively [11].

2.2. Measurement Challenges. In the real world, identifyingmaternal deaths and determining their causes, in orderto differentiate deaths due to direct or indirect causes, orto incidental causes, poses many challenges. Most devel-oping countries do not have civil registration systems torecord the death of a woman of reproductive age, andidentification of maternal deaths usually requires specialmethods of investigation, such as household surveys. Inaddition, there are well-documented problems with regard tothe identification of cause of death in general and maternaldeath in particular, which are exacerbated in the contextof HIV [12–14]. Ideally, HIV-associated maternal deaths

should be counted using the ICD-10 classification system,along with other causes of maternal deaths—direct andindirect. In practice, this is challenging, due to insufficientor inaccurate information on the circumstances of death andlimited capacities to correctly use standards to assign andcode causes of death. In particular, information about thewoman’s pregnancy status or HIV status is often missing orwithheld. Even when deaths are medically certified, applyingthe ICD rules when deaths are associated with multiplecomorbidities is not straightforward [2, 10]. With ICD-10, WHO recommends the inclusion of a checkbox on thedeath certificate for recording a woman’s pregnancy statusat the time of death, but this has not been implemented inmany countries, and therefore some AIDS deaths in preg-nant women are not identified as pregnancy-related. Theoverall result is that most countries report incomplete andinaccurate data on numbers and causes of maternal andpregnancy-related deaths, and the precise contribution ofHIV is poorly documented.

2.3. Methods of Estimating HIV-Associated Maternal Deaths.In the run-up to the 2015 deadline to achieve the MillenniumDevelopment Goals, increased efforts are being made to trackprogress in reducing maternal mortality. In 2010, two majorreports provided estimates of maternal mortality trends atglobal, regional, and country levels [1, 2]. Maternal mortalityestimates were based on data collected through a varietyof methods, including civil registration systems, householdsurveys, special studies, censuses, and verbal autopsies [2].Different statistical modeling approaches were used to es-timate the contribution of HIV to maternal mortality.

Hogan and colleagues, at the Institute for Health Metricsand Evaluation (IHME) and other academic institutions,estimated levels and trends in maternal mortality for 181countries between 1980 and 2008 and assessed the impact ofHIV on these estimates. Details on the IHME methodologicalapproach can be found in the published report [1], webannex[15], and model output data published on the IHME website(http://www.who.int/gho/maternal health/mortality/mater-nal/en/index.html). In summary, IHME’s multivariate modelincluded HIV prevalence as one of several covariates toestimate total maternal deaths, including what the authorscall “HIV-related maternal deaths.” To distinguish the HIV-related deaths among the total number of maternal deaths,IHME conducted a counterfactual analysis that explored theeffect on maternal mortality of reducing HIV prevalence tozero, by setting the HIV covariates of the estimated modelto zero values. The difference between the two outputsproduced the number of HIV-related maternal deaths. Theprimary source of data used in many countries was surveys orcensuses, which do not exclude incidental deaths in pregnantwomen. Therefore, the numbers reported by IHME in thosecountries relate to our definition of pregnancy-related HIVdeath, including both HIV-associated maternal deaths andHIV deaths that are incidental to the pregnancy. Hoganand colleagues indicate that this approach would tend tobias the maternal mortality estimates upward, but note that

Page 36: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Journal of Pregnancy 3

HIV/AIDS deaths

All of reproductive age

Pregnancy-related deaths

Maternal deaths

HIV-associatedmaternal death

Incidental pregnancy-related HIV deaths

Non-pregnancy-related HIV deaths

deaths of women

Figure 1: The intersecting epidemics of HIV and maternal mortality.

there is no evidence-based analytical strategy to identify theproportion of pregnancy-related deaths that are incidental.

The UN-affiliated Maternal Mortality Estimation Inter-Agency Group (MMEIG), together with academics from theUniversity of California at Berkeley, developed estimates ofindirect maternal deaths due to HIV as part of a largereffort to estimate country, regional, and global maternalmortality levels and trends between 1990 and 2008 [2, 16].The interagency group also took available data on totaldeaths in women of reproductive age as the starting point,but applied different statistical models and assumptions togenerate estimates of maternal deaths. In the first instance,only maternal, non-HIV-related deaths were computed, bysubtracting the fraction of estimated HIV deaths in thepregnancy-related period (which corresponds to our defi-nition of pregnancy-related HIV deaths). Then the numberof indirect maternal deaths due to HIV was computed andadded back to obtain the total number of estimated maternaldeaths.

The steps in this process were as follows (see also [2], datafiles available on the WHO web site http://www.who.int/gho/maternal health/mortality/maternal/en/index.html). First, astatistical model estimated the fraction of HIV deaths inwomen of reproductive age occurring during pregnancy orwithin 42 days of termination of pregnancy. The calculationsrequired assumptions about the relative risk of dying fromHIV of a pregnant versus a nonpregnant woman (reflectingboth the decreased fertility of HIV-positive women and the

increased mortality risk of HIV-positive pregnant women).This latter parameter was developed through a combinationof expert opinion and tests of model fitness to reach a valueof 0.4, which was applied to all countries. These deathswere removed from the computation of maternal deathsdue to direct obstetric causes or to indirect causes otherthan HIV. Then a constant fraction of these deaths, countedas indirect maternal deaths, was added back to calculatetotal maternal deaths. For the estimation of these HIV-related indirect maternal deaths, it was assumed that halfof the total pregnancy-related deaths among HIV-infectedwomen were indirect maternal deaths, the other half beingincidental to the pregnancy. This value was chosen tominimize distribution errors given the scarcity of appropriatedata for estimating correct proportion of HIV deaths that areincidental (which can vary over time and place).

3. Results Based on the Different Methods

3.1. Worldwide HIV-Associated Maternal Deaths. The inter-agency group estimated 21,000 HIV-associated maternaldeaths (uncertainty interval 15,341–29,120), or 6% of theworldwide total of 358,000 maternal deaths in 2008. Forthe same year, the IHME calculated that 61,400 fewermaternal deaths (18%) would have occurred in the absenceof HIV (uncertainty interval 58,200–66,400), of an estimatedtotal of 342,900 maternal deaths. As noted above, the

Page 37: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

4 Journal of Pregnancy

interagency model computed pregnancy-related HIV deathsand assumed that half of those were indirect maternaldeaths. Direct maternal deaths in which HIV was present,and possibly a contributing factor, were not counted. Theother model computed primarily pregnancy-related HIVdeaths. Should the IHME model have used the interagencyassumption about the proportion of pregnancy-related HIVdeaths that are truly maternal, their estimated numbers ofHIV-associated maternal death and total number of maternaldeaths would have been reduced accordingly.

3.2. Regional Variation in HIV-Associated Maternal Deaths.Both models show that the vast majority of HIV-associatedmaternal deaths occur in sub-Saharan Africa (86% in theinteragency model and 92% in the IHME model) (Table 1).As a fraction of all maternal deaths, HIV-associated maternaldeaths are also highest in the sub-Saharan region (9% underthe interagency model and 32% in the IHME model). Inother regions, HIV-associated maternal deaths account for6% or less of all maternal deaths.

3.3. HIV-Associated Maternal Deaths by Country. The rank-ing of countries with the highest number of HIV-relatedmaternal deaths varies substantially by model (Table 2).Nigeria is consistently ranked first, although the estimatednumber of deaths per year ranges widely, between 2,472for the interagency model and 10,422 for the IHME model.Six countries make the top ten under both model es-timates—Kenya, Malawi, Mozambique, Nigeria, Tanzania,and Uganda. However, their ranking differs under eachmodel. Only one country outside sub-Saharan Africa, India,is in the top ten.

The countries with the highest fraction of HIV-relatedmaternal deaths are shown in Table 3. The list is dominatedby countries in Southern Africa that have been hardest hitby the HIV epidemic. In many of these countries, HIV-as-sociated maternal deaths represent a substantial proportionof total maternal deaths. Both the interagency and IHMEapproaches estimate the fraction at greater than 70% forBotswana and Swaziland and over one-third for all countriesin the top ten.

3.4. Relative Impact of HIV on Pregnancy-Related and Ma-ternal Mortality at Country Level. Figure 2 shows the sub-stantial impact HIV has on the health of women in Zambia,a high-prevalence country. Based on the interagency groupestimates, 21,532 or two-thirds of the 32,510 deaths ofwomen of reproductive age in 2008 can be attributed toHIV. Of the total of 3,673 pregnancy-related deaths, 2,551are maternal deaths. Of maternal deaths, 943 (37%) areestimated to be HIV associated. Another 943 women are esti-mated to die from incidental HIV deaths during pregnancy.

Figure 3 shows the pattern in the Democratic Republicof Congo. In that country, a smaller proportion of deathsamong women of reproductive age are reported to be as-sociated with HIV, and a larger proportion of deaths aredue to complications of pregnancy and childbirth. The op-posite would seem to be the case in Zimbabwe, where the

overwhelming majority of deaths among women of repro-ductive age are associated with HIV, as shown in Figure 4.Similar charts can be produced to support more detailedassessments of the situation in other affected countries.

4. Comparison of Methods and Results

We examined some of the challenges in defining and count-ing HIV-associated maternal deaths and reviewed two recentmethodological approaches for their estimation. (This paperdoes not attempt to explain why the interagency and IHMEestimates for maternal deaths might differ. For further dis-cussion, see Section 3.6 of [2].) Both approaches address thewell-known conceptual and practical difficulties in assessinga pregnancy-related death in the presence of HIV. Theyuse statistical models to estimate the impact of HIV onlevels and trends of maternal mortality, combining robustdata sources with assumptions regarding key parametersabout which little information exists. The interagency groupestimates of indirect maternal deaths due to HIV rest onassumptions about the relative risk of HIV mortality inpregnant versus nonpregnant women and the proportionof pregnancy-related HIV mortality that can be consideredmaternal. IHME uses a counterfactual method that comparesresults with and without the HIV epidemic.

There are major differences in the way each model iden-tifies pregnancy-related HIV deaths and examines the asso-ciation of HIV with maternal deaths. The IHME modelproduces somewhat higher results, with the number of HIV-associated maternal deaths (61,400) about three times higherthan the interagency estimates. In part because the IHMEmodel has a lower overall number of maternal deaths, thefraction of maternal deaths associated with HIV is sub-stantially higher when compared to the interagency model.Regional estimates of both numbers and affected fractionfollow a similar pattern between models. At the country level,the models produce similar rankings in terms of the mostaffected countries, although the estimated numbers and af-fected fractions vary.

These comparisons should be made with caution, partlybecause of the different model approaches and in how theresults are presented. The interagency estimate distinguishesbetween incidental pregnancy-related HIV deaths and indi-rect HIV-related maternal deaths, to keep strictly to theICD-10 definition of maternal death, without being able tofully delineate the differences between the two. The IHMEmodel notes that the evidence to identify the proportion ofpregnancy-related deaths that are incidental is scant, and itsestimates are based on total numbers of pregnancy-relateddeaths in which HIV was present, while recognizing that thisstrategy biases the estimates upwards.

Despite these differences, estimates using the differentmethods clearly reinforce that the HIV epidemic is having aprofound influence on maternal mortality and, more broad-ly, pregnancy-related mortality. Worldwide estimates of thefraction of maternal deaths due to HIV range between 6and 18%, with the majority of HIV-associated maternal

Page 38: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Journal of Pregnancy 5

Table 1: HIV-associated maternal deaths by region, 2008, by model.

MDG regionNumber of HIV-associated maternal deaths Proportion of all maternal deaths

Interagency IHME Interagency IHME

World total 21,000 61,436 5.8% 17.9%

Developed regions 90 45 5.6% 3.4%

CIS countries 70 35 4.7% 2.6%

Developing regions 21,000 61,356 5.8% 18.0%

Africa 18,000 56,446 8.9% 31.5%

Northern Africa 10 28 0.3% 1.2%

Sub-Saharan Africa 18,000 56,418 9.0% 31.9%

Asia 1,700 4,358 1.2% 2.9%

Eastern Asia 80 42 1.0% 0.5%

South Asia 1,300 3,632 1.2% 3.0%

South-Eastern Asia 310 682 1.7% 3.9%

Western Asia — 2 0.0% 0.0%

Latin America and the Caribbean 480 510 5.2% 6.4%

Oceania 10 42 1.1% 5.8%

Table 2: 10 Countries with the highest number of HIV-associated maternal deaths, 2008, by model.

Interagency IHME

Ranking Country #HIV-associated maternal deaths Country #HIV-associated maternal deaths

1 Nigeria 2472 Nigeria 10422

2 South Africa 1920 Malawi 4689

3 Zimbabwe 1574 Ethiopia 3971

4 Tanzania 1552 Tanzania 3941

5 Uganda 1512 India 3531

6 India 1264 Mozambique 3448

7 Mozambique 1217 Kenya 3006

8 Kenya 1100 Cote d’Ivoire 2871

9 Malawi 961 Uganda 2611

10 Ethiopia 948 Zambia 2403

Table 3: 10 Countries with the highest fraction of HIV-associated maternal deaths, 2008, by model.

Interagency IHME

Ranking CountryFraction of HIV-associated

maternal deathsCountry

Fraction of HIV-associatedmaternal deaths

1 Botswana 78% Botswana 84%

2 Swaziland 75% Swaziland 84%

3 Lesotho 59% Lesotho 83%

4 Zimbabwe 53% Zimbabwe 82%

5 Namibia 50% South Africa 78%

6 South Africa 43% Zambia 73%

7 Zambia 37% Namibia 73%

8 Belize 35% Malawi 69%

9 Malawi 32% Mozambique 66%

10 Trinidad and Tobago 28% Uganda 51%

Page 39: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

6 Journal of Pregnancy

Pregnancy-related deaths

(3, 673)

Maternal deaths (2, 551)

HIV-associated maternal

ofreproductive age

Pregnancy-related deaths

Maternal deaths

Zambia 2008

Incidental pregnancy-related HIV deaths (943)

deaths (943) HIV-associatedmaternal

deaths

Incidental pregnancy-related HIV deaths

Deaths of women of reproductive age

HIV deaths of women ofreproductive age (21, 532)

women of reproductiveage (32, 510)

HIV deaths of women

Outer square: deaths of

Figure 2: Intersection of HIV and maternal mortality in Zambia, 2008. Source: http://www.who.int/gho/maternal health/mortality/maternal/en/index.html.

reproductive age (11, 073)

Pregnancy-related deaths(21, 800)

Maternal deaths (19, 216)

HIV-associated maternaldeaths (506)

Incidental pregnancy-related HIV deaths (506)

women of reproductiveage (97, 511)

Pregnancy-related deaths

Maternal deaths

D CR 2008

HIV-associatedmaternal

deaths

Incidental pregnancy-related HIV deaths

Deaths of women of reproductive age

ofreproductive ageHIV deaths of women

HIV deaths of women of

Outer square: deaths of

Figure 3: Intersection of HIV and maternal mortality in DRC, 2008. Source: http://www.who.int/gho/maternal health/mortality/maternal/en/index.html.

Page 40: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Journal of Pregnancy 7

HIV-associated maternal

Pregnancy-related deaths Maternal deaths

HIV-associatedmaternal

deaths

Incidental pregnancy-related HIV deaths

deaths (1, 574)

Incidental pregnancy-related HIV deaths (1, 574)

Maternal deaths (2, 984)

Pregnancy-related deaths(4, 714)

women of reproductiveage (68, 604)

reproductive age (6 , 4 )1 59

Zimbabwe 2008

HIV deaths of women of

Deaths of women of reproductive age

ofreproductive ageHIV deaths of women

Outer square: deaths of

Figure 4: Intersection of HIV and maternal mortality in Zimbabwe, 2008. Source: http://www.who.int/gho/maternal health/mortality/maternal/en/index.html.

deaths in the sub-Saharan region, where the affected fractionis between 9–32%. Country rankings between models aresimilar and provide a general but not precise estimate of theHIV epidemic’s contribution to maternal mortality.

More detailed examination at country level of the impactof HIV on maternal and pregnancy-related deaths showsdiverse patterns. National estimates are usually based on fewdata points and are subject to all the caveats mentionedearlier about the identification of pregnancy-related deathsand the classification of cause of death. Nonetheless, they givean idea of the severe adverse effect of the HIV epidemic onwomen’s health in high-burden countries.

5. Conclusions

Despite methodological challenges, assessments of trends inmaternal mortality clearly indicate that HIV has become aleading cause of death during pregnancy and the postpartumperiod in countries with high HIV prevalence and thatthe HIV epidemic has slowed down progress in improvingmaternal health.

Improved knowledge of the contribution of HIV to ma-ternal and pregnancy-related mortality should help to directscarce resources to appropriate policy and programmaticresponses and spur better collaboration between HIV andreproductive health services [1, 17]. We hope that ourexamination of different approaches used to assess the inter-actions between HIV and maternal mortality will stimulate

improvements in systems to document maternal deaths andassess their causes.

It would be useful to distinguish between maternal HIV-associated death due to direct obstetric causes or indirectHIV-related causes and incidental pregnancy-related HIVdeaths. The new ICD-10 amendment introducing a newcode for an “indirect maternal death” is a step forward.However, this is likely to continue to remain a sourceof misclassification, given the ongoing clinical uncertaintyaround the relationship between HIV and other causes ofmaternal death. Further discussion among experts in ICD-compliant death certification and coding is needed to helpclarify how best such interrelated causes of deaths shouldbe classified. Adoption of a common terminology amongresearchers and health service providers would be a furtherbenefit.

There is also an urgent need to carry out systematicreviews of existing studies on the key parameters that un-derpin the estimates. These include studies on the relativerisk of maternal death in HIV-infected versus uninfectedpregnant women and the relative risk of dying from HIVdisease of a pregnant versus a nonpregnant woman. Morestudies are also needed to elucidate the mechanisms by whichHIV can contribute to direct or indirect maternal deaths.

Nonetheless, given the close, almost inextricable inter-actions between HIV and other causes of death duringand around pregnancy and inadequate systems for civilregistration in most countries, it will remain difficult toquantify the precise impact of HIV on maternal mortality

Page 41: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

8 Journal of Pregnancy

in the foreseeable future. This should not serve as an excusefor inaction. Urgent measures are required to provide themany girls and young women who face the dual risks ofHIV infection and pregnancy with the services that theyneed. In settings with a high HIV burden and continuinghigh maternal mortality ratios, especially in sub-SaharanAfrica, this means scaling up comprehensive and integratedprogrammes that include both improved HIV treatment andcare and improved reproductive health services [9].

Acknowledgment

The authors thank Carla Abou-Zahr, Eleanor Gouws, PeterGhys, and Lale Say for their careful review and helpful adviceand Terhi Aaltonen for her thorough editorial support.

References

[1] M. C. Hogan, K. J. Foreman, M. Naghavi et al., “Maternal mor-tality for 181 countries, 1980–2008: a systematic analysis ofprogress towards Millennium Development Goal 5,” The Lan-cet, vol. 375, no. 9726, pp. 1609–1623, 2010.

[2] WHO, UNICEF, UNFPA, World Bank, Trends in MaternalMortality: 1990 to 2008. Estimates Developed by WHO,UNICEF, UNFPA and the World Bank, World Health Organi-zation, Geneva, Switzerland, 2010.

[3] J. Wilmoth, C. Mathers, L. Say, and S. Mills, “Maternal deathsdrop by one-third from 1990 to 2008: a United Nationsanalysis,” Bulletin of the World Health Organization, vol. 88,no. 10, pp. 718–718A, 2010.

[4] W. J. Graham and M.-L. Newell, “Seizing the opportunity: col-laborative initiatives to reduce HIV and maternal mortality,”The Lancet, vol. 353, no. 9155, pp. 836–839, 1999.

[5] J. McIntyre, HIV in Pregnancy: A Review, World HealthOrganization, Geneva, Switzerland, 1998.

[6] World Health Organization, Statistical Classification of Dis-eases and Related Problems. 10th revision, World Health Organ-ization, Geneva, Switzerland, 1992.

[7] J. McIntyre, “Maternal health and HIV,” Reproductive HealthMatters, vol. 13, no. 25, pp. 129–135, 2005.

[8] R. French and P. Brocklehurst, “The effect of pregnancy onsurvival in women infected with HIV: a systematic review ofthe literature and meta-analysis,” BJOG, vol. 105, no. 8, pp.827–835, 1998.

[9] J. Moodley, R. C. Pattinson, C. Baxter, S. Sibeko, and Q.Abdool Karim, “Strengthening HIV services for pregnantwomen: an opportunity to reduce maternal mortality rates inSouthern Africa/sub-Saharan Africa,” BJOG, vol. 118, no. 2,pp. 219–225, 2011.

[10] World Health Organization, The WHO Classification of DeathsDuring Pregnancy, Childbirth and the Puerperium. Applicationof the ICD to Maternal Mortality: ICD-MM, Childbirth andPuerperium.

[11] World Health Organization, Women and Health: Today’sEvidence, Tomorrow’s Agenda, World Health Organization, Ge-neva, Switzerland, 2009.

[12] C. Ronsmans and W. J. Graham, “Maternal mortality: who,when, where, and why,” The Lancet, vol. 368, no. 9542, pp.1189–1200, 2006.

[13] S. Cross, J. S. Bell, and W. J. Graham, “What you count is whatyou target: the implications of maternal death classificationfor tracking progress towards reducing maternal mortality in

developing countries,” Bulletin of the World Health Organiza-tion, vol. 88, no. 2, pp. 147–153, 2010.

[14] K. S. Khan, D. Wojdyla, L. Say, A. M. Gulmezoglu, and P. F. A.van Look, “WHO analysis of causes of maternal death: a sys-tematic review,” The Lancet, vol. 367, no. 9516, pp. 1066–1074,2006.

[15] M. C. Hogan, K. J. Foreman, M. Naghavi et al., “Supplemen-tary webappendix. Supplement to: maternal mortality for 181countries, 1980–2008: a systematic analysis of progress to-wards Millennium Development Goal 5,” The Lancet, vol. 375,no. 9726, pp. 1609–1623, 2010.

[16] J. Wilmoth, S. Zuerick, N. Mizoguchi, M. Inoue, and M.Oestergaard, Levels and Trends of Maternal Mortality in theWorld: The Development of the New Estimates by the UnitedNations, WHO, UNICEF, UNFPA, and the World Bank, 2010.

[17] Q. Abdool-Karim, C. AbouZahr, K. Dehne et al., “HIV andmaternal mortality: turning the tide,” The Lancet, vol. 375, no.9730, pp. 1948–1949, 2010.

Page 42: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Hindawi Publishing CorporationJournal of PregnancyVolume 2012, Article ID 841979, 4 pagesdoi:10.1155/2012/841979

Research Article

Risk Factors Associated with False Positive HIVTest Results in a Low-Risk Urban Obstetric Population

Tamara T. Chao,1 Jeanne S. Sheffield,1 George D. Wendel Jr.,1

M. Qasim Ansari,2 Donald D. McIntire,1 and Scott W. Roberts1

1 Department of Obstetrics and Gynecology, The University of Texas Southwestern Medical Center, Dallas,TX 75390-9032, USA

2 Department of Pathology, The University of Texas Southwestern Medical Center, Dallas,TX 75390-9032, USA

Correspondence should be addressed to Tamara T. Chao, [email protected]

Received 21 April 2011; Accepted 22 June 2011

Academic Editor: Karen Odberg-Petterson

Copyright © 2012 Tamara T. Chao et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Objective. To examine risk factors for false positive HIV enzyme immunoassay (EIA) testing at delivery. Study Design. A review ofpregnant women who delivered at Parkland Hospital between 2005 and 2008 was performed. Patients routinely received serumHIV EIA testing at delivery, with positive results confirmed through immunofluorescent testing. Demographics, HIV, hepatitis Bsurface antigen (HBsAg), and rapid plasma reagin (RPR) results were obtained. Statistical analyses included Pearson’s chi-squareand Student’s t-test. Results. Of 47,794 patients, 47,391 (99%) tested negative, 145 (0.3%) falsely positive, 172 (0.4%) positive, and86 (0.2%) equivocal or missing HIV results. The positive predictive value of EIA was 54.3%. Patients with false positive results weremore likely nulliparous (43% versus 31%, P < 0.001) and younger (23.9 ± 5.7 versus 26.2 ± 5.9 years, P < 0.001). HIV positivepatients were older than false positive patients and more likely positive for HBsAg and RPR. Conclusion. False positive HIV testingat delivery using EIA is associated with young maternal age and nulliparity in this population.

1. Introduction

The Centers for Disease Control and Prevention (CDC)and United States Preventive Services Task Force (USPSTF)recommend universal HIV screening for all pregnant womenentering prenatal care [1, 2]. This screening enables HIV-infected women and their infants to benefit from appropriateand timely interventions such as antiretroviral medications.When the recommended antiretroviral and obstetric inter-ventions are used, a woman who knows of her HIV infectionearly in pregnancy now has a less than 2% chance ofdelivering an HIV-infected infant. Without intervention, thisrisk is approximately 25% in the United States [3–6].

Testing for HIV began in 1985 with the introductionof the enzyme immunoassay (EIA). In order to account forfalse positive results using screening tests in a low-prevalencepopulation, confirmatory testing has been implemented

using a Western blot or immunofluorescence assay. In a low-prevalence population, the false positive rate using the EIAis increased compared to a high-prevalence population, andthe positive predictive value of any test will always depend onthe prevalence of the condition in the population tested. Intesting performed by the CDC, the EIA positive predictivevalue ranges from 71 to 83% in populations with HIVprevalence from 0.5 to 1% [7].

Pregnancy has been observed to be associated withfalse positive HIV testing. Some investigators believe thatthe presence of alloantibodies account for the increasedfalse positive rate associated with pregnancy, transfusions,transplantation, and autoimmune diseases [8]. However,risk factors specific to pregnancy that account for this arepoorly understood. Conversely, a recent large retrospectivestudy found that the false positive HIV EIA rate was lowerin pregnant women compared to nonpregnant individuals

Page 43: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

2 Journal of Pregnancy

(0.14% versus 0.21%) [9]. Our objective was to determineif any maternal characteristics correlated with false positiveHIV EIA testing at delivery.

2. Materials and Methods

This was a review of all women who delivered at ParklandMemorial Hospital in Dallas, Tex, from October 1, 2005through September 30, 2008. All women routinely receivedserum HIV testing at their initial prenatal visit and at time ofpresentation to labor and delivery for delivery via the opt-out approach with the Abbott Commander HIV AB HIV-1/HIV-2 (rDNA) EIA (Abbott Laboratories, Abbott Park,Ill). HIV test results performed at the time of deliverywere analyzed in this study. A woman was considered HIVnegative if EIA testing was negative. Positive test results wereconfirmed with the fluorognost immunofluorescent assay(IFA) HIV-1 (Sanochemia Corp, Stamford, Conn, USA).Women were considered to be falsely positive if EIA resultswere positive and the IFA was negative. Women deliveredin this time period were identified through the obstetricoperations database and linked to the pathology database forHIV, hepatitis B surface antigen (HBsAg), and rapid plasmareagin (RPR) results. Maternal characteristics, includingrace, parity, age, singleton versus multiple gestation, and adiagnoses of diabetes or hypertension were obtained usingthe obstetrics operations database. Laboratory results drawn28 days prior to delivery through seven days after deliverywere included.

The study was approved by the Institutional ReviewBoard at the University of Texas Southwestern MedicalCenter. Categorical data were reported as frequencies, andstatistical significance was determined using χ2 analysis.Two-group comparisons were made using Student’s t-test. Pvalues less than 0.05 were considered statistically significant.Statistical analyses were performed using SAS, Version 9.2(SAS Institute, Cary, NC).

3. Results

A total of 47,794 women were identified who deliveredduring the study time frame. Demographic and obstetricalcharacteristics of the patient population are shown inTable 1. Compared to HIV negative patients, false positivepatients were more likely to be nulliparous (43% versus 31%,P < 0.001) and younger (mean age 23.9±5.7 versus 26.2±5.9,P < 0.001). HIV positive patients were significantly olderthan false positive patients (27.4 versus 23.9, P < 0.001) andHIV negative patients (27.4 versus 26.2, P = 0.012).

Of the 47,794 women, 47,391 (99%) were HIV negative,145 (0.3%) had a false positive test, 172 (0.4%) weretrue positives, and 86 (0.2%) tested equivocal or weremissing HIV results. The specificity of the HIV EIA testwas 99.7% with a positive predictive value of 54.3%. Thesensitivity of the EIA test was presumed to be near 100%,as the false negative rate using the EIA has been previouslydemonstrated to be negligible in studies performed by themanufacturer [10].

Table 2 shows the prevalence of diabetes, hypertension,hepatitis B, and results of syphilis testing in the studypopulation. HIV positive women were more likely to testpositive for hepatitis B surface antigen (1% versus 0%, P =0.002) and RPR (2% versus 0%, P = 0.02). There wasno significant difference in the prevalence of diabetes orhypertension between the three groups. There was also nosignificant difference in the rate of HBsAg and RPR positivitybetween the HIV negative and false positive groups.

When evaluating the interaction between nulliparityand age, there is a significant correlation between the twovariables; that is, parous patients are likely to be older and,therefore, nulliparity and age are not independent predictorsof HIV false positivity. However, when the HIV testinggroups (positive, false positive, negative) are stratified byparity, the comparison of age across the three groups remainssignificant only in nulliparous women (Table 3, P = 0.0003).The interaction between parity and age means that thedifference in age between nulliparous and parous women isdifferent depending on the HIV status. For example, in theHIV positive group the nulliparous and parous women arecloser in age than in the false positive group.

4. Comment

This was the first population-based study to evaluate riskfactors for HIV false positivity in pregnant women pre-senting for delivery at a large urban institution. We foundthat younger and nulliparous women were more likely tohave false positive testing using the HIV EIA. The observedpositive predictive value (PPV) of 54.3% was lower thanthe previously reported by the CDC (PPV 71–83%) in anonpregnant population [7], suggesting that pregnancy maybe associated with a higher rate of false positivity. The lowpositive predictive value of the HIV EIA in our study mayhave been impacted by the relatively low HIV prevalence(0.4%) in our population. This information could be usefulfor counseling women who test positive for HIV at deliveryand emphasizes the limitations of EIA testing if used asa rapid test to determine the need for intrapartum andneonatal antiretroviral prophylaxis.

The false positive rates of currently available rapid HIVtests have been reported to be much lower than has beenfound with EIA testing in this study. Tung et al. evaluated theHIV false positive rate in over 900 pregnant women, most ofwhom were Hispanic, using both the EIA and a rapid point-of-care (POCT) test, OraQuick [11]. They found that whilethere were no false positive tests with the OraQuick, therewere seven false positives using the EIA (PPV 100% versus35.7%). In the Mother-Infant Rapid Intervention at Delivery(MIRIAD) study, Jamieson et al. found that the PPV of theOraQuick test was 90% while the PPV of the EIA was 74%[12, 13].

Current recommendations by the CDC and the Ameri-can College of Obstetricians and Gynecologists include rapidHIV screening for women presenting to labor and deliverywith undocumented HIV status and for repeat HIV testing inthe third trimester for women at high risk or who live in areas

Page 44: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

Journal of Pregnancy 3

Table 1: Demographic and obstetric characteristics of HIV positive, false positive, and negative women.

CharacteristicPositiveN = 172

False positiveN = 145

NegativeN = 47391

P

Positive versusfalse positive

Positive versusnegative

False positiveversus negative

Age: mean ± std 27.4 ± 6.2 23.9 ± 5.7 26.2 ± 5.9 <0.001 0.012 <0.001

Age <0.001 0.021 0.002

<18 2 (1) 16 (11) 2494 (5)

18–35 152 (88) 124 (86) 41412 (87)

>35 18 (10) 5 (3) 3485 (7)

Race <0.001 <0.001 0.102

Black 116 (67) 19 (13) 4422 (9)

White 16 (9) 8 (6) 1955 (4)

Hispanic 38 (22) 118 (81) 39975 (84)

Other 2 (1) 0 (0) 1039 (2)

Nulliparous 45 (26) 63 (43) 14488 (31) 0.001 0.210 <0.001

Parity 0.002 0.009 0.004

0 45 (26) 63 (43) 14488 (31)

1 53 (31) 47 (32) 14732 (31)

2 35 (20) 21 (14) 10690 (23)

3 19 (11) 9 (6) 4893 (10)

>3 20 (12) 5 (3) 2588 (5)

Multiple gestation 1 (1) 0 (0) 531 (1) 0.358 0.502 0.200

Data expressed as N (%) or mean ± standard deviation.

Table 2: Comorbidities, hepatitis B, and RPR results in HIV positive, false positive, and negative women.

CharacteristicPositiveN = 172

False positiveN = 145

NegativeN = 47391

P

Positive versusfalse positive

Positive versusnegative

False positiveversus negative

Diabetes 13 (8) 8 (6) 3015 (6) 0.467 0.521 0.677

Hypertension 20 (12) 16 (11) 4164 (9) 0.868 0.189 0.340

HBsAg 0.002 <0.001 0.580

No result 12 (7) 0 (0) 224 (0.5)

Negative 158 (92) 145 (100) 47038 (99)

Positive 2 (1) 0 (0) 129 (0.3)

RPR 0.020 <0.001 0.618

No result 5 (3) 0 (0) 31 (0.07)

Nonreactive 163 (95) 145 (100) 47078 (99)

Reactive 4 (2) 0 (0) 282 (0.6)

Data expressed as N (%).

with high HIV prevalence [7, 14, 15]. A woman testingpreliminarily positive for HIV in labor should be counseledthat she may have HIV infection and that her neonate may beexposed, and immediate antiretroviral prophylaxis should berecommended without waiting for confirmatory test results.The results of our study may aid in counseling womenif they test preliminarily positive for HIV using the EIA,while awaiting confirmatory testing results. It remains to bedetermined whether the same risk factors for false positiveHIV EIA testing apply to the POCT tests used in a real lifesetting.

Our study found that the positive predictive value of EIAtesting was only 54.3% and that younger nulliparous womenwere more likely to test falsely positive. The reasons for thesefindings are not entirely clear. Importantly, our data mayrepresent the real world performance of the EIA testing ina large obstetric population-based setting. Investigators alsohave noted a significant relationship between influenza vac-cination and false positive screening for HIV antibodies [16–18]. A potential reason for this cross reactivity is that thereare similarities in homology between the transmembranedomains of the influenza envelope protein hemagglutinin

Page 45: HIV/AIDS, Tuberculosis, and Malaria in Pregnancydownloads.hindawi.com/journals/specialissues/946815.pdf · HIV/AIDS, Tuberculosis, and Malaria in Pregnancy ... Fulgencio Proverbio,

4 Journal of Pregnancy

Table 3: HIV testing groups stratified by age and parity.

Nulliparous

HIV result≤19 yearsN = 4925

>19 yearsN = 9671

P

Positive 10 (0.20) 35 (0.36)

False positive 35 (0.71) 28 (0.29) 0.0003

Negative 4880 (99.1) 9608 (99.4)

Parous

HIV result≤19 yearsN = 1568

>19 yearsN = 31544

P

Positive 7 (0.45) 120 (0.38)

False positive 6 (0.38) 76 (0.24) 0.5

Negative 1555 (99.2) 31348 (99.4)

Data expressed as N (%).

and the HIV-1 envelope proteins [19]. In our population,there were no significant differences in age or parity andinfluenza vaccination acceptance rates (unpublished data).

There were several limitations to our study. Women inour study were from a single institution, with a predomi-nantly Hispanic background, and therefore the results maynot be applicable to all populations. Reasons for false positiveHIV serology may vary depending on the geographicalregion. While we did find a significant association betweenyoung age, nulliparity, and HIV false positive testing, ourstudy does not have the capability to identify a causalrelationship or explain why this association may exist.In addition, our study cannot address if these same riskfactors for false positive testing apply to all other HIVtests. Further studies are needed to confirm our findings,elucidate the biological mechanisms for increased HIV EIAfalse positivity in young, nulliparous women, and comparethis conventional testing method with contemporary rapidscreening assays, including POCT.

Acknowledgment

This work was presented at The Society for Maternal-FetalMedicine 31st Annual Meeting, February 11, 2011, SanFrancisco, CA, USA.

References

[1] Centers for Disease Control and Prevention, “Revised rec-ommendations for HIV testing of adults, adolescents, andpregnant women in health-care settings,” Morbidity andMortality Weekly Report, vol. 55, no. 14, pp. 1–17, 2006.

[2] U.S. Preventive Services Task Force, “Screening for HIV: rec-ommendation statement,” AHRQ Publication, no. 07-0597-EF-2, March 2011, http://www.uspreventiveservicestaskforce.org/uspstf05/hiv/hivrs.htm.

[3] “Recommendations of the U.S. Public Health Service TaskForce on the use of zidovudine to reduce perinatal tran-simission of human immunodeficiency virus,” Morbidity andMortality Weekly Report, vol. 43, no. 11, pp. 1–20, 1994.

[4] N. A. Wade, G. S. Birkhead, B. L. Warren et al., “Abbreviatedregimens of zidovudine prophylaxis and perinatal transmis-sion of the human immunodeficiency virus,” New EnglandJournal of Medicine, vol. 339, no. 20, pp. 1409–1414, 1998.

[5] P. M. Garcia, A. Kalish, J. Pitt et al., “Maternal levels of plasmaimmunodeficiency virus type 1 RNA and the risk of perinataltransmission,” New England Journal of Medicine, vol. 341, pp.394–402, 1999.

[6] L. M. Mofenson, J. S. Lambert, E. R. Stiehm et al., “Risk factorsfor perinatal transmission of human immunodeficiency virustype 1 in women treated with zidovudine,” New EnglandJournal of Medicine, vol. 341, no. 6, pp. 385–393, 1999.

[7] Centers for Disease Control and Prevention, “Rapid HIV-1 antibody testing during labor and delivery for women ofunknown HIV status: a practical guide and model proto-col,” November 2010, http://www.cdc.gov/hiv/topics/testing/resources/guidelines/rt-labor&delivery.htm.

[8] T. I. Doran and E. Parra, “False-positive and indeterminatehuman immunodeficiency virus test results in pregnantwomen,” Archives of Family Medicine, vol. 9, no. 9, pp. 924–929, 2000.

[9] L. G. Wesolowski, K. P. Delaney, M. A. Lampe, and S.R. Nesheim, “False-positive human immunodeficiency virusenzyme immunoassay results in pregnant women,” PLoS ONE,vol. 6, no. 1, Article ID e16538, 2011.

[10] Abbott Laboratories HIVABTM HIV-1/HIV-2 (rDNA) EIA,March 2011, http://www.fda.gov/downloads/BiologicsBlood-Vaccines/SafetyAvailability/TissueSafety/ucm095983.pdf.

[11] C. S. Tung, H. Sangi-Haghpeykar, and J. Levison, “Rapidversus standard testing for prenatal HIV screening in apredominantly Hispanic population,” Journal of Perinatology,vol. 30, no. 1, pp. 30–32, 2010.

[12] D. J. Jamieson, M. H. Cohen, R. Maupin et al., “Rapid humanimmunodeficiency virus-1 testing on labor and delivery in 17US hospitals: the MIRIAD experience,” American Journal ofObstetrics and Gynecology, vol. 197, no. 3, supplement, pp.S72–S82, 2007.

[13] M. Bulterys, D. J. Jamieson, M. J. O’Sullivan et al., “RapidHIV-1 testing during labor: a multicenter study,” Journal ofthe American Medical Association, vol. 292, no. 2, pp. 219–223,2004.

[14] B. M. Branson, H. H. Handsfield, M. A. Lampe et al., “Revisedrecommendations for HIV testing of adults, adolescents,and pregnant women in health-care settings,” Morbidity andMortality Weekly Report, vol. 55, no. 14, pp. 1–17, 2006.

[15] American College of Obstetricians and Gynecologists, “Pre-natal and perinatal human immunodeficiency virus testing:expanded recommendations,” Committee Opinion, no. 418,September 2008.

[16] W. R. Mac Kenzie, J. P. Davis, D. E. Peterson, A. J. Hibbard,G. Becker, and B. S. Zarvan, “Multiple false-positive serologictests for HIV, HTLV-1, and hepatitis C following influenza vac-cination, 1991,” Journal of the American Medical Association,vol. 268, no. 8, pp. 1015–1017, 1992.

[17] L. Simonsen, J. Buffington, C. N. Shapiro et al., “Multiplefalse reactions in viral antibody screening assays after influenzavaccination,” American Journal of Epidemiology, vol. 141, no.11, pp. 1089–1096, 1995.

[18] C. P. Erickson, T. McNiff, and J. D. Klausner, “Influenzavaccination and false positive HIV results,” New EnglandJournal of Medicine, vol. 354, no. 13, pp. 1422–1423, 2006.

[19] D. C. Chan, D. Fass, J. M. Berger, and P. S. Kim, “Core structureof gp41 from the HIV envelope glycoprotein,” Cell, vol. 89, no.2, pp. 263–273, 1997.