Are skilled birth attendants really skilled? A measurement method, some disturbing results and a potential way forward

Transcription

1 A mesurement method, some disturbing results nd potentil wy forwrd Steven A Hrvey, Yudy Crl Wong Blndón, b Affette McCw-Binns, c Ivette Sndino, d Luis Urbin, b Césr Rodríguez, b Ivonne Gómez, b Ptricio Aybc, e Sbou Djibrin f & the Nicrgun mternl nd neontl helth qulity improvement group Objective Delivery by skilled birth ttendnt (SBA) serves s n indictor of progress towrds reducing mternl mortlity worldwide the fifth Millennium Development Gol. Though WHO trcks the proportion of women delivered by SBAs, we know little bout their competence to mnge common life-thretening obstetric complictions. We ssessed SBA competence in five high mternl mortlity settings s bsis for inititing qulity improvement. Methods The WHO Integrted Mngement of Pregnncy nd Childbirth (IMPAC) guidelines served s our competency stndrd. Evlution included written knowledge test, prtogrph (used to record ll observtions of womn in lbour) cse studies nd ssessment of procedures demonstrted on ntomicl models t five skills sttions. We tested purposive smple of 166 SBAs in Benin, Ecudor, Jmic nd Rwnd (Phse I). These initil results were used to refine the instruments, which were then used to evlute 1358 SBAs throughout Nicrgu (Phse II). Findings On verge, Phse I prticipnts were correct for 56% of the knowledge questions nd 48% of the skills steps. Phse II prticipnts were correct for 62% of the knowledge questions. Their verge skills scores by re were: ctive mngement of the third stge of lbour 46%; mnul removl of plcent 52%; bimnul uterine compression 46%; immedite newborn cre 71%; nd neontl resuscittion 55%. Conclusion There is wide gp between current evidence-bsed stndrds nd provider competence to mnge selected obstetric nd neontl complictions. We discuss the significnce of tht gp, suggest pproches to close it nd describe briefly current efforts to do so in Ecudor, Nicrgu nd Niger. Bulletin of the World Helth Orgniztion 2007;85: الرتجمة العربية لهذه الخالصة يف نهاية النص الكامل لهذه املقالة. espñol. Une trduction en frnçis de ce résumé figure à l fin de l rticle. Al finl del rtículo se fcilit un trducción l Introduction Bckground Ech yer obstetric complictions kill over women worldwide. 1,2 Skilled ttendnce during lbour, delivery nd in the erly postprtum period could prevent mny of these deths, though estblishing cusl link between skilled ttendnce nd mternl survivl remins problemtic. 2 6 Still, the proportion of deliveries ssisted by skilled birth ttendnt (SBA) hs become n indictor for mesuring mternl mortlity reduction, including the 75% reduction clled for by the fifth Millennium Development Gol (MDG-5). 7,8 WHO defines n SBA s someone trined to proficiency in the skills needed to mnge norml (uncomplicted) pregnncies, childbirth nd the immedite postntl period, nd in the identifiction, mngement nd referrl of complictions in women nd newborns. 9 WHO uses household survey dt to estimte the percentge of women delivered by n SBA in ech country. 10 (An explntion of the methodology for determining the percentge of women delivered by SBAs is vilble t: ttendnt.html#methodology. WHO s most recent estimtes of the proportion of women delivered by SBAs re vilble t: Typiclly, these surveys sk women wht type of helth professionl, if ny, ssisted t their most recent deliveries Most surveys report results by cdre: doctor, midwife, nurse, trditionl birth ttendnt, reltive nd other. Some ctegorize doctors, midwives nd nurses s helth personnel to distinguish them from untrined ttendnts. Since surveys mke no clim to scertin provider skill, using survey dt to estimte the proportion of SBA-ssisted deliveries ssumes tht ll helth personnel qulify s SBAs. 15 (In this pper, consistent with the Interntionl Confedertion of Midwives Qulity Assurnce Project, University Reserch Co., LLC, 7200 Wisconsin Avenue, Suite 600, Bethesd, MD 20814, USA. Correspondence to Steven A Hrvey (e-mil: shrvey@urc-chs.com). b Qulity Assurnce Project, University Reserch Co., LLC, Mngu, Nicrgu. c Deprtment of Community Helth nd Psychitry, University of the West Indies, Mon, Jmic. d United Ntions Children s Fund (UNICEF), Mngu, Nicrgu. e Qulity Assurnce Project, University Reserch Co., LLC, Quito, Ecudor. f Qulity Assurnce Project, University Reserch Co., LLC, Nimey, Niger. doi: /BLT (Submitted: 2 November 2006 Finl revised version received: 10 Mrch 2007 Accepted: 14 Mrch 2007) Bulletin of the World Helth Orgniztion October 2007, 85 (10) 783

2 Reserch Steven A Hrvey et l. definition, the term midwife or nursemidwife mens provider who hs grduted from certified or ccredited midwifery trining course in the country of prctice. These progrmmes vry in content nd durtion nd my or my not include medicl or nursing trining beyond midwifery skills. Nurse refers to provider who hs completed nursing trining tht is not prt of certified or ccredited midwifery progrmme, even when midwifery functions re performed.) 15 But re skilled birth ttendnts relly skilled? Do the helth personnel enumerted by household surveys fit WHO s definition? This question ws the focus of our two-phse study. In Phse I, we developed nd piloted evlution instruments, then crried out smll-scle competency ssessments in four countries. In Phse II, shortcomings identified in Phse I were corrected nd the revised instruments used to conduct lrger-scle ssessment. The results contribute to ongoing competencyimprovement efforts. Methods We evluted 166 helth providers in Benin, Ecudor, Jmic nd Rwnd during Phse I. In Phse II, we tested 1358 Nicrgun providers. In ech country, our im ws to evlute the professionls responsible for most deliveries those most likely to pper in WHO estimtes. Like WHO, we excluded trditionl birth ttendnts. Tble 1 summrizes mternl helth chrcteristics by country. Instrument development We defined competence s possessing skills nd knowledge sufficient to comply with predefined clinicl stndrds. In multi-country setting, this presupposes stndrds ccepted by ll providers nd sites. Since no such stndrds exist, we mesured competence ginst WHO s Integrted Mngement of Pregnncy nd Childbirth (IMPAC) guidelines, n pproprite evidencebsed yrdstick. 16 The three leding direct cuses of mternl deth re hemorrhge, preeclmpsi nd eclmpsi, nd sepsis. 17,18 We designed instruments to test competence relted to prevention, dignosis nd mngement of these complictions, plus mngement of uncomplicted lbour nd delivery. We evluted knowledge with written test nd skills with ntomicl models. We ruled out direct clinicl observtion for three resons: Complictions occur in pproximtely 15% of pregnncies nd cnnot be predicted. 19,20 It would be prohibitively time-consuming nd expensive to observe ech prticipnt mnging ech compliction. Ethics would compel intervention if cliniclly experienced observer observed sub-stndrd cre. In clinicl settings, mny fctors ffect performnce. Without essentil drugs, equipment or supplies, highly competent provider might perform poorly. With competent ssistnts nd ll necessities t hnd, mrginl provider might perform well. Since environment vries by fcility, we elected to test competency in nonclinicl venues where we could control for environmentl fctors. The initil knowledge test ws modelled on MotherCre, nd Mternl nd Neontl Helth Progrm (MNH) mterils. 21,22 Additionl sources included IMPAC guidelines, prticipting countries norms, nd the Sfe Motherhood Inter-Agency Group s SBA competencies. 5,16,23,24 We dpted two MNH cse studies in order to evlute prtogrph skills. For initil skills testing, we dpted MNH structured observtion checklists for neontl resuscittion, mnul removl of the plcent nd bimnul uterine compression. 21 We developed n intrvenous (IV) insertion skills checklist. Senior clinicins used these checklists to ssess ech prticipnt s skill. To stndrdize observtion criteri, we trined clinicl observers using the following regimen: 1. The country study coordintor reviewed ech checklist line-by-line with observers, clrifying ny perceived mbiguities. 2. One observer performed the first procedure on n ntomicl model. All others used the first checklist to rte the skill demonstrted in this mock evlution. 3. Observers then compred checklists step-by-step to resolve differences. The coordintor served s finl rbiter. 4. The group repeted steps 2 nd 3, with different observers role-plying the evluee, until 100% greement ws chieved. Typiclly, observers chieved 100% consensus fter four to five prctice observtions. The principl investigtor (Steven Hrvey) supervised trining in Benin, Ecudor nd Nicrgu. Coinvestigtors supervised trining in Jmic (Affette McCw-Binns) nd Rwnd (Sbou Djibrin). We piloted the instruments in Ecudor in November 2001, then conducted smll-scle evlutions in the Phse I countries from Mrch to June Phse I prticipnts needed much longer thn nticipted to complete the evlution. Mny struggled with complex cse-bsed questions in the knowledge test. For Phse II, we simplified the test lnguge, eliminted cse-bsed questions nd dopted formt tht ws ll multiple-choice. We dded mteril from the Americn College of Nurse-Midwives (ACNM) life-sving skills curriculum nd JHPIEGO guidelines for the ssessment of skilled providers. 25,26 On expert dvice, we replced IV insertion, nd mouth-tomouth nd nose resuscittion checklists with one focused on ctive mngement of the third stge of lbour (AMTSL) nd nother on immedite newborn cre. In both phses, we relied on multiple rounds of expert review to estblish construct nd content vlidity for our instruments. Reviewers included obstetricins/gynecologists, peditricins nd midwives t the interntionl level nd in ech study country. To ensure ccurte trnsltion, we reviewed ech instrument line-by-line with locl clinicins before ech evlution. To ssess obstetric skills, we used the Gumrd Advnced Childbirth Simultor S500; for neontl skills we used the Simulids Sni-Bby CPR mnnequin in Phse I nd the Gumrd S320 Airwy Triner Newborn in Phse II. Finlly, we reduced prtogrph evlution to one cse. Study sites nd smple chrcteristics In Phse I, we selected prticipnts purposively to represent the full rnge of fcilities where women give birth nd the cdre(s) primrily responsible for ttending fcility-bsed births in ech country. We included t lest one ntionl-level tertiry cre fcility, t lest two district-level hospitls, nd mix of rurl nd urbn fcilities. All prticipnts completed ll evlution ctivities. For Phse II, we evluted helth personnel in 20 hospitls, t lest 1 in 784 Bulletin of the World Helth Orgniztion October 2007, 85 (10)

3 Steven A Hrvey et l. Reserch Tble 1. Mternl helth chrcteristics of study countries Country WHO-estimted MMR % births occurring in helth fcility b % births ssisted by skilled ttendnt b Attendnt most likely to ssist t in-fcility delivery Benin Midwife Ecudor c 74.1 c Medicl resident or intern Jmic Midwives (85%), doctors (15%) Nicrgu Medicl resident or intern Rwnd Nurse, nursing ssistnt MMR, mternl mortlity rtio. MMR is defined s the number of mternl deths per live births. Source: Mternl mortlity in 2000: estimtes developed by WHO, UNICEF nd UNFPA. Genev: WHO; b Sttistics from: Benin 2001 DHS; Ecudor 2004 ENDEMAIN; Jmic Register Generl s Deprtment. Vitl Sttistics Report, 2001, Twickenhm Prk: 2003; Nicrgu 2001 DHS; Rwnd 2000 DHS. c In Ecudor, the birth percentges represent two different time intervls. Percentge of births occurring in helth fcility tkes ccount of ll births from Jnury 2002 to the time of the survey (n = 2798). Percentge of births ssisted by skilled ttendnt tkes ccount of ll births from Jnury 1999 to the time of the survey (n = 6140). ech of Nicrgu s 17 helth res. This included 3 teching hospitls, 2 mternl nd child hospitls, nd 15 district hospitls. We lso evluted personnel from 43 primry helth centres, t lest 1 from every helth re except Crzo, Mngu, Msy nd Rivs. Phse II selection criteri required ll prticipting fcilities to be qulified bsic or comprehensive emergency obstetric cre (BEOC or CEOC) fcilities nd, in the cse of primry fcilities, locted in helth re prticipting in some obstetric cre improvement inititive. 19 In the South Atlntic Autonomous Region (RAAS), we enrolled ll BEOC fcilities. Elsewhere fcilities were selected by convenience, principlly relted to physicl ccessibility. UNICEF, CARE, the Pn Americn Helth Orgniztion (PAHO), the Qulity Assurnce Project (QAP) nd the Nicrgun Ministry of Helth jointly determined the smpling strtegy. Phse II prticipnts were evluted by function. Obstetricins nd gynecologists, obstetrics nd gynecology residents, nd generl prctitioners in gynecology prticipted only in testing relted to lbour, delivery nd obstetric complictions. In Nicrgu, these cdres were not tested on peditric skills s they work in higherlevel fcilities nd provide no peditric cre. For the sme reson, Nicrgun peditricins, peditric residents nd peditric generl prctitioners prticipted only in ctivities relted to immedite newborn cre nd neontl resuscittion. Hospitl emergency personnel nd nonspecilized medicl nd nursing personnel from peripherl fcilities completed both components (Tble 2, vilble t: volumes/85/10/ /en/index. html). The study ws reviewed for complince with QAP ethics guidelines, designed to weigh potentil risks nd benefits, nd ensure prticipnt confidentility, informed consent nd hostcountry ethics review. Ministry of helth ethics committees or their equivlent pproved the protocol in ech country. We obtined verbl consent from ll study prticipnts. To protect confidentility, we observed the following mesures: Prticipnts were ssigned study numbers; no nmes were recorded. Evlutors were prohibited from observing ny prticipnt with whom they hd supervisory reltionship. Dt were stored off-site in loction ccessible only to study tem members. Helth fcility mngers nd supervisors hd no ccess. Results were reported in ggregte, minly by district nd cdre. Results were not disggregted by fcility when this might hve exposed prticipnt s identity. Results Tble 3 summrizes Phse I results, which re reported in detil elsewhere. 27 The text describes Phse II results, presented by provider ctegory in Tble 4. Knowledge test Grouping ll provider ctegories nd ll knowledge test versions (obstetric, peditric, generl), prticipnts were correct on 62% of test questions (Tble 4). By topic, results rnged from 80% correct for hemorrhge during pregnncy to 16% correct for infection prevention. By cdre, doctors were correct on 72% of questions overll, medicl students 67%, professionl nurses 57% nd uxiliry nurses 51%. These differences were sttisticlly significnt (nlysis of vrince, ANOVA F = 265, P < 0.001), s were individul between-group differences (Bonferroni post hoc P in ll cses). Only five nurse-midwives (less thn 0.5% of the smple) prticipted in Phse II s Nicrgu hs stopped trining this cdre. Prtogrph cse study Doctors nd medicl students performed similrly on the prtogrph test, nswering 65 70% of questions correctly. Professionl nd uxiliry nurses scores were lower: 42% nd 33% respectively for written questions, nd 19% nd 10% for grphing. Differences were sttisticlly significnt (ANOVA F = 199, P < for combined written nd grphing scores). Doctors nd medicl students hd sttisticlly identicl scores (Bonferroni post hoc P = 1). Differences between ll other groups on ll test components were significnt (P 0.001). Skills sttions AMTSL scores using the childbirth simultor rnged from 53% for doctors to 36% for uxiliry nurses. Doctors nd medicl students scored virtully identiclly (ANOVA Bonferroni post hoc P = 1). Nurses scored significntly lower thn doctors (Bonferroni P < 0.001) nd mrginlly lower thn medicl students (P = 0.09). Score differences between Bulletin of the World Helth Orgniztion October 2007, 85 (10) 785

4 Reserch Steven A Hrvey et l. Tble 3. Men Phse I competency scores (%) by country, provider cdre nd topic Topic Competency scores by country Competency scores by provider cdre Benin Ecudor Jmic Rwnd Doctors b Midwives b All providers b Knowledge test (n = 43) (n = 25) (n = 64) (n = 34) (n = 25) (n = 54) (n = 166) Infection prevention 60.5% 41.1% 48.2% 34.9% 47.4% 54.5% 47.6% Uncomplicted lbour nd delivery 52.6% 64.8% 59.5% 56.2% * 62.3% * 55.2% 57.9% Immedite newborn cre 49.1% 62.6% 65.8% 44.1% 57.8% 52.7% 56.5% Postprtum hemorrhge 63.4% 68.0% 64.2% 54.2% * 72.7% * 63.1% 62.5% Pregnncy-induced hypertension 54.8% 78.2% 68.9% 51.6% ** 74.7% ** 58.2% 63.1% Sepsis 38.0% 53.3% 47.9% 39.2% * 56.7% * 42.0% 44.4% Active mngement 3rd stge lbour 10.5% 14.0% 39.1% 7.4% 24.0% 16.7% 21.4% Totl knowledge test score 52.5% 61.8% 59.8% 47.9% * 61.9% * 54.2% 55.8% Prtogrph test (n = 42) (n = 24) (n = 27) (n = 20) (n = 45) (n = 93) Written questions only 67.4% 54.5% NA 50.1% 67.2% 61.9% 59.0% Grphing only 63.5% 33.0% NA 48.5% 59.6% 54.8% 51.3% Totl prtogrph test score 66.7% 50.7% NA 49.8% 65.9% 60.7% 57.7% Skills evlution (n = 42) (n = 25) (n = 62) (n = 19) (n = 22) (n = 49) (n = 148) Mnul removl of plcent 64.2% 46.8% 20.1% 51.1% 57.1% 56.8% 41.1% Bimnul uterine compression 7.9% 28.6% 22.2% 40.2% ** 42.8% ** 12.4% 21.5% Neontl resuscittion with mbu bg 58.6% 39.5% 67.9% 43.3% 52.2% 59.0% 57.3% Neontl resuscittion mouth-to-mouth 59.7% 26.0% 69.1% 44.7% * 45.0% * 57.5% 56.0% nd nose Infection prevention 55.2% 38.0% 43.2% 53.7% 49.4% 52.0% 47.1% IV insertion 79.4% 66.7% 60.2% 73.7% 74.5% 77.4% 68.5% Totl skills score 54.4% 41.8% 46.1% 50.2% 54.3% 52.7% 48.2% IV, intrvenous; NA, non-pplicble. Section totls do not represent cumultive men scores from ech section (knowledge test, prtogrph test, etc.) becuse their respective sub-sections contined different numbers of questions. b Scores for doctors nd midwives re reported seprtely becuse these were the only two cdres present in ll four Phse I countries. The men scores for doctors, midwives nd ll providers represent pooled individul scores. Scores re not weighted by country. Sttisticl significnce for the difference in scores between doctors nd midwives estblished by t-test: * P < 0.05; ** P < professionl nd uxiliry nurses were not significnt (P = 0.49). Doctors correctly performed 53% of the steps for mnul removl of the plcent; medicl students 45% (t = 2.6, P = 0.009). Bimnul uterine compression scores were lower: 48% for doctors versus 37% for medicl students (t = 3.2, P = 0.001). Since Nicrgun norms prohibit nurses nd uxiliry nurses from performing either procedure, they did not prticipte in these evlutions. Scores were higher nd less vried for immedite newborn cre, rnging from 76% for doctors to 64% for uxiliry nurses. There ws no score difference between doctors nd medicl students or between professionl nd uxiliry nurses (Bonferroni P = 1, both cses), but professionl nd uxiliry nurses scored significntly lower thn doctors nd medicl students (P < 0.05, ll cses). Neontl resuscittion scores followed the sme pttern, rnging from 62% for doctors to 45% for uxiliry nurses. With the exception of infectionprevention knowledge, Phse II scores correlte highly with professionl profile: doctors scored highest, followed by medicl students, professionl nurses nd uxiliry nurses. Professionl nurses outscored medicl students (but not doctors) on infection prevention, though no group exceeded 21%. The verge score ws 16%. Discussion While our results show significnt vritions in competency between different evlution components nd different cdres, the generlly low scores re troubling. Different countries nd cdres show different strengths nd weknesses, but severl ptterns emerge: Mny prticipnts scored poorly on bsic questions relted to infection prevention (hnd-wshing, proper hndling of contminted instruments, proper disposl of medicl wste). In Nicrgu, intrmusculr (IM) oxytocin use immeditely fter birth becme provisionl stndrd in 2003, nd AMTSL knowledge is high. In other countries, AMTSL ws not routine t the time of the study. This my explin why mny providers could not identify its components (IM oxytocin immeditely fter delivery of the foetus, controlled cord trction, uterine mssge) nd did not know tht it should be prctised universlly Mny providers did not recognize the distolic blood pressure level indictive of severe pre-eclmpsi or identify the use of mgnesium sulfte nd rpid termintion of the pregnncy s the preferred mngement strtegies for this condition. Ability to correctly use nd interpret the prtogrph ws low. Skills scores generlly were lower thn knowledge scores. For exmple: the men AMTSL knowledge score 786 Bulletin of the World Helth Orgniztion October 2007, 85 (10)

5 Steven A Hrvey et l. Reserch Tble 4. Men Phse II competency scores (%) by provider cdre nd topic Topic Doctors (n) Medicl students (n) Professionl nurses b (n) Auxiliry nurses (n) Totl (n) Knowledge test Infection prevention 21.4% (506) 12.2% (148) 16.2% (339) 9.6% (365) ** 15.9% (1358) Uncomplicted lbour nd delivery 73.1% (357) 71.7% (117) 52.6% (256) 47.5% (307) ** 60.3% (1037) Immedite newborn cre 65.3% (347) 60.8% (120) 44.8% (243) 39.0% (313) ** 51.9% (1023) Hemorrhge during pregnncy 88.4% (357) 86.8% (117) 75.9% (256) 70.3% (307) ** 79.8% (1037) Postprtum hemorrhge 81.1% (357) 75.9% (117) 60.4% (256) 57.8% (307) ** 68.5% (1037) Pregnncy-induced hypertension 60.5% (357) 57.2% (117) 45.5% (256) 43.0% (307) ** 51.2% (1037) Sepsis 76.3% (357) 73.4% (117) 61.5% (256) 53.0% (307) ** 65.4% (1037) Active mngement of third stge lbour 84.1% (357) 82.6% (117) 67.8% (256) 63.4% (307) ** 73.8% (1037) Totl knowledge score 71.5% (506) 67.5% (148) 56.7% (339) 51.4% (365) 62.0% (1358) Prtogrph test Written questions only 67.0% (343) 65.8% (116) 41.9% (118) 32.9% (89) ** 57.8% (666) Grphing only 67.3% (343) 69.1% (116) 19.3% (118) 9.9% (89) ** 51.5% (666) Totl prtogrph test score 67.1% (343) 66.6% (116) 36.3% (118) 27.1% (89) 56.2% (666) Skills evlution Active mngement of third stge lbour 52.5% (170) 48.7% (41) 40.9% (93) 36.4% (81) ** 45.9% (385) Mnul extrction of the plcent 53.1% (170) 45.0% (41) ** 51.5% (211) Bimnul uterine compression 48.4% (170) 37.2% (41) ** 46.2% (211) Immedite newborn cre 76.4% (159) 76.8% (40) 67.4% (86) 63.6% (84) ** 71.5% (369) Neontl resuscittion with mbu bg 61.6% (159) 57.7% (40) 50.1% (86) 45.0% (84) ** 54.7% (369) Sttisticl significnce for the difference in scores between provider cdres determined by ANOVA: * P < 0.05; ** P < b As in Tble 2, five nurse-midwives who prticipted in Phse II were included in the ctegory professionl nurses. in Nicrgu ws 74%; the men skills score ws 46%. Similrly, the men Phse I knowledge score on mngement of postprtum hemorrhge ws 63%; the skills scores for mnul removl of the plcent nd bimnul uterine compression bsic evidence-bsed interventions to control postprtum hemorrhge were 41% nd 22% respectively. This suggests tht knowledge of procedure is no gurntee tht it cn be performed correctly. Though the pttern ws less cler in Phse I, Phse II doctors nd medicl students generlly scored higher thn midwives, professionl nurses nd uxiliry nurses. Tble 5 (vilble t: volumes/85/10/ /en/index. html) presents the pir-wise score differences for ech ssessment re by provider type. While predictble, this outcome my not be the most desirble for birthing women. In isolted rurl settings where technology is limited nd the nerest referrl fcility hours wy, womn is much more likely to be ttended by midwife or nurse thn doctor, so the bsic life-sving skills of these cdres my be crucil when complictions rise. Since Nicrgu no longer trins midwives nd few remin in prctice, birthing womn there is most likely to be ttended by professionl or uxiliry nurse in settings where no doctor is vilble. Even when doctors nd technology re more ccessible, nursing stff often ttend most deliveries nd perform routine tsks such s completing the prtogrph nd monitoring for postprtum hemorrhge. Limittions Since the smples were non-rndom, we cnnot be certin tht they represent ll professionl helth workers who ttend births in the study countries. Further, we could not control for mny potentil confounders: differences in pre-service trining, popultion helth sttus, helth system structure, ntionl norms nd prctices, nd inter-observer greement between countries. The nlysis here is bivrite. A multivrite nlysis tht controlled for helth fcility type, rurl versus urbn setting nd work experience might yield different results. However, none of these fctors seems likely to produce downwrd bis in scores. If nything, the inclusion of tertiry cre nd teching hospitl personnel might inflte scores since these providers perform procedures more frequently nd hve ccess to better resources thn rurl clinicins. The higher Nicrgun scores my be ttributble to severl fctors. Phse II instruments were shorter nd simpler. In contrst to Phse I, we evluted Phse II prticipnts only on functions they ctully perform. Finlly, Nicrgun prticipnts were ll ssessed in Spnish, their ntive lnguge. Beninese nd Rwndn prticipnts were ssessed in French second lnguge for some. But mesurement error lone is unlikely to explin our results. In spite of these differences, there were remrkble consistencies between Phse I nd Phse II results, suggesting serious cuse for concern cross rnge of settings. Wht is the significnce of this gp between evidence-bsed stndrds nd provider competence? One perspective holds tht proposed interntionl stndrds re simple nd strightforwrd; nyone eligible to be designted s n SBA should be competent to implement them. It sends the wrong messge if some providers re held to these norms but others re excused. This rgument hs prticulr logic in rurl res where highly trined providers often re unvilble. Another perspective holds tht Bulletin of the World Helth Orgniztion October 2007, 85 (10) 787

6 Reserch Steven A Hrvey et l. there is fr from universl greement on these best prctices. Some ntionl norms contrdict some IMPAC guidelines or even prohibit their use. In urbn res where provider functions re more specilized, perhps not everyone needs to be competent t everything. Thus, judging provider competency ginst stndrds to which their own helth systems do not subscribe csts them in n unfirly negtive light. These conflicting perspectives underscore the need to strengthen consensus on which prctices mximize opportune identifiction nd mngement of life-thretening complictions nd how best to implement them. A centrl premise of mternl helth progrmming holds tht delivery by skilled ttendnt nd redy ccess to BEOC or CEOC fcility re fundmentl to reducing mternl deth. 3,19,31,32 In their recent contribution to the Lncet series on mternl survivl, Cmpbell nd Grhm reiterte tht helth centre intrprtum cre is the most promising strtegy for reducing mternl mortlity in time to chieve MDG Sufficient numbers of competent birth ttendnts re essentil to this strtegy. A helth worker shortge is one importnt brrier, but indequte competence mong existing helth workers my be eqully importnt. 8 Our findings pper to confirm this conclusion: womn who delivers t forml helth fcility ssisted by so-clled skilled ttendnt cnnot necessrily ssume she will receive competent cre. A potentil wy forwrd In light of our results, prticipting countries re now tking steps to close the competency gp. QAP, United Ntions Popultion Fund (UNFPA) nd the Ecudorin Ministry of Helth hve temed up to develop nd implement n eight-session trining progrmme focused on improving compliction mngement. By the end of 2006, this group hd trined 81 triners nd 74 clinicins in 5 provinces, plus 12 instructors t Quito s midwifery school. Provincil triners re scheduled to trin lrge number of provincil clinicins in Quito s midwifery school is integrting this progrmme into its pre-service curriculum; other Ecudorin midwifery nd medicl schools re considering its doption. In Nicrgu, similr effort is under wy, coordinted by the Ministry of Helth, QAP, UNICEF, CARE nd PAHO. By December 2006, this inititive hd trined 428 birth ttendnts in 14 of the country s 17 helth res. In Niger, comprble inititive which begn in erly 2006 hd trined 239 providers in 28 of the country s 37 reference fcilities by the yer s end. Correct performnce of AMTSL rose from 25% to 97% in prticipting fcilities, ffecting n verge of 2369 births monthly. Projects in Bngldesh, Benin, Eritre nd Keny hve begun to integrte SBA competency evlution into mternl helth progrmming. QAP s competency ssessment instruments, long with mnul on how to conduct n ssessment, cn be downloded from strtsfemotherhood.html/sbcomp. html. Trining, however, is only one prt of the eqution. No mount of trining will led to more hnd-wshing if helth fcilities lck sop nd wter. Helth personnel cnnot be expected to identify mgnesium sulfte s the drug of choice for pre-eclmpsi nd eclmpsi if it is unvilble s it ws in Benin during this study. Systemic problems require more comprehensive qulityimprovement inititives to ddress drug vilbility nd distribution, equipment supply nd mintennce, ineffective supervision, low morle nd other problems tht ffect helth services in mny high mternl-mortlity settings. However, while work continues to resolve systemic problems, much effort must be directed t rising bsic competencies if helth personnel re to ttin the proficiency nd fulfil the functions nticipted by WHO, the Interntionl Federtion of Gyncology nd Obstetrics (FIGO) nd the Interntionl Confedertion of Midwives (ICM) definition. Only then will SBAs be truly skilled nd their deliveries become n ccurte indictor of progress towrds reducing mternl mortlity. Acknowledgements Reserch ctivities were supported by QAP nd the United Sttes Agency for Interntionl Development (USAID). UNICEF nd CARE provided dditionl funding for ctivities in Nicrgu. PAHO s Nicrgun office contributed to study design. In ddition to the listed uthors, members of the Nicrgun mternl nd neontl helth qulity improvement group include Ivette Aráuz nd Dysi Nvrro (CARE), Wilmer Betet, Roberto Jiménez, Rmiro López nd Auror Velásquez (Nicrgun Ministry of Helth), Alm Fbiol Morles (PAHO), Oscr Nuñez (QAP, Mngu), Cludi Evens (USAID, Mngu) nd Cludi Grnj (UNICEF). We thnk: MoH officils, helth fcility directors, prticipting helth personnel, nd field tem members who mde this study possible in ech country. Also P Annie Clrk nd Deborh Gordis (ACNM), Ptrici Gomez nd Brbr Deller (JHPIEGO), nd Cryl Feldcker for their help in revising nd updting study instruments; Sourou Gbngbde nd Denn Ashley for informtion on trining nd functions of helth personnel in Benin nd Jmic respectively. Finlly, we thnk Beth Goodrich (QAP) for her ssistnce editing the mnuscript. Competing interests: None declred. Résumé Les «ccoucheurs/euses qulifiés/es» ont-ils/elles réellement les compétences pour prtiquer des ccouchements? Présenttion d une méthode de mesure de ces compétences, de certins résultts dérngents et d une voie de progression potentielle Objectif Le tux d ccouchement pr un ccoucheur qulifié sert dns le monde entier d indicteur des progrès rélisés dns l réduction de l mortlité mternelle (cinquième objectif du Millénire pour le développement). Si l OMS suit l proportion de femmes ccouchées pr des ccoucheurs qulifiés, elle sit peu de choses sur l cpcité de ces gents de snté à prendre en chrge des complictions obstétricles courntes potentiellement ftles. Nous vons évlué les compétences de ces ccoucheurs dns cinq étblissements présentnt une forte mortlité mternelle, en tnt que point de déprt pour une méliortion de l qulité des presttions. 788 Bulletin of the World Helth Orgniztion October 2007, 85 (10)

7 Steven A Hrvey et l. Méthodes Les directives de l Prise en chrge intégrée de l grossesse et de l ccouchement de l OMS (PCIGA) nous ont servi de norme pour l évlution des compétences. Celle-ci été rélisée sur l bse d un test de connissnces écrit, d études de cs utilisnt un prtogrphe (ppreil cpble d enregistrer l ensemble des prmètres pour une femme en trvil) et de l démonstrtion pr les prticipnts de différentes opértions sur des modèles ntomiques dns cinq postes d évlution des compétences. Nous vons testé cette évlution sur un échntillon choisi à dessein de 166 ccoucheurs qulifiés du Bénin, de l Equteur, de l Jmïque et du Rwnd (phse I). Les résultts initiux obtenus ont été utilisés pour perfectionner les instruments d évlution, qui ont ensuite été ppliqués à 1358 ccoucheurs qulifiés dns l ensemble du Nicrgu (phse II). Résultts En moyenne, les prticipnts à l phse I ont fourni une Reserch presttion correcte pour 56 % des questions de connissnces et pour 48 % des étpes de compétences. Les prticipnts à l phse II ont répondu correctement à 62 % des questions de connissnces. En moyenne, leurs résultts en mtière de compétences se réprtissient pr domines comme suit. Prise en chrge ctive du troisième stde du trvil : 46 %, retrit mnuel du plcent : 52 %, compression utérine à deux mins : 46 %, soins immédits u nouveu-né : 71 % et rénimtion néontle : 55 %. Conclusion Il existe un lrge écrt entre les exigences de l norme ctuelle reposnt sur une bse fctuelle et l cpcité des prestteurs à prendre en chrge certines complictions obstétricles et néontles. Nous exminons l importnce de cet écrt, proposons des pproches pour le combler et décrivons brièvement les efforts entrepris dns cette voie en Equteur, u Nicrgu et u Niger. Resumen Qué tn clificdos están los sistentes de prterí clificdos? Un método de medición, lgunos resultdos preocupntes y un posible pso delnte Objetivo L tención del prto por personl de slud clificdo Rund (fse I). Con esos resultdos iniciles se refinron los es un indicdor que vlor el progreso relizdo pr reducir l instrumentos, que volvieron ser utilizdos pr evlur 1358 mortlidd mtern nivel mundil, y que se ve reflejdo en el personl clificdo trvés de Nicrgu (fse II). quinto Objetivo de Desrrollo del Milenio. Aunque l OMS sigue Resultdos En promedio, los prticipntes en l fse I contestron de cerc l proporción de mujeres que dn luz tendids por un correctmente el 56% de ls pregunts sobre los conocimientos personl clificdo, es poco lo que sbemos sobre su competenci teóricos. Así mismo, desempeñron correctmente el 48% de pr mnejr complicciones obstétrics comunes potencilmente los psos de ls pruebs práctics. Los prticipntes en l fse II mortles. Evlumos l competenci del personl clificdo en cinco respondieron correctmente l 62% de ls pregunts teórics. Sus entornos de lt mortlidd mtern como punto de prtid pr puntuciones promedis por áre de práctic fueron ls siguientes: empezr mejorr l clidd de l sistenci. mnejo ctivo del tercer período del prto, 46%; extrcción mnul Métodos El grdo de competenci se determinó emplendo como de l plcent, 52%; compresión uterin bimnul, 46%; tención referenci l guí de l OMS, Mnejo integrdo del embrzo y el inmedit l recién ncido, 71%; y renimción neontl, 55%. prto (IMPAC). L evlución incluyó un exmen escrito sobre los Conclusión Existe un brech importnte entre ls norms ctules conocimientos en l mteri, estudios de csos del prtogrm bsds en l evidenci y ls ptitudes del personl pr mnejr (registro de vribles durnte l vigilnci del trbjo de prto) y determinds complicciones obstétrics y neontles. Trs nlizr un evlución de l práctic demostrd con modelos ntómicos ls implicciones de est brech, se propone medids pr corregirlo en cinco tipos de ptitudes. Se nlizó un muestr intenciond y se describe brevemente ls ctividdes que ctulmente se llev de 166 sistentes clificdos en Benin, el Ecudor, Jmic y cbo con ese fin en el Ecudor, Nicrgu y el Níger. ملخص هل تتمت ع املولدات املاهرات فعال باملهارة طريقة للقياس وبعض النتائج املقلقة والخطوات املقبلة املحتملة الغرض: تعد الوالدة تحت إرشاف مولدة ماهرة مؤرشا عىل التقد م الم ح رز تجاه الحد من وفيات األمهات يف جميع أنحاء العامل وهو ما ينص عليه املرمى الخامس من املرامي اإلمنائية لأللفية. وبرغم تقصي منظمة الصحة العاملية لنسبة السيدات الاليت يلدن تحت إرشاف مولدات ماهرات إال أننا ال نعرف إال القليل عن كفاءتهن يف تدبري املضاعفات التوليدية الشائعة املهدي دة للحياة. وقد ق منا يف هذه الدراسة بتقييم كفاءة املولدات املاهرات يف خمسة مواقع ترتفع فيها معدالت وفيات األمهات كأساس للبدء يف تحسني جودة العمل يف هذه املواقع. الطريقة: اعت ربت الدالئل اإلرشادية للتدبري املتكامل للحمل والوالدة الصادرة عن منظمة الصحة العاملية معيار الكفاءة يف هذه الدراسة. وقد اشتمل التقييم عىل اختبار كتايب ملستوى املعارف ومخططا بيانيا للمخاض )يستخدم لتسجيل جميع املالحظات الخاصة باملرأة أثناء املخاض( ودراسات حالة وتقييام لإلجراءات عىل النامذج الترشيحية يف مراكز املهارات الخمسة. وق منا باختبار عيي نة مقصودة قوامها 166 مولدة ماهرة يف بنني واإلكوادور وجامايكا ورواندا )املرحلة األوىل(. واست خدمت هذه النتائج األولية لتنقيح الوثائق واستخدمت بعد ذلك لتقييم 1358 مولدة ماهرة يف جميع أنحاء نيكاراغوا )املرحلة الثانية(. املوجودات: من واقع املتوسطات لوحظ أن إجابات املشاركني يف املرحلة األوىل كانت صحيحة يف %56 من األسئلة املتعلقة باملعارف ويف %48 من األسئلة املتعلقة بخطوات املهارات. وكانت إجابات املشاركني يف املرحلة الثانية صحيحة يف %62 من األسئلة املتعلقة باملعارف. وكانت األحراز املتوسطة للمهارات بحسب املجال عىل النحو التايل: التدبري النشط للمرحلة الثالثة للمخاض 46 % والنزع اليدوي للمشيمة 52 % والضغط بكلتا اليدين عىل الرحم 46 % والرعاية العاجلة للوليد 71 % وإنعاش الوليد %55. االستنتاج: خلصت الدراسة إىل وجود فجوة عميقة بني املعايري الحالية الم س ندة بالبينات وبني كفاءة مقدي مي الخدمات يف تدبري بعض املضاعفات التوليدية واملضاعفات التي تصيب الولدان. ونناقش يف هذه الورقة حجم هذه الفجوة ونقترح أساليب لسد هذه الفجوة ونبنيي بإيجاز الجهود الحالية يف هذا الشأن يف كل من اإلكوادور ونيكاراغوا والنيجر. Bulletin of the World Helth Orgniztion October 2007, 85 (10) 789

8 Reserch References 1. Hill K, AbouZhr C, Wrdlw T. Estimtes of mternl mortlity for Bull World Helth Orgn 2001;79: Mternl mortlity in 2000: estimtes developed by WHO, UNICEF nd UNFPA. Genev: World Helth Orgniztion; Thddeus S, Mine D. Too fr to wlk: mternl mortlity in context. Soc Sci Med 1994;38: Liljestrnd J. Strtegies to reduce mternl mortlity worldwide. Curr Opin Obstet Gynecol 2000;12: Grhm WJ, Bell JS, Bullough CH. Cn skilled ttendnce t delivery reduce mternl mortlity in developing countries? In: De Brouwere V, Vn Lerberghe W, editors. Sfe motherhood strtegies: review of the evidence; Studies in Helth Services Orgnistion nd Policy, 17. Antwerp: ITG Press; pp Donny F. Mternl survivl in developing countries: Wht hs been done, wht cn be chieved in the next decde? Int J Gynecol Obstet 2000;70: AbouZhr C, Wrdlw T. Mternl mortlity t the end of decde: signs of progress? Bull World Helth Orgn 2001;79: Koblinsky M, Mtthews Z, Hussein J, Mvlnkr D, Mridh MK, Anwr I, et l. Going to scle with professionl skilled cre. Lncet 2006;368: Mking pregnncy sfer: the criticl role of the skilled ttendnt: joint sttement by WHO, ICM nd FIGO. Genev: WHO; WHO/UNFPA/UNICEF/World Bnk. Reduction of mternl mortlity: joint sttement. Genev: WHO; Enquête Démogrphique et de Snté u Bénin, Clverton: Institut Ntionl de l Sttistique et de l Anlyse Économique nd ORC Mcro; Encuest Nicrgüense de demogrfí y slud Clverton: Instituto Ncionl de Estdístics y Censos (INEC) y ORC Mcro; Enquête démogrphique et de snté, Rwnd Kigli nd Clverton: Ministère de l Snté, Office Ntionl de l Popultion et ORC Mcro; Ordoñez J, Stupp P, Monteith D, Ruiz A, Goodwin M, McCrcken S, et l. ENDEMAIN 2004 (Encuest demográfic y de slud mtern e infntil): informe finl. Quito: CEPAR; Progress since the World Summit for Children: sttisticl review. New York: UNICEF; September WHO, UNFPA, UNICEF, World Bnk. IMPAC - Mnging complictions in pregnncy nd childbirth: guide for midwives nd doctors. Genev: WHO; Wlsh JA, Feifer CM, Meshm AR, Gertler PJ. Mternl nd perintl helth. In: Jmison DT, Mosley WH, Meshm AR, Bobdill JL, eds. Disese control priorities in developing countries. New York: Oxford University Press; Grhm WJ, Cirns J, Bhttchry S, Bullough CHW, Quyyum Z, Rogo K. Mternl nd perintl conditions. In: Jmison DT, Mosley WH, Meshm AR, Bobdill JL. Disese control priorities in developing countries. 2nd ed. New York: Oxford University Press; pp doi: / /Chpt-26. Steven A Hrvey et l. 19. Mine D, Wrdlw TM, Wrd VM, McCrthy J, Birbum A, Aklin MZ, et l. Guidelines for monitoring the vilbility nd use of obstetric services. 2nd edn. New York: UNICEF, WHO, UNFPA; Sy L, Pttinson RC, Gulmezoglu AM. WHO systemtic review of mternl morbidity nd mortlity: the prevlence of severe cute mternl morbidity (ner miss). Reprod Helth 2004;1: Mternl nd Neontl Helth Progrm. MNH Progrm: evlution of MNH clinicl trining for service providers (DRAFT). Bltimore: JHPIEGO; October McDermott J, Beck D, Buffington ST, Anns J, Suprtikto G, Prenggono D, et l. Two models of in-service trining to improve midwifery skills: how well do they work? J Midwifery Womens Helth 2001;46: Snté fmilile Benin: politique normes et stndrds. Cotonou: Ministère de l Snté Publique du Benin, Direction de l Snté Fmilile; Norms y procedimientos pr l tención de l slud reproductiv. Quito: Ministerio de Slud Públic Repúblic del Ecudor; Mrshll MA, Tebben Buffington S. Life-sving skills mnul for midwives, 3rd ed. Silver Spring: Americn College of Nurse-Midwives; Mternl nd neontl helth progrm. Guidelines for ssessment of skilled providers fter trining in mternl nd newborn helthcre. Bltimore: JHPIEGO; Hrvey SA, Aybc P, Bucgu M, Djibrin S, Edson WN, Gbngbde S, et l. Skilled birth ttendnt competence: n initil ssessment in four countries, nd implictions for the Sfe Motherhood movement. Int J Gynecol Obstet 2004;87: Festin MR, Lumbignon P, Tolos JE, Finney KA, B-Thike K, Chipto T, et l. Interntionl survey on vritions in prctice of the mngement of the third stge of lbour. Bull World Helth Orgn 2003;81: Neilson JP. Evidence-bsed intrprtum cre: evidence from the Cochrne Librry. Int J Gynecol Obstet 1998;63:S Shne B. Preventing postprtum hemorrhge: mnging the third stge of lbor. Mternl & neontl helth project Bltimore Outlook 2001; McCrthy J, Mine D. A frmework for nlyzing the determinnts of mternl mortlity. Stud Fm Plnn 1992;23: Mine D, Rosenfield A. The Sfe Motherhood Inititive: why hs it stlled? Am J Public Helth 1999;89: Cmpbell OM, Grhm WJ. Strtegies for reducing mternl mortlity: getting on with wht works. Lncet 2006;368: Bulletin of the World Helth Orgniztion October 2007, 85 (10)

9 Steven A Hrvey et l. Reserch Tble 2. Number of Phse II prticipnts by evlution component nd professionl cdre Professionl cdre Totl Knowledge component Obstetric/ gynecologic test Peditric test Generl test Prtogrph exercise Active mngement Mnul removl Skills component Bimnul compression Immedite newborn cre Neontl resuscittion Doctors Obstetricin/gynecologist Peditricin Obstetric/gynecologic resident Peditric resident Generl prctitioner gynecologist Generl prctitioner peditricin Generl prctitioner ER Medicl students Socil service physicin Medicl intern Professionl nurses Nurse-midwife Mternl nd child helth nurse Professionl nurse Auxiliry nurses Technicl nurse Auxiliry nurse Totl ER, emergency room. Nurse-midwives hve received both nursing nd midwifery trining, nd therefore theoreticlly belong in seprte ctegory from nurses. However, Nicrgu no longer trins this cdre nd very few remin in prctice. Since our smple contined only five nurse-midwives (< 0.5% of totl), we grouped them with professionl nurses. Bulletin of the World Helth Orgniztion October 2007, 85 (10) A

10 Reserch Steven A Hrvey et l. Tble 5. Men difference in score (%) by provider cdre Topic Doctor Medicl student Professionl nurse Knowledge test Infection prevention Medicl student ** 9.2% Professionl nurse * 5.2% 4.1% Auxiliry nurse ** 11.8% 2.6% ** 6.6% Uncomplicted lbour nd delivery Medicl student 1.5% Professionl nurse ** 20.5% ** 19.1% Auxiliry nurse **25.6% ** 24.2% * 5.1% Immedite newborn cre Medicl student 4.5% Professionl nurse ** 20.6% ** 16.1% Auxiliry nurse ** 26.3% ** 21.8% * 5.7% Hemorrhge during pregnncy Medicl student 1.6% Professionl nurse ** 12.5% ** 10.9% Auxiliry nurse ** 18.1% ** 16.5% 5.6% Postprtum hemorrhge Medicl student 5.2% Professionl nurse ** 20.7% ** 15.5% Auxiliry nurse ** 23.3% ** 18.1% 2.6% Pregnncy-induced hypertension Medicl student 3.3% Professionl nurse ** 15.0% ** 11.7% Auxiliry nurse ** 17.5% ** 14.2% 2.5% Sepsis Medicl student 2.9% Professionl nurse ** 14.8% ** 11.8% Auxiliry nurse ** 23.3% ** 20.4% ** 8.5% Active mngement of third stge lbour Medicl student 1.5% Professionl nurse ** 16.3% ** 14.8% Auxiliry nurse ** 20.7% ** 19.2% 4.4% Prtogrph test Written questions Medicl student 1.2% Professionl nurse ** 25.0% ** 23.9% Auxiliry nurse ** 34.1% ** 32.9% * 9.0% Grphing questions Medicl student 1.8% Professionl nurse ** 48.0% ** 49.8% Auxiliry nurse ** 57.5% ** 59.3% 9.4% B Bulletin of the World Helth Orgniztion October 2007, 85 (10)

11 Steven A Hrvey et l. Reserch (Tble 5, cont.) Topic Doctor Medicl student Professionl nurse Skills evlution Active mngement of third stge lbour Medicl student 3.8% Professionl nurse ** 11.6% * 7.8% Auxiliry nurse ** 16.1% * 12.3% 4.5% Mnul extrction of the plcent b Medicl student * 8.2% Professionl nurse NA NA Auxiliry nurse NA NA NA Bimnul uterine compression b Medicl student * 11.2% Professionl nurse NA NA Auxiliry nurse NA NA NA Immedite newborn cre Medicl student 0.3% Professionl nurse * 9.0% * 9.3% Auxiliry nurse ** 12.8% * 13.1% 3.8% Neontl resuscittion with mbu bg Medicl student 3.9% Professionl nurse ** 11.5% 7.6% Auxiliry nurse ** 16.5% * 12.7% 5.1% NA, non-pplicble. Sttisticl significnce for the difference in scores by provider cdre (pir-wise comprison) determined by Bonferroni post hoc test: * P < 0.05; ** P < b Skill in mnul removl of plcent nd bimnul uterine compression were evluted for doctors nd medicl students only. Professionl nd uxiliry nurses re prohibited from crrying out these procedures in Nicrgu. Since only two ctegories of provider were compred, sttisticl significnce ws determined by t-test (insted of ANOVA) nd post hoc pir-wise comprison ws unnecessry. Bulletin of the World Helth Orgniztion October 2007, 85 (10) C