trichomoniasis and bacterial vaginosis in pregnancy

7/28/2019 trichomoniasis and bacterial vaginosis in pregnancy

1/10

297Bulletin o the World Health Organization | April 2007, 85 (4)

Objective To measure the prevalence o Trichomonas vaginalis(TV) inection and bacterial vaginosis (BV) among pregnant womenin Botswana, and to evaluate the syndromic approach and alternative management strategies or these conditions in pregnancy.Methods In a cross-sectional study, 703 antenatal care attendees were interviewed and examined, and specimens were collectedto identiy TV, BV, Candida species, Chlamydia trachomatisand Neisseria gonorrhoeae. Inormation on reproductive tract inectionsearlier in pregnancy was obtained rom a structured interview and the antenatal record.Findings TV was ound in 19% and BV in 38% o the attendees. Three-ourths o women with TV or BV were asymptomatic.Syndromic management according to the vaginal discharge algorithm would lead to substantial under-diagnosis and over-treatmento TV and BV. Signs o vaginal discharge were more predictive o the presence o these conditions than were symptoms. Among the546 attendees on a repeat antenatal visit, 142 (26%) had been diagnosed with vaginal discharge earlier in their pregnancy 14 othem twice. In 143 cases, an attendee was diagnosed with vaginal discharge in the second or third trimester; however, metronidazolehad been prescribed only 17 times (12%).

Conclusion Diagnosis and treatment o TV and BV among pregnant women in sub-Saharan Arica presents major challenges.Hal the pregnant women in this study were diagnosed with TV or BV, but these conditions were not detected and treated duringantenatal care with syndromic management. Also, health workers did not adhere to treatment guidelines. These results indicate thatmanagement guidelines or TV and BV in antenatal care should be revised.

Bulletin o the World Health Organization 2007;85:297-304.

Une traduction en ranais de ce rsum fgure la fn de larticle. Al fnal del artculo se acilita una traduccin al espaol.

Trichomoniasis and bacterial vaginosis in pregnancy:inadequately managed with the syndromic approachM Romoren,a M Velauthapillai,b M Rahman,c J Sundby,d E Klouman e & P Hjortdahl d

IntroductionTrichomonas vaginalis (V) inection

is the most common curable sexuallytransmitted inection (SI) worldwide.1In studies o low-risk women in sub-Saharan Arica, the prevalence rangesrom 1031%.2,3 Bacterial vaginosis (BV)is a syndrome characterized by a shitin vaginal ora; it is particularly com-mon in the sub-Saharan region, whereprevalences up to 50% are not uncom-mon.4 Tese two vaginal conditions arethought to cause substantial morbidityamong women in developing countries.Both inections have been linked topreterm delivery and low birth weight 5and, as reproductive tract inections(RIs), they are likely to increase bothinectiousness o HIV and susceptibilityto the disease.6,7 It appears to be criti-cal to diagnose and treat V and BV inpregnancy, especially in high-prevalencesettings.8,9

.

a Faculty o Medicine, University o Oslo, Oslo, Norway. Correspondence to Maria Romoren (e-mail: [email protected]).b National Health Laboratory, Ministry o Health, Gaborone, Botswana.c Sexual and Reproductive Health Associates, Gaborone, Botswana.d Faculty o Medicine, University o Oslo, Oslo, Norway.e Norwegian Institute o Pub lic Health, Oslo, Norway.doi: 10.2471/BLT.06.031922(Submitted: 23 March 2006 Final revised version received: 2 October 2006 Accepted: 16 November 2006)

Tere are ew studies rom devel-oping countries on eective strategies

to prevent the adverse outcomes associ-ated with V and BV in pregnancy. Sys-tematic reviews rom developed coun-tries o antibiotic treatment or theseconditions in asymptomatic pregnant

women show no signicant reductionsin adverse pregnancy outcomes.5,1012However, antibiotic treatment or BVmay reduce the risk o low birth weightand preterm rupture o the membranesamong pregnant women with previouspreterm deliveries.13 Provision o treat-ment or V or BV in symptomaticpregnant women has not been adequate-ly evaluated.11,12

Diagnosis o V and BV in womenin sub-Saharan Arica is based on thevaginal discharge syndrome the mostcommon syndrome in the syndromicapproach (i.e. treating symptoms andsigns o disease based on the organ-isms most commonly responsible or

the particular syndrome). In the early1990s, the World Health Organization

developed syndromic managementguidelines or symptomatic SI patientsor countries without laboratory sup-port. Easily recognized symptoms andsigns are combined using owcharts,and patients are then treated with two ormore antibiotic regimens.14 In Botswana,

women reporting vaginal discharge orlower abdominal pain are managed us-ing the vaginal discharge algorithm.15Based on a risk assessment and clinicalsigns, the women are provided withtreatment or V and BV and/or chla-mydia and gonorrhoea and/or candi-diasis. Where a woman has chlamydiaand gonorrhoea, partner treatment isalways recommended; however, wherea woman has V, partner treatmentis only recommended i the womanssymptoms persist.

For pregnant women, the BotswanaSI manual states that asymptom-


2/10

298 Bulletin o the World Health Organization | April 2007, 85 (4)

ResearchBacterial vaginosis in pregnancy M Romoren et al.

atic women with a history o previouspreterm delivery should be examinedor vaginal discharge to detect andtreat BV.15 Syndromic management oasymptomatic antenatal care attendeesin general is not recommended in ei-

ther the national or the World HealthOrganizations SI management guide-lines. In practice, however, all antenatalcare attendees in Botswana are clinicallyscreened or RIs because the countrysantenatal care guidelines recommend aroutine speculum examination at therst antenatal visit to exclude genitalinections, abnormalities and pelvictumours.16 It is not uncommon orabnormal vaginal discharge to be oundin women not displaying symptoms. Tenurses will act on pathological ndings,

and asymptomatic women with signs ovaginal discharge are thus provided withsyndromic treatment. Tis managementbypasses the original entry point o thesyndromic algorithms: symptoms thatlead patients to seek health care.

Te aim o this paper is to pres-ent results on the prevalence o V andBV among antenatal care attendees inBotswana, to examine the use o thevaginal discharge algorithm earlier inthe current pregnancy, and to evaluatethe syndromic approach and clinicalscreening in the diagnosis o these twoconditions in pregnancy.

MethodsA total o 703 pregnant women par-ticipated in this study. Te women wereselected rom those visiting the 13 mainacilities providing antenatal care (12 pri-mary health clinics and one outpatientdepartment) in Gaborone, Botswana,between October 2000 and February2001. A proportionate sample o attend-ees was recruited rom each location,

based on the percentage o all antenatalcare attendees who attended that acil-ity the previous year. In most clinics,all attendees were included in the datacollection; a sample o the attendees wasincluded rom the busiest clinics. All par-ticipants gave written, inormed consent;the study was approved by the nationalcommittees or research ethics in Bo-tswana and in Norway. Te only exclu-sion criterion was the use o antibioticsduring the previous two weeks.

A structured interview and data

rom the patient-held antenatal recordwere used to obtain inormation onsociodemographic actors, currentRI symptoms, and diagnosis and

prescribed treatment or RIs earlier inthe pregnancy. All attendees underwenta genital examination by a medicaldoctor, and clinical signs rom externaland internal genitalia were recorded.

Amount, consistency, colour and odouro vaginal discharge were described andcategorized as normal, Candida-likeor non-Candida-like discharge.

One high vaginal swab was placedin Stuart transport medium; another

was used or a vaginal smear. Specimenswere transported at ambient temperatureto the National Health Laboratory orurther processing. Wet-mounts maderom the swabs in transport media wereexamined or motile trichomonads bylight microscopy. Te swabs were thenagitated into bottles o Diamonds modi-ed medium, the bottles were incubatedin Oxoid gaspack jars, and wet-mounts

were examined or trichomonads oncedaily or up to ve days. Te vaginalsmears were Gram-stained and scored orBV according to Nugents criteria17 by anexperienced laboratory technician.

Culture oCandidaspecies was initi-ated by direct inoculation o Saboraudplates at the clinic, incubated at 35 C(5% CO2), and examined ater 24 and48 hours. Smears o colonies rom posi-tive cultures were Gram-stained andexamined or budding yeast cells and

pseudohyphae. Te wet-mounts andGram-stained smears were also exam-ined or Candida. Presence oCandida

was veried by positive growth and/ormicroscopy. Cervical swabs were ob-tained or ligase chain reaction (LCR)amplication, or direct, qualitativedetection o specic target nucleic acidsequences oChlamydia trachomatisandNeisseria gonorrhoeae.18

Data were analysed using the sta-tistical package SPSS, Version 11. oevaluate the clinical diagnosis o V and

BV, univariate logistic regression analy-ses were used to assess the associationbetween laboratory-veried diagnosesand genital symptoms and signs. So-ciodemographic risk actors and genitalsymptoms and signs that, in univariateanalysis, were associated at a 0.2 level(P-value o odds ratios [OR]), were in-cluded in multivariate logistic regressionanalysis. Validity o the vaginal dischargealgorithm and o the clinical screening

were assessed by measuring sensitivity,specicity, positive and negative likeli-

hood ratios (LR+ and LR), and positiveand negative predictive values, using thelaboratory diagnosis o V and BV asthe reerence standard.

ResultsGeneral characteristicsTe median age o the 703 antenatal careattendees was 25 years (range o 1543)and median gestational age 30 weeks(range o 842). Selected background

characteristics, genital symptoms andsigns, and the prevalence o RIs areshown in able 1. V and/or BV waspresent in 359 (51%) o the women,Chlamydia and/or gonorrhoea in 67(10%), and Candida species in 416(59%). Among the 132 women withV, 100 (76%) reported no symptomso vaginal discharge or lower abdominalpain. Among the 268 women with BV,205 (76%) were asymptomatic.

Vaginal discharge and its

association with TV and BVable 2 (available at http://www.who.int/bulletin) shows selected genitalsymptoms and signs, and their univari-ate association with V and BV. O thesymptoms, only vaginal discharge wasassociated with V, albeit weakly (OR1.6; 95% condence interval (95%CI),1.0 to 2.5), and only in the univariateanalysis. None o the genital symptomsevaluated were signicantly associated

with BV. Vaginal discharge and genitalitching were, however, signicantly as-

sociated with Candidaspecies. Candidawas identied in 84 (71%) o the 119women with symptoms o vaginal dis-charge, compared with 332 (57%) othe 585 women without this symptom(P< 0.01).

In the clinical evaluation o thedischarge, non-Candida-like vaginaldischarge was associated with increasedprevalence o V and BV, whereasCandida-like discharge was not (able2, available at http://www.who.int/bul-letin). Runny, rothy and malodorous

discharges were strongly associated withboth V and BV; results rom women

with one or more o these discharge char-acteristics were OR 7.1 (95%CI, 4.7 to10.8) or V, and OR 3.3 (95%CI, 2.3to 4.8) or BV. Adjusted odds ratios areshown in able 3.

able 4 compares the diagnosticaccuracy o the vaginal discharge al-gorithm, screening or signs o vaginaldischarge and or specic dischargecharacteristics. Te vaginal dischargealgorithm ailed to detect most cases

o V and BV. Also, women diagnosedby the algorithm as diseased were notsignicantly more likely to have Vand BV than women not diagnosed


3/10

299Bulletin o the World Health Organization | April 2007, 85 (4)

ResearchBacterial vaginosis in pregnancyM Romoren et al.

Table 1. Background characteristics, genital symptoms and signs and prevalence oreproductive tract inections (RTIs) among 703 antenatal care attendeesin Gaborone, Botswana

Characteristics n (%)

Age groups

1519 76 (11)2024 249 (35)2529 183 (26)3034 126 (18)3543 69 (10)

EducationPrimary school or less 168 (24)Junior secondary school 310 (44)Secondary school or higher 225 (32)

Marital statusMarried 114 (16)Non-marital steady partner 572 (81)

Single 17 (2)Living with husband/partner 353 (50)

Not living with husband/partner 350 (50)

Pregnancy number1st pregnancy 243 (35)2nd pregnancy 208 (30)3rd pregnancy 122 (17)4th+ pregnancy 130 (18)

Antenatal care visit number1st visit 157 (22)24th visit 300 (43)5th+ visit 246 (35)

Sel-reported symptoms o RTIsVaginal discharge 119 (17)Itching/soreness 58 (8)Lower abdominal pain 53 (8)Genital warts 16 (2)Genital ulcer 8 (1)Dysuria 8 (1)

Clinical signs o RTIsVaginal discharge (not Candida-like) 227 (32)Candida-like vaginal discharge 81 (12)Genital warts 29 (4)Genital ulcer 5 (1)

Presence o pathogensChlamydia trachomatis 53 (8)Neisseria gonorrhoeae 21 (3)Trichomonias vaginalis 132 (19)Bacterial vaginosis 268 (38)Candida species 416 (59)

as diseased (LR+ 1.35; 95%CI, 0.97to 1.89).

Signs o non-Candida-like vaginaldischarge gave an LR+ o 3.00 (95%CI,2.31 to 3.92) in the diagnosis o thetwo conditions combined. Screen-

ing the women or specic dischargecharacteristics increased the LR+ to6.66 (95%CI, 4.25 to 10.5), but alsoincreased the proportion o undetectedinections.

Diagnosis and treatment ovaginal discharge earlier incurrent pregnancy

An indicator o eectiveness o the cur-rent RI management in antenatal careis the prevalence o inections amongthe repeat attendees because, unlike new

attendees, they have already been pro-vided with standard care. V prevalencewas 15% or new and 20% or repeatattendees (P= 0.22) and BV prevalence

was 41% or new and 37% or repeatattendees (P= 0.44).

Among the 546 repeat attendees,142 (26%) had been diagnosed withvaginal discharge earlier in their cur-rent pregnancy 14 o them twice.Prevalence o V and BV was similar,

whether or not the women had a historyo vaginal discharge.

reatment guidelines or vaginaldischarge were usually not ollowedappropriately. At the time o the study,the recommended regimen to coverV and BV was 400 mg metronidazoletwice daily or 7 days; treatment orthese conditions in the rst trimestero pregnancy was not recommended.Consistent with the guidelines, met-ronidazole had not been prescribed toany o the 13 women diagnosed withvaginal discharge in the rst trimester.In 143 cases, however, attendees werediagnosed with vaginal discharge in thesecond or third trimester, but metro-nidazole had been prescribed only 17times (12%).

O the 17 attendees or whom met-ronidazole was prescribed, none hadV, compared to 132 (19%) o the 686

women who had not been prescribedthe drug (P< 0.05). Tree (18%) o the17 attendees had BV, compared to 265(39%) o the women who had not beenprescribed metronidazole (P> 0.1).

DiscussionIn our study o RIs in pregnantBatswana women, we ound high preva-lence o V and BV, and no indication

that syndromic management reduces theprevalence o these conditions. Te vagi-nal discharge algorithm was extensivelyused by the health workers, but had lowaccuracy in diagnosing V and BV. In

addition, sta oten did not adhere totreatment guidelines. V and BV maycontribute substantially to the risk opreterm delivery, low birth weight and

increased HIV transmission in sub-Saharan Arica. Apparently, diagnosisand treatment strategies or V and BVamong pregnant women in this settingrequire reconsideration.

Te study population is represen-tative o antenatal care attendees inGaborone. Te level o health care iscomparable throughout Botswana, and


4/10



HIV prevalence among pregnant womenis similar in urban and rural areas.19Combined with an antenatal coverageo 95%, this situation leads us to believe

that we present an accurate picture omanagement challenges o V and BVin pregnant women in the country as a

whole.20

Table 3. Univariate and multivariate logistic regression analysesa o determinants o trichomoniasis and bacterial vaginosisamong 703 antenatal care attendees in Gaborone, Botswana

Trichomoniasis N Crude odds ratio(95% confdence

interval)

Adjusted odds ratio(95% confdence interval)

% Positive(n= 132)

% Negative(n= 571)

Sociodemographic actors

Age groups< 20 20 9 76 4.55 (2.358.82) 7.05 (3.2315.37)2029 65 61 432 2.18 (1.293.66) 3.00 (1.665.43)30+ 15 31 195 1 1

EducationPrimary school or less 30 23 168 2. 65 (1.524.65) 5.25 (2.7510.02)Junior secondary school 54 42 310 2.56 (1.544.25) 2.68 (1.564.60)Senior secondary or higher 17 35 225 1 1

Antenatal care visitNew client 17 24 157 1 1Repeat client 83 77 546 1.45 (0.892.37) 1.67 (0.982.85)

SymptomsMore discharge than usual

No 77 84 584 1 1Yes 23 16 119 1.59 (1.002.54) 1.09 (0.631.87)

Clinical signs

Vaginal dischargeNegative 36 61 395 1 1Candida-like discharge 3 14 81 0.38 (0.131.07) 0.32 (0.110.96)Other vaginal discharge 61 26 227 3.93 (2.625.91) 3.61 (2.315.65)

Bacterial vaginosis

% Positive(n= 268)

% Negative(n= 435)

Sociodemographic actors

Age groups< 20 14 9 76 2.10 (1.223.60) 1.73 (0.751.62)2029 62 61 432 1.32 (0.921.89) 1.10 (0.751.62)30+ 24 30 195 1 1

Marital statusUnmarried 88 81 589 1.63 (1.052.52) 1.37 (0.852.19)Married 12 19 114 1 1

Symptoms

Malodorous dischargeNo 94 96 671 1 1Yes 6 4 32 1.66 (0.823.38) 1.34 (0.622.88)

Clinical signs

Vaginal dischargeNegative 50 59 395 1 1Candida-like discharge 4 16 81 0.27 (0.140.54) 0.26 (0.130.52)Other vaginal discharge 46 24 227 2.28 (1.633.18) 2.13 (1.512.99)

a Only actors that were associated with the respective conditions at a 0.2 level ( P-value o odds ratio) in the univariate analysis are presented in the table.

Clinical management o TV andBV in pregnancyIn antenatal care, health-care workersmanage women who report vaginal

discharge and women in whom vaginaldischarge is ound at the routine exami-nation. Te vaginal discharge algorithmrecognizes that women reporting vaginal

discharge commonly suer rom V, BVor vulvovaginal candidiasis and, in rarercases, a cervical inection.2123 Symptomso vaginal discharge or lower abdominal

pain are unspecic, especially in preg-nancy, where physiological dischargeand the presence o candidiasis increase.

We ound that V and BV were equally


5/10


6/10



Revised guidelines in Botswana rec-ommend metronidazole 2 g in a singledose to cover V in pregnancy, and250 mg three times daily or 7 days tocover BV.15 Tis management is basedon evidence o optimal treatment eect

or each condition, and would minimizemetronidazole use in pregnancy i thewomen were treated or only one othe conditions.13,37,42 However, as anetiological diagnosis cannot be madein primary health care, this strategy re-quires the prescription o two regimensto pregnant women with vaginal dis-charge a situation that is both conusingand unnecessary. According to recentlyupdated guidelines rom the Centers orDisease Control and Prevention, metro-nidazole 500 mg twice daily or 7 days is

the only regimen which eectively treatsboth conditions.37

Simple diagnostic tests toidentiy TV and BV

Accurate diagnosis and prompt treat-ment o SIs is important rom a publichealth perspective, and this includesdetection o asymptomatic cases. Withthe demonstrated shortcomings o themanagement o V in Botswana, andthe lack o treatment or partners o

women whose symptoms do not persist,

the high prevalence o V is likely tocontinue.

A ocus on the implementationo the syndromic approach may havesuperseded an exploration o the use opoint-o-care tests or V and BV indeveloping countries. Contrary to manyother RIs, simple and cheap tests or

V and BV are available. V can beidentied immediately with a simplelatex agglutination test 43 or by saline

wet preparation. BV can be diagnosedon-site, with or without the aid o amicroscope.44

Introduction o point-o-care testsor V and BV in the antenatal carein countries where prevalences o RIsand HIV are high requires urther ex-ploration. Antenatal screening wouldidentiy asymptomatic cases, and testingattendees with vaginal discharge would

reduce over-treatment with metroni-dazole in pregnancy. Health workersprescriptions and patients compliance

would probably improve with a specicdiagnosis, and treatment regimens couldbe optimized. reating sexual partners o

women with vaginal discharge has beendebated because the majority o theidentied women do not have an SI.45

An important benet o testing or V isthe relative ease o notiying and treatingpartners o patients with a positive test.

Botswana has a relatively well-

unctioning health system and couldserve as an exploratory site or the use

o point-o-care tests or V and BV.Trough the programme to preventmother-to-child transmission o HIV, allhealth posts and clinics have lay workers

who perorm rapid tests or HIV. Simpletests or V and BV perormed by clini-

cians or lay workers could contribute toimproving diagnosis and reducing thedisease burden o these conditions insub-Saharan Arica.

ConclusionTe inaccuracy o vaginal discharge inpredicting pathological conditions inpregnancy and the magnitude o asymp-tomatic V and BV is a challenge indeveloping countries, as is the qualityo care provided. Guidelines are inad-equate and there is a lack o adherenceto guidelines by health workers. Ourresults rom Botswana indicate thatnational health authorities should revisethe diagnosis and treatment guidelinesor V and BV in antenatal care. Also,the results o this study may be useulin the process o continuous revision o

World Health Organization guidelinesor the management o these conditionsin developing countries. O

Competing interests: None declared.

Rsum

Insufsances de lapproche syndromique pour la prise en charge de la trichomonase et de la vaginitebactrienne chez la emme enceinteObjecti Dterminer la prvalence des inections Trichomonasvaginalis(TV) et des vaginites bactriennes (VB) chez les emmesenceintes du Botswana et mener une valuation compare delapproche syndromique et dautres stratgies pour la prise encharge de ces pathologies pendant la grossesse.

Mthodes Dans le cadre dune tude transversale, on a interroget examin 703 emmes se prsentant aux visites prnatales eton a prlev chez elles des chantillons pour recherche de TV,dune ventuelle VB, de Candida, de Chlamydia trachomatiset deNeisseria gonorrhoeae. Des entretiens structurs et lexploitationdes dossiers prnatals ont permis de recueillir des lments surles inections de lappareil reproducteur survenues un stadeantrieur de la grossesse.Rsultats On a diagnostiqu une inection TV chez 19%des emmes examines et une VB chez 38% dentre elles. Troisquarts des emmes atteintes dinection TV ou VB taientasymptomatiques. Une prise en charge syndromique selonlAlgorithme pour les pertes vaginales de lOMS aurait conduit

un sous-diagnostic notable de certaines pathologies et unsurtraitement des inections TV et des VB. Les pertes vaginales

se sont rvles plus ortement prdictives de la prsence deces aections que les symptmes. Parmi les 546 emmes reuesdans le cadre des visites prnatales, 142 (26 %) avaient dj tdiagnostiques comme prsentant des pertes vaginales un stadeantrieur de la grossesse ( deux reprises pour 14 dentre elles). On

a relev 143 cas de pertes vaginales au deuxime ou au troisimetrimestres de grossesse parmi les emmes examines lors de cesvisites prnatales. Nanmoins, ces pertes nont donn lieu que 17ois la prescription de mtronidazole (12 % des cas).Conclusion Le diagnostic et le traitement des inections TVet des VB chez les emmes enceintes dArique subsaharienneprsentent des dicults majeures. Cette tude a conduit diagnostiquer une inection TV ou une VB chez la moiti desemmes enceintes, mais ces aections nont t ni dtectes, nitraites, par la prise en charge syndromique applique lors desvisites prnatales. De mme, le personnel soignant ne suivait parles recommandations thrapeutiques. Ces rsultats tmoignentde la ncessit dune rvision du Guide pour la prise en charge

des IST TV et des VB dans le cadre des soins prnatals.


7/10


8/10



11. Gulmezoglu AM. Interventions or trichomoniasis in pregnancy (CochraneReview). In: The Cochrane Library, Issue 3, 2005.

12. McDonald H, Brocklehurst P, Parsons J. Antibiotics or treating bacterialvaginosis in pregnancy (Cochrane Review). In: The Cochrane Library, Issue 3,2005.

13. Koumans EH, Markowitz LE, Hogan V. Indications or therapy and treatmentrecommendations or bacterial vaginosis in nonpregnant and pregnant

women: a synthesis o data.Clin Inect Dis

2002;35:S152-72.14. Guidelines or the management o sexually transmitted inections. Geneva:WHO; 2003. Available at: http://whqlibdoc.who.int/publications/2003/9241546263.pd

15. Botswana Ministry o Health. Management o Sexually TransmittedInections. Reerence Manual or Health Workers. Gaborone: MoH; 2005.

16. Botswana Ministry o Health. Guidelines or antenatal care and themanagement o obstetric emergencies and prevention o mother-to-child

transmission o HIV. Gaborone: Ministry o Health; 2005.17. Nugent RP, Krohn MA, Hillier SL. Reliability o diagnosing bacterial vaginosis

is improved by a standardized method o gram stain interpretation. J ClinMicrobiol1991;29:297-301.

18. Romoren M, Rahman M, Sundby J, Hjortdahl P. Chlamydia and gonorrhoeain pregnancy: eectiveness o diagnosis and treatment in Botswana. SexTransm Inect2004;80:395-400.

19. Botswana Ministry o Health. Botswana 2005 second generation HIV/AIDSsurveillance. Gaborone: MoH; 2006.20. National AIDS Coordinating Agency. Botswana AIDS Impact Survey II 2004.

Gaborone; 2005.21. Dallabetta GA, Gerbase AC, Holmes KK. Problems, solutions, and challenges

in syndromic management o sexually transmitted diseases. Sex Transm Inect1998;74:S1-11.

22. Mayaud P, ka-Gina G, Cornelissen J, Todd J, Kaatano G, West B, et al.Validation o a WHO algorithm with risk assessment or the clinicalmanagement o vaginal discharge in Mwanza, Tanzania. Sex Transm Inect1998;74:S77-84.

23. Fonck K, Kidula N, Jaoko W, Estambale B, Claeys P, Ndinya-Achola J, et al.Validity o the vaginal discharge algorithm among pregnant and non-pregnant women in Nairobi , Kenya. Sex Transm Inect2000;76:33-8.

24. Anderson MR, Klink K, Cohrssen A. Evaluation o vaginal complaints. JAMA

2004;291:1368-79.25. Tann CJ, Mpairwe H, Morison L, Nassimu K, Hughes P, Omara M, et al. Lacko eectiveness o syndromic management in targeting vaginal inections inpregnancy in Entebbe, Uganda. Sex Transm Inect2006;82:285-9.

26. Hylton-Kong T, Brathwaite AR, Del Rosario GR, Kristensen S, Kamara P,Jolly PE, et al. Marginal validity o syndromic management or reproductivetract inections among pregnant women in Jamaica. Int J STD AIDS2004;15:371-5.

27. Blankhart D, Muller O, Gresenguet G, Weis P. Sexually transmitted inectionsin young pregnant women in Bangui, Central Arican Republic. Int J STDAIDS1999;10:609-14.

28. Sturm AW, Wilkinson D, Ndovela N, Bowen S, Connolly C. Pregnant womenas a reservoir o undetected sexually transmitted diseases in rural SouthArica: implications or disease control. Am J Public Health 1998;88:1243-5.

29. Smith KS, Tabrizi SN, Fethers KA, Knox JB, Pearce C, Garland SM. Comparison

o conventional testing to polymerase chain reaction in detection oTrichomonas vaginalis in indigenous women living in remote areas. Int J STDAIDS2005;16:811-5.

30. van Der Schee C, van Belkum A, Zwijgers L, van Der Brugge E, ONeill EL,Luijendijk A, et al. Improved diagnosis o Trichomonas vaginalis inection byPCR using vaginal swabs and urine specimens compared to diagnosis bywet mount microscopy, culture, and fuorescent staining. J Clin Microbiol1999;37:4127-30.

31. Boonstra E, Lindbaek M, Klouman E, Ngome E, Romoren M, Sundby J.Syndromic management o sexually transmitted diseases in Botswanas

primary health care: quality o care aspects.Trop Med Int Health

2003;8:604-14.32. Vuylsteke BL, Ettiegne-Traore V, Anoma CK, Bandama C, Ghys PD,

Maurice CE, et al. Assessment o the validity o and adherence to sexuallytransmitted inection algorithms at a emale sex worker clinic in Abidjan,Cote dIvoire. Sex Transm Dis2003;30:284-91.

33. Kane BG, Degutis LC, Sayward HK, DOnorio G. Compliance with the Centersor Disease Control and Prevention recommendations or the diagnosis andtreatment o sexually transmitted diseases. Acad Emerg Med2004;11:371-7.

34. Moses S, Elliott L. Sexually transmitted diseases in Manitoba: evaluation ophysician treatment practices, STD drug utilization, and compliance withscreening and treatment guidelines. Sex Transm Dis2002;29:840-6.

35. Mak DB, Holman CD. STDs arent sexy: health proessionals lack oadherence to clinical guidelines in an area o high STD endemicity. J Public

Health Med2000;22:540-5.36. Owen MK, Clenney TL. Management o vaginitis. Am Fam Physician 2004;70:2125-32.

37. Centers or Disease Control and Prevention. Sexually transmitted diseasestreatment guidelines 2006. Available at: www.cdc.gov/std/treatment/2006/toc.htm

38. Caro-Paton T, Carvajal A, Martin de Diego I, Martin-Arias LH, AlvarezRequejo A, Rodriguez Pinilla E. Is metronidazole teratogenic? A meta-analysis.Br J Clin Pharmacol1997;44:179-82.

39. Burtin P, Taddio A, Ariburnu O, Einarson TR, Koren G. Saety o metronidazolein pregnancy: a meta-analysis. Am J Obstet Gynecol1995;172:525-9.

40. Klebano MA, Carey JC, Hauth JC, Hillier SL, Nugent RP, Thom EA, et al.Failure o metronidazole to prevent preterm delivery among pregnantwomen with asymptomatic Trichomonas vaginalis inection. N Engl J Med2001;345:487-93.

41. Kigozi GG, Brahmbhatt H, Wabwire-Mangen F, Wawer MJ, Serwadda D,Sewankambo N, et al. Treatment o Trichomonas in pregnancy and adverseoutcomes o pregnancy: a subanalysis o a randomized trial in Rakai, Uganda.Am J Obstet Gynecol2003;189:1398-400.

42. Forna F, Gulmezoglu AM. Interventions or treating trichomoniasis in women(Cochrane Review). In: The Cochrane Library, Issue 3, 2005.

43. Adu-Sarkodie Y, Opoku BK, Danso KA, Weiss HA, Mabey D. Comparisono latex agglutination, wet preparation, and culture or the detection oTrichomonas vaginalis. Sex Transm Inect2004;80:201-3.

44. Sexually transmitted and other reproductive tract inections. A guide toessential practice. Geneva: WHO; 2005. Available at: http://www.who.int/reproductive-health/publications/rtis_gep/rtis_gep.pd

45. Hawkes S, Mabey D, Mayaud P. Partner notication or the control o sexuallytransmitted inections. BMJ2003;327:633-4.


9/10

ABulletin o the World Health Organization | April 2007, 85 (4)

ResearchBacterial vaginosis in pregnancyM Romoren et al.

Table 2. Association between symptoms and signs and laboratory-verifed trichomoniasis or bacterial vaginosis among 703antenatal care attendees in Gaborone, Botswana

Trichomoniasis N Odds ratio(95% confdence interval)

P-value

% Positive(n= 132)

% Negative(n= 571)

Symptoms

More discharge than usualNo 77 84 584 1Yes 23 16 119 1.59 (1.002.54) 0.049

Itching or sorenessNo 91 92 645 1Yes 9 8 58 1.14 (0.592.22) 0.70

Lower abdominal painNo 96 92 650 1Yes 5 8 53 0.53 (0.221.27) 0.15

Clinical signs

Moderate/prouse dischargeNo 46 68 447 1Yes 55 32 256 2.52 (1.723.71) 0.000

Runny dischargeNo 62 93 615 1Yes 38 7 88 8.55 (5.2313.84) 0.000

Malodorous dischargeNo 84 96 659 1Yes 16 4 44 4.51 (2.418.43) 0.000

Frothy dischargeNo 56 90 590 1Yes 44 10 113 7.35 (4.7311.44) 0.000

Specic discharge characteristicsNegative 46 86 552 1Runny/malodorous/rothy 54 14 151 7.14 (4.7110.83) 0.000

Vaginal dischargeNegative 36 61 395 1Candida-like discharge 3 14 81 0.38 (0.131.07) 0.068Other vaginal discharge 61 26 227 3.93 (2.625.91) 0.000

Bacterial vaginosis

% Positive(n= 268)

% Negative(n= 435)

Symptoms

More discharge than usual

No 82 84 584 1Yes 18 16 119 1.12 (.0751.67) 0.585

Itching or sorenessNo 91 92 645 1Yes 9 8 58 1.25 (0.732.16) 0.415

Lower abdominal painNo 92 93 650 1Yes 8 7 53 1.17 (0.662.06) 0.598

Clinical signs

Moderate/prouse dischargeNo 62 65 447 1Yes 38 35 256 1.15 (0.841.58) 0.383

Runny dischargeNo 82 91 615 1Yes 18 9 88 2.16 (1.373.38) 0.001


10/10

B Bulletin o the World Health Organization | April 2007, 85 (4)


Bacterial vaginosis N Odds ratio(95 % confdence interval)

P-value

% Positive(n= 268)

% Negative(n= 435)

Malodorous discharge

No 88 97 659 1Yes 12 3 44 4.25 (2.188.27) 0.000

Frothy dischargeNo 74 90 590 1Yes 26 10 113 3.08 (2.044.67) 0.000

Specic discharge characteristicsNegative 66 86 552 1Runny/malodorous/rothy 34 14 151 3.33 (2.304.84) 0.000

Vaginal dischargeNegative 50 59 395 1Candida-like discharge 4 16 81 0.27 (0.140.54) 0.000Other vaginal discharge 46 24 227 2.28 (1.633.18) 0.000

(Table 2, cont.)

trichomoniasis and bacterial vaginosis in pregnancy

Documents