Cheongetal.BMCFamilyPractice2013,14:19 http://www.biomedcentral.com/1471-2296/14/19 RESEARCH ARTICLE Open Access How useful is a history of rubella vaccination for determination of disease susceptibility? A cross-sectional study at a public funded health clinic in Malaysia Ai Theng Cheong1*, Seng Fah Tong2 and Ee Ming Khoo3 Abstract Background: Identification of pregnant women susceptible to rubella is importantas vaccinationcan be given postpartum to prevent future risksof congenital rubella syndrome. However,in Malaysia, rubella antibody screening is not offered routinely to pregnant women in public funded health clinics due to cost constraint. Instead, a history ofrubella vaccination is asked to be provided to establish the women’srisk for rubella infection. The usefulness of this history, however, is not established. Thus, this paper aimed to determine theusefulness of a history of rubella vaccination indetermining rubella susceptibility inpregnant women. Methods: A cross-sectional study was conducted on500 pregnant women attending a public funded health clinic. Face-to-face interviews were conducted, and demographic data and history of rubella vaccination were obtained. Anti-rubella IgG test was performed. Results: A majority ofthe women (66.6%)had a positive vaccination history. Of these, 92.2%womenwere immune. A third (33.4%) of the women had a negative or unknown vaccinationhistory, but 81.4% of them were immuneto rubella. The sensitivity and specificityof a history ofrubella vaccination inidentifying disease susceptibility was 54.4% (95%CI: 40.7, 67.4%)and 69.3% (95%CI: 64.7, 73.5%)respectively; the positive predictivevalue was 18.6% (95%CI:13.1, 25.5%)and thenegativepredictivevalue was 92.2% (95% CI:88.6, 94.7%). Conclusions: A vaccination history ofrubella had a poor diagnostic value in predicting rubella susceptibility. However, obtaining a vaccinationhistory is inexpensive compared with performing a serological test. A cost-utility analysis would be useful indetermining which test (history versus serological test)is more cost-effective in a country with resource constraint. Keywords: Rubella susceptibility,History, Vaccination, Sensitivity, Specificity Background Although some women may acquire natural immunity Rubella infection is a mild disease when it affects against rubella by virtue of being infected in childhood, children. However, when it affects pregnant women in which is often asymptomatic [3], it is still important to the early trimester, it can cause serious complications vaccinate susceptible women as CRS carries serious such asmiscarriage tothe motherandcongenital rubella consequences. syndrome (CRS) to the fetus [1,2]. CRS is an important Cutts et al. (1999) estimated a total of 110,000 CRS cause of severe birth defects with ophthalmic, auditory, cases in the developing countries [4]. The incidence rate cardiacandneurologicalabnormalities. ofCRSindevelopingcountriesrangedfrom0.4to4.3per 1000 live births [5]. In Malaysia, a retrospective review of rubella serology from 1993–1998 at University Hospital, *Correspondence:[email protected] 1DepartmentofFamilyMedicine,FacultyofMedicineandHealthSciences, Kuala Lumpur had reported an incidence rate of 0.5 per UniversitiPutraMalaysia,Serdang,Selangor43400,Malaysia 1000 live births for CRS [5,6]. In recent years, sporadic Fulllistofauthorinformationisavailableattheendofthearticle ©2013Cheongetal.;licenseeBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsoftheCreative CommonsAttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,distribution,and reproductioninanymedium,providedtheoriginalworkisproperlycited. Cheongetal.BMCFamilyPractice2013,14:19 Page2of5 http://www.biomedcentral.com/1471-2296/14/19 cases of CRS have been reported in literature [7,8], to45yearsattendingtheclinicforantenatalbookingwere althoughtheexactnumberisunknown. invitedtoparticipateinthestudy.Womenwithsymptoms One of the ways to reduce CRS is to vaccinate all ofpossibleacuterubellainfectionassignifiedbyfeverand women before they reach reproductive age. The World rashwereexcludedfrom thestudy.Pregnantwomenwho Health Organization (WHO) has reported that 131 out consented to the study were recruited. For women aged of 193 WHO member countries (68%) have included ru- lessthan18years,writtenconsentsweretakenfromboth bella vaccination in their national immunization sched- the women and their guardians. The study was approved ule [9]. Malaysia started the rubella vaccination program by the Medical Ethics Committee of University Malaya in 1988 targeting school girls aged 12 to 15 years, and MedicalCentreandtheMinistryofHealth,Malaysia. women in the reproductive age group (15 – 44 years) [10]. Since 2002, the program has replaced rubella Setting vaccine with the 2-dose measles-mumps-rubella (MMR) Malaysia is a multiracial country with a population of vaccination for all children aged 1 and 7 years [11]. In 28.3 million according to the population census of 2010. 2011, the uptake of the MMR vaccination for children The ethnic distribution consists of 67.4% Malay, 24.6% aged 1 to 2 years was about 95% [12]. Despite the Chinese, 7.3% Indian and 0.7% others [25]. Selangor is vaccination program, outbreak of rubella has been the most densely populated among the fourteen states of reported among 16-year old students in a military voca- the country. It has a population of 5.46 million and is tional training school in 2006 [13] and the prevalence of highly urbanized. The health clinic in this study served a rubella susceptibility in pregnant women has been population of 67,578 [25]. It has a maternal and child reported to range between 8% and 11% in Malaysia health clinic that provides free service to Malaysian citi- [14,15]. This prevalence is relatively high compared with zens. The ethnic distribution of this study-population of some countries in the Asia-pacific region such as female aged 15–45 years was similar to the Selangor and Australia (2.7%) [16] and Japan (6.7%) [17], but is lower Malaysiancensus[25,26]. than Singapore (15.8%), Thailand (18.0%), Taiwan (16.7%) and Sri Lanka (24%) [18-21]. Therefore, efforts Datacollection are still needed to reduce rubella susceptibility among Face-to-face interviews were conducted using structured pregnantwomen. data collection forms to obtain socio-demographic data, To achieve a reduction in rubella susceptibility in history of rubella vaccination and past history of rubella women, apart from implementing rubella vaccination as infection. (Additional file 1). For history of rubella vacci- part of the childhood immunization schedule, supple- nation, the following question was asked: “Have you ever mentary vaccination can be employed as an additional been vaccinated for rubella?” There were three options strategy [22]. In some countries, screening for rubella for response: “yes”, “no” or “don’t know”. Those who antibodies is carried out duringantenatal period to iden- answered “no” or “don’t know” were considered suscep- tify susceptible women who are then vaccinated post tible to rubella infection. The history of rubella vaccin- partum [22-24]. In Malaysia, rubella antibody is not rou- ation was self reported as there was no documentation of tinely screened for in pregnant women in public funded vaccination available for verification. The women were health clinics due to cost constraint. Instead, these alsoaskedforpasthistoryofrubellainfection.Thosewho women are routinely asked for a history of rubella vac- answeredyeswereaskediftheyhadthediagnosisverified cination to identify their risk for rubella infection as eitherbyadoctororbyabloodtest. some studies have shown that rubella vaccination status All recruited women then had their serum IgG rubella is significantly associated with susceptibility to rubella antibodylevelmeasured,andthislevelwasusedasthegold infection [20,21]. Women who have no history of rubella standardforrubellasusceptibility.TheIgGrubellaantibody vaccination are then offered the vaccination post par- was measured using the AxSYM Rubella IgG assay based tum. The aim of this study was thus to determine the on the Microparticle Enzyme Immunoassay (MEIA) tech- usefulness of a history of rubella vaccination as a scree- nology.ArubellaIgGantibodylevelof10IU/mLorgreater ning tool for rubella susceptibility in countries such as indicates immunity to rubella infection either through Malaysia that do not offer routine antenatal rubella anti- previous rubella infection or induced by previous vaccine. body screening. ArubellaIgGantibodyleveloflessthan10IU/mLisconsi- deredsusceptibletorubellainfection. Methods Studydesignandparticipants Samplesize A cross-sectional study was carried out at a public Using epi-info version 6.0, the sample size was calcu- funded health clinic from June to October 2005 in the lated based on the expected prevalence of rubella state of Selangor, Malaysia. All pregnant women aged 15 susceptibility of 8% and the worst acceptable prevalence Cheongetal.BMCFamilyPractice2013,14:19 Page3of5 http://www.biomedcentral.com/1471-2296/14/19 of 11% with a confidence level of 95%. The estimated Table2Sensitivity,specificity,positiveandnegative samplesizewas314. predictivevalueofrubellavaccinationhistoryagainst rubellasusceptibility Statisticalanalysis Rubellavaccination Rubellasusceptibility Total Data were analyzed using SPSS version 15. Sensitivity history (anti-rubellaIgGtest) (%) and specificity, the positive and negative predictive Susceptible Immune values of a history of rubella vaccination in identifying (IgG<10IU/mL) (IgG≥10IU/mL) N(%) N(%) rubella susceptibility were determined using anti-rubella IgG level as the gold standard. In this study, the “test” Nohistoryofvaccinationor 31(18.6%) 136(81.4%) 167 unknownvaccinationstatus (100%) was a “history of rubella vaccination”. The aim of the (rubellasusceptible) “test” was to correctly identify women who were suscep- Historyofvaccination 26(7.8%) 307(92.2%) 333 tible to rubella. The gold standard for rubella suscep- (rubellaimmuned) (100%) tibility was those with anti-rubella IgG<10 IU/mL. Total 57 443 500 Sensitivity of the test is defined as the percentage of all Sensitivityofrubellavaccinationhistoryforrubellasusceptibility=31/(31+26) susceptible women who indeed reported to be not vacci- x100=54.4%(95%CI:40.7,67.4%). nated or to be unknown whether vaccinated. Specificity Specificityofrubellavaccinationhistoryforrubellasusceptibility=307/(307 +136)x100=69.3%(95%CI:64.7,73.5%). is defined as the percentage of all protected women that Positivepredictivevalueofrubellavaccinationhistory=31/(31+136)x100= indeed reported to be vaccinated. Positive predictive 18.6%(95%CI:13.1,25.5%). value of the test is the chance that a woman who Negativepredictivevalueofrubellavaccinationhistory=307/(307+26)x100 =92.2%(95%CI:88.6,94.7%). reportedtobenot vaccinated wassusceptible torubella. our previous analysis have shown that a negative vacci- Results nation history was a significant predictor for rubella Five hundred and two (502) patients who satisfied the susceptibility (odd ratio=2.7, 95% CI: 1.5, 4.7) after con- inclusion criteria were approached. Two patients refused trolling for age, ethnicity, parity, education level and to participate. The response rate was 99.6%. A majority of occupation[15]. women were Malay (55.2%), followed by Chinese (23.0%), Thesensitivityofahistoryofrubellavaccinationiniden- Indian (8.6%) andothers (13.2%). The age ranged from 16 tifying rubella susceptibility using serological test as gold to 42 years, the mean age being 27.2 (SD 4.8) years; the standardwas54.4%(95%CI:40.7,67.4%)andthespecificity medianagewas27.0years(IQR6.0).Halfthewomenwere was 69.3% (95% CI: 64.7, 73.5%). The positive predictive primigravida(49.2%).Noneofthemreportedapasthistory value was 18.6% (95% CI: 13.1, 25.5%) and the negative ofrubellainfectionthatwasconfirmedbyclinicaldiagnosis predictivevaluewas92.2%(95%CI:88.6,94.7%)(Table2). of a physician or by blood test. The prevalence of rubella susceptibilityinthisstudywas11.4%[95%CI:8.6,14.2%]. Discussion Most of the women (66.6%) had a positive vaccination The aim of the test (vaccination history) was to identify history. Of these, 92.2% were immune. (Table 1) A third women susceptible to rubella in order to offer them (33.4%) of the women had a negative vaccination history vaccination to prevent rubella in future pregnancies. For (no history of previous vaccination or unknown vaccin- this purpose, sensitivity is the most important element ation status); nevertheless, 81.4% of them were immune of the test. The sensitivity of the test was 54%. Hence, to rubella. (Table 2) A history of rubella vaccination on the basis of just a history of rubella vaccination, was significantly associated with rubella susceptibility almost half the susceptible women would be overlooked (χ2 = 12.737, P ≤0.001). Logistic regression results from and remain unprotected against rubella. On the other hand,comparedwithnoserologicaltestingatall,anega- tive vaccination history would at least identify half of the Table1Selfreportedrubellavaccinationhistoryand rubellasusceptibility susceptible women. The specificity of the test was 69%; however, it is not a clinical problem that 31% of the Selfreported Rubellasusceptibility TotalN=500 rubella (anti-rubellaIgGtest) N(%) women were protected while they reported to be not vaccination Susceptible Non-susceptible vaccinated. This could be due to natural immunity they history (IgG<10IU/mL) (IgG≥10IU/mL) acquired during childhood subclinical infection that N=57N(%) N=443N(%) often goes unnoticed. The positive predictive value of Nohistoryof 22(33.3) 44(66.7) 66(100.0) the test is only 19%, which means that there is a one in vaccination five chance that a woman who reported to be not vacci- Unknown 9(8.9) 92(91.1) 101(100.0) nated was susceptible to rubella, the majority being pro- vaccinationstatus tected. The negative predictive value was 92%, which Vaccinated 26(7.8) 307(92.2) 333(100.0) meansthatthemajority ofwomen whoreportedpositive Cheongetal.BMCFamilyPractice2013,14:19 Page4of5 http://www.biomedcentral.com/1471-2296/14/19 vaccination history were protected. The relatively low design and thesocio-economicstatus ofthe participants, sensitivity and positive predictive value of the rubella who were mainly from the middle and lower socio- vaccination history might not be useful in screening for economic groups that may differ from patients seen in rubella-susceptible women. Serological testing, the gold other settings. Therefore, caution is necessary in genera- standard test, is superior in correctly identifying women lizing the findings. Nevertheless, this study that has pro- who are susceptible to rubella. Nonetheless, serological vided us the preliminary evidence of an inexpensive test, test is expensive compared with an enquiry of vacci- namely, a history of rubella vaccination, is not a useful nation history. A cost-utility analysis is needed to deter- screening tool for rubella susceptibility. Further study is mine which test is preferable from a public health point needed to confirm this as it may have implications on of view: a history of vaccination “test” or a serological the healthcare policy regarding rubella antibody testing test,especiallyfor aresource constrainedcountry. inpregnant women. 7.8% of the women who reported having had vacci- nation were found to have low antibody titer level and Conclusions were considered to be not protected based on the cut-off A test based on the history of rubella vaccination has level of anti-rubella IgG of 10 IU/mL. This proportion is low sensitivity, specificity and predictive values in deter- substantial given the serious complications resulting from mining rubella susceptibility. Serological testing needs to congenital rubella syndrome. Therefore, screening for be considered in all women of childbearing age to effec- rubella susceptibilityusingahistoryofrubellavaccination tively prevent the occurrence of congenital rubella syn- would have overlooked this group of women and denied drome. However, as serology test is more expensive than them the benefit of rubella booster dose to prevent con- taking a vaccination history, a cost-utility analysis is genital rubella syndrome. Some studies have reported the recommendedincaseofresource constraint. possibilityofagradualdeclineintherubellaantibodytiter from a protective level to a non-protective level in some Additional file individuals who had the vaccination [27-29]. Therefore, it is possible for rubella to occur during pregnancy even in Addistionalfile1:Structureddatacollectionform. womenwhohavehadthevaccination. About two-third of the women who reported negative Competinginterests rubella vaccination history had protective antibody titers. Theauthorsdeclarenocompetinginterests. Some of these women may have acquired the immunity through subclinical infections they were not aware of Authors’contributions Allauthorscontributedtothedesign,analysisandwriteupofthisstudy. when they were young [3]. Some women may have expe- ATCproducedthefirstdraftofthemanuscript.Allauthorscontributedto rienced fever and rash and did not relate it to rubella therevisionofthedraft.Thefinalmanuscripthasbeenreadandapproved infection as these symptoms mimic other common viral byallauthors. infections. Locally, rubella antibody test is not performed Acknowledgements inwomenforconfirmationofthediseaseasthediseaseis TheauthorswishtothanktheDirectorGeneralofHealth,Malaysia,for mildandself-limiting,anddoesnotrequirenotification. permissiontopublishthiswork.Wewouldalsoliketothankthestaffsatthe clinicfortheirsupportandthepatientswhotookpartinthisstudy. Inthisstudy,weusedanti-rubellaIgGlevelof10IU/mL asthecut-offpointforprotectionagainstrubellainfection. Authordetails This was the recommended level based on earlier epi- 1DepartmentofFamilyMedicine,FacultyofMedicineandHealthSciences, UniversitiPutraMalaysia,Serdang,Selangor43400,Malaysia.2Departmentof demiological studies when rubella infection was prevalent FamilyMedicine,FacultyofMedicine,UniversitiKebangsaanMalaysia,Jalan [30]. Although a recent study on the sero-prevalence of YaacobLatifCheras,KualaLumpur56000,Malaysia.3DepartmentofPrimary rubella antibody over time has suggested that the cut-off CareMedicine,FacultyofMedicine,UniversityofMalaya,KualaLumpur 50603,Malaysia. level of anti-rubella IgG for protection against infection should be adjusted according to an immunized antenatal Received:6June2012Accepted:25January2013 population rather than following the older recommenda- Published:31January2013 tion[30],welackedlocalserologicaldataregardingrubella References antibody level for women of childbearing age to ascertain 1. CooperLZ:Thehistoryandmedicalconsequencesofrubella.RevInfect the level of protection for rubella infection. Until such Dis1985,7(Suppl1):S2–S10. data becomes available, the cut-off level of 10 IU/mL 2. MillerE,Cradock-WatsonJE,PollockTM:Consequencesofconfirmed maternalrubellaatsuccessivestagesofpregnancy.Lancet1982, wouldbeused. 2:781–784. The strength of this study was the recruitment of par- 3. GreenRH,BalsamoMR,GilesJP,KrugmanS,MirickGS:Studiesofthe ticipants from a public funded health clinic, which pro- naturalhistoryandpreventionofrubella.AmJDisChild1965, 110:348–365. vided most of the antenatal care for pregnant women in 4. CuttsFT,VynnyckyE:Modellingtheincidenceofcongenitalrubella the country. The study is limited by its cross-sectional syndromeindevelopingcountries.IntJEpidemiol1999,28:1176–1184. Cheongetal.BMCFamilyPractice2013,14:19 Page5of5 http://www.biomedcentral.com/1471-2296/14/19 5. RobertsonSE,FeatherstoneDA,Gacic-DoboM,HershBS:Rubellaand 28. DavidkinI,PeltolaH,LeinikkiP,ValleM:Durationofrubellaimmunity congenitalrubellasyndrome:globalupdate.RevPanamSaludPublica inducedbytwo-dosemeasles,mumpsandrubella(MMR)vaccination.A 2003,14:306–315. 15-yearfollow-upinFinland.Vaccine2000,18:3106–3112. 6. ChuaKB,LamSK,HooiPS,ChuaBH,LimCT:Retrospectivereviewof 29. AsahiT,UedaK,HidakaY,MiyazakiC,TanakaY,NishimaS:Twenty-three-year serologicrubellaactivityinUniversityHospitalKualaLumpur. follow-upstudyofrubellaantibodiesafterimmunizationinaclosed MedJMalaysia2000,55:299–303. population,andserologicalresponsetorevaccination. 7. OoiHL,CheongSM,YogesweryS,NorizahI,ZuridahH,KumarasamyV, Vaccine1997,15:1791–1795. ChuaKB:Congenitalrubellasyndromewithpositiveserologyandvirus 30. ByrneL,BrantL,ReynoldsC,RamsayM:SeroprevalenceoflowrubellaIgG isolation.MedJMalaysia2006,61:248–250. antibodylevelsamongantenatalwomeninEnglandtestedbyNHS 8. KandasamyY,SomasundramP:AreviewofOrangAslinewborns bloodandtransplant:2004–2009.Isrubellasusceptibilityincreasing? admittedtoaneonatalunitinaMalaysiangeneralhospital.Singapore Vaccine2012,30:161–167. MedJ2007,48:926–928. 9. CountriesusingrubellavaccineinNationalImmunisationSchedule.WHO/IVB doi:10.1186/1471-2296-14-19 database:2010.http://www.who.int/immunization_monitoring/diseases/ Citethisarticleas:Cheongetal.:Howusefulisahistoryofrubella Rubella_map_schedule.jpg vaccinationfordeterminationofdiseasesusceptibility?Across-sectional 10. MinistryofHealth(MOH)annualReport.Malaysia:MinistryofHealth; studyatapublicfundedhealthclinicinMalaysia.BMCFamilyPractice 1988:36–39. 201314:19. 11. Planofaction,revisednationalimmunizationprogrammeforchildrenwitha specialfocusonHibandMMRimmunization.Malaysia:DivisionofFamily HealthDevelopment,MinistryofHealth;2002. 12. HealthFacts:Ministryofhealth.Malaysia;2012,2012:6. 13. YusofABM,SelvanesanS,NorizahI,ZuridahH,KumarasamyV,MariamM, ChuaKB:Rubellaoutbreakamongstresidentialstudentsinamilitary vocationalschoolofMalaysia.MedJMalaysia2006,61:296–301. 14. SekawiZ,MuizatulWMN,MarlynM,JamilMAY,IlinaI:Rubellavaccination programmeinMalaysia:analysisofaseroprevalencestudyinan antenatalclinic.MedJMalaysia2005,60:345–348. 15. CheongATKhooEM:Prevalenceofrubellasusceptibilityamong pregnantmothersinacommunity-basedantenatalclinicinMalaysia:a cross-sectionalstudy.AsiaPacJPublicHealth2008,20:340–346. 16. NardoneA,TischerA,AndrewsN,BackhouseJ,TheetenH,GatchevaN, ZarvouM,KrizB,PebodyRG,BarthaK,O’FlanaganD,CohenD,DuksA, GriskeviciusA,MossongJ,BarbaraC,PistolA,SlacikováM,ProsencK, JohansenK,MillerE:Comparisonofrubellaseroepidemiologyin17 countries:progresstowardsinternationaldiseasecontroltargets.Bull WorldHealthOrgan2008,86:118–125. 17. OkudaM,YamanakaM,TakahashiT,IshikawaH,EndohM,HiraharaF: Positiveratesforrubellaantibodyinpregnantwomenandbenefitof post-partumvaccinationinaJapaneseperinatalcenter.JObstetGynaecol Res2008,34:168–173. 18. AngLW,ChuaLT,JamesL,GohKT:Epidemiologicalsurveillanceand controlofrubellainSingapore,1991–2007.AnnAcadMedSingap2010, 39:95–101. 19. BoonruangS,BuppasiriP:Rubellaantibodiesinnormalpregnantwomen atSrinagarindHospital,KhonKaen.Thailand.MedAssocThai2005, 88:455–459. 20. WangI-J,HuangL-M,ChenH-H,HwangK-C,ChenC-J:Seroprevalenceof rubellainfectionafternationalimmunizationprograminTaiwan: vaccinationstatusandimmigrationimpact.JMedVirol2007,79:97–103. 21. PalihawadanaP,WickremasingheAR,PereraJ:Seroprevalenceofrubella antibodiesamongpregnantfemalesinSriLanka.SoutheastAsianJTrop MedPublicHealth2003,34:398–404. 22. RobertsonSE,CuttsFT,SamuelR,Diaz-OrtegaJL:Controlofrubellaand congenitalrubellasyndrome(CRS)indevelopingcountries,part2: vaccinationagainstrubella.BullWorldHealthOrgan1997,75:69–80. 23. CheffinsT,ChanA,KeaneRJ,HaanEA,HallR:Theimpactofrubella immunisationontheincidenceofrubella,congenitalrubellasyndrome andrubella-relatedterminationsofpregnancyinSouthAustralia. Submit your next manuscript to BioMed Central BrJObstetGynaecol1998,105:998–1004. and take full advantage of: 24. GiraudonI,FordeJ,MaguireH,ArnoldJ,PermallooN:Antenatalscreening andprevalenceofinfection:surveillanceinLondon,2000–2007. EuroSurveill2009,14:8–12. • Convenient online submission 25. PopulationofhousingcensusofMalaysia.Populationdistributionandbasic • Thorough peer review demographiccharacteristic2010.Malaysia:DepartmentofStatistic;2010. • No space constraints or color figure charges 26. GeneralreportofthepopulationandhousingcensusofMalaysia2000. • Immediate publication on acceptance Malaysia:DepartmentofStatistic;2000. 27. WysokiñskaT,JanaszekW,BucholcB,GorskaP,GniadekG,SlusarczykJ, • Inclusion in PubMed, CAS, Scopus and Google Scholar RawiczM:Theprevalenceofanti-rubellaantibodiesinwomenof • Research which is freely available for redistribution childbearingageinPoland.Vaccine2004,22:1899–1902. Submit your manuscript at www.biomedcentral.com/submit