Early Impact of an Integrated MNCH Program on Newborn and Child Health Outcomes, Northern Nigeria, 2009 to 2011

Transcription

1 Early Impact of a Itegrated MNCH Program o Newbor ad Child Health Outcomes, Norther Nigeria, 2009 to 2011 Sally E. Fidley, Omolara T. Uwemedimo, Hery V. Doctor, Cathy Gree, Fatima Adamu ad Godwi Y. Afeyadu BACKGROUND Of the seve millio deaths of childre uder the age of five that occurred worldwide i 2011, at least two-thirds could have bee preveted by low-cost, itegrated ewbor ad child heath (NCH) itervetios.[1, 2] The majority of global childhood deaths, largely due to eoatal problems, peumoia, diarrhea ad malaria, occur without ay cotact with the formal health system. More simply put, most childre die i their ow homes.[3] Accordigly, recet evidece from the 2008 Lacet Alma-Ata series suggests that itervetios cetered o scalig up commuity ad household care, i particular, have had a sigificat impact o ewbor ad child survival.[4] Such itervetios have icluded promotio of early iitiatio of breastfeedig, early postatal follow-up care of ewbors, exclusive breastfeedig for at least six moths, icreasig recogitio of dager sigs of illess amog caregivers, ad case maagemet of acute febrile illesses durig early childhood.[2, 3, 5, 6] As itegrated materal, ewbor, ad child health (MNCH) packages are ow beig delivered to-scale across may low-icome coutries, there has bee a acceleratio i the declie of global childhood mortality sice I Sub-Sahara Africa, where 1 i 9 childre uder five will die each year, reductios i childhood mortality have bee slower tha i the rest of the world, icludig Souther Asia.[1] While there are a umber of cotributory factors to this regioal disparity (e.g., extreme poverty, low female educatio ad autoomy, iadequate health system ifrastructure), a major obstacle is the iaccessibility to huma resources for health.[7] Coutries with higher desity of health professioals per capita have bee foud to have higher rates of ifat ad child survival. Sub-Sahara Africa has the lowest health worker desity i the world at 2.3 per 1,000 populatio.[8] This associatio is also otable withi coutries betwee rural ad urba areas, resultig i a geographic misdistributio where workers are most cocetrated i urba regios.[9] I light of these obstacles to care, the use of commuity health workers (CHWs) has emerged as a solutio with the strogest potetial to stregthe primary healthcare delivery i sub-sahara Africa.[10, 11] CHWs are described as members of the commuities where they work, should be selected by the commuities, should be aswerable to the commuities for their activities, should be supported by the health system but ot ecessarily a part of its orgaizatio, ad have shorter traiig tha professioal workers.[12] The effective use of CHWs has the capacity to address the three major gaps i service delivery: coverage, equity ad quality.[7] Limited data has suggested that this cadre of health worker is uiquely capable of reachig childre most at risk, those from the poorest families ad those livig i remote areas.[4, 13] I Nigeria, the most populous coutry i Africa ad a coutry with the secod highest burde of child deaths i the world, the eed to improve child survival is paramout.[3] Further, withi Nigeria there are marked differetials i child mortality rates, with rates i the orther states two to three times higher tha i the souther states. [14] I respose to this eed, the Partership for Revivig Routie Immuizatio i Norther Nigeria (PRRINN) was established i 2006 i four orther states of Nigeria (Jigawa, Katsia, Yobe ad Zamfara) ad the, i 2008, expaded to iclude materal, ewbor ad child health (PRRINN-MNCH). The program is comprehesive, ecompassig multiple aspects of the health system icludig huma resources, health goverace, health iformatio, stregtheig of cliical services, ad commuity egagemet i order to reduce materal, ewbor ad child mortality. The strategy adopted is to focus o revitalizig comprehesive primary care services usig a cluster approach which builds capacity for the provisio of emergecy obstetrical care services at selected facilities, with strog primary care facilities support care ad referrals to these desigated ceters i each cluster. Attetio is paid to traiig of health care workers at all levels i this cluster, as well as buildig the demad for health care services withi commuities served by these desigated emergecy care facilities. The program utilizes a operatios research approach that promotes progressive learig, with studies supportig cotiuous improvemet of program activities.

2 A key elemet of this itegrated strategy is the developmet of a etwork of CHWs, who bridge betwee the household ad the health facility. This paper will focus o the effects of the commuity based service delivery program withi PRRINN-MNCH from 2009 to Specifically, we will report chages i ewbor ad child health care kowledge ad behaviors amog caregivers, ad chages i ewbor ad child morbidity. METHODS Itervetio Desig The focus of this study is o the impact of the materal, ewbor, ad child health itervetios which were implemeted three of the four orther Nigeria states where PRRINN has expaded its MNCH activities, amely Katsia, Yobe, ad Zamfara, with respective populatios of 5.8, 2.3, ad 3.3 millio, accordig to the 2006 populatio cesus of Nigeria. The program desig focuses o improvig materal, ewbor ad child health (MNCH) care by clusters of local govermet areas (LGAs) per state, which each comprise a catchmet area for emergecy obstetrical care (EOC) services. A total of 15 LGAs were selected as the first itervetio clusters, 4-6 per state. The remaiig LGAs were desigated as either low-itesity areas, characterized by havig statewide policy chages without activities to improve health system ifrastructure ad MNCH care demad, or cotrol areas. The health system stregtheig compoet of the itervetio icludes upgradig EOC services withi local health facilities, midwife traiig ad postig through the Nigeria govermet s Midwife Service Scheme, establishig plaig ad maagemet techiques withi existig facilities ad establishig the Primary Health Care Uder Oe Roof, which cosolidates ad coordiates the differet compoets of primary care i oe health cliic or post. Complemetig these supply-side chages, are activities that create demad for MNCH services. Selected groups of villages served by primary care facilities liked to the upgraded EOC facility participate i a commuity egagemet process, which aims to icrease awareess of ad chage health behaviors to respod to MNCH barriers. Core to this process is a commuity discussio group methodology, facilitated by traied commuity voluteers (CVs), which provides a space for reflectio ad problem solvig for the most prevalet MNCH problems affectig the commuity. CVs are recruited i each commuity ad traied to do outreach ad social mobilizatio, emphasizig the use of commuity discussio groups ad jigles ad other visual-auditory cues to educate about critical MNCH issues, such as dager sigs for a pregacy or the timig of childhood vacciatios. I additio to these health educatio roles, the CVs also aid i idetifyig at risk wome ad childre ad referrig them to the earest facility for care. Betwee 2010 ad 2013 almost 30,000 CVs, primarily wome, have bee recruited, mobilized ad supported i their work i over 3600 commuities. Commuity discussio group participats are ecouraged ad supported to establish emergecy systems to tackle key barriers of access to ad affordability of MNCH services, icludig establishmet of blood door groups, commuity emergecy savigs schemes, commuity emergecy trasport schemes ad a mother s helpers system. Members of the discussio groups are ecouraged to share what they kow with their families ad peers betwee sessios, leadig to rapid saturatio of the etire commuity with ew ideas. This work is reiforced by mass commuicatio activities, icludig the use of radio jigles to promote birth preparedess or childhood immuizatios. Lastly, i the year before the mid-term household survey the program also developed a small cadre of CHW providig commuity-based service delivery (CHW-CBSD), to provide primary health services directly to families through rotatig visits or exteded availability through residece i the commuities. This category of CHW was recruited amog uemployed but previously traied Juior Commuity Health Extesio Workers (traied by the state Schools of Health Techology), who were the give additioal traiig ad tool kits to eable them to make home visits, egage mothers usig supportive commuicatio techiques, provide basic prevetive ateatal care (ANC) ad NCH services, as well as limited treatmet per Itegrated Maagemet of Newbor ad Childhood Illess, ad refer to the primary health care facility for treatmet as eeded. The CHW CBSD are provided with trasport to eable visit families i commuities o a regular basis, spedig most of their time visitig ad providig prevetive ad basic treatmet services i the commuity. At the time of the mid-term household survey these CHWs were active i 25 commuities, all of which also had CVs supportig their work through commuity educatio ad mobilizatio.

3 Evaluatio Desig The assessmet of the impact of the CBSD programs uses a quasi-experimetal desig usig pre- ad post-itervetio household surveys i both the itervetio ad cotrol commuities. The pre-itervetio or baselie household survey (BHS) was coducted i 2009 ad the post-itervetio survey, the midterm household survey (MHS) was coducted i This program is grouded i the hypothesis that this multicompoet itervetio will lead to chages i health kowledge ad behaviors ad attitudes towards existig services, resultig i icreased service utilizatio ad improved health outcomes. The evaluatio of the impact of this itegrated MNCH package takes ito accout both availability of program ad actual idividual participatio i ay of the program s commuity-based service activities. Availability of the program activities was assessed by compariso of itervetio ad cotrol areas. Idividual exposure to the program was assessed by the woma s resposes regardig the source of iformatio or health care advice, which allowed for differet sources correspodig to the alterative CBSD strategies. Study sample The samplig pla was a stratified two-stage cluster, radom sample, with oversamplig of idividuals i the MNCH itervetio clusters. Idividuals from MNCH clusters were oversampled usig a ratio of 2:1, eve though MNCH clusters cover a sigificatly lower proportio of the populatio of each state. I the BHS there were 24 Local Govermet Areas (LGAs), with 3,901 households sampled i the itervetio area ad 2,444 i the cotrol areas. For the MHS, the same itervetio LGAs were icluded, but we excluded LGAs of the state capitals (cosidered ot a appropriate cotrol for the largely rural itervetio). This left 15 LGAs i the samplig frame for the MHS. The MHS sample size was 770 per state, yieldig a sample of 1,577 households i the itervetio areas ad 733 i the cotrol areas. I both the BHS ad MHS, the umber of households at the first stage was proportioal to the size of the uit, the eumeratio area i the baselie ad the LGA i the MHA. I the MHS, commuities i the itervetio LGAs were icluded i the itervetio if it was cofirmed that PRRINN-MNCH programs had bee active i the commuity. I the MHS, samplig withi each commuity was also proportioal to size withi each itervetio ad cotrol LGA. The samplig fractio for each commuity was determied by iformatio o the total households from the commuity leadership. For both surveys, households withi each selected commuity were radomly sampled usig a procedure similar to that used i the WHO-EPI cluster surveys, amely by umberig the samplig households accordig to the commuity samplig fractio alog radomly selected paths leadig out from the ceter of the village. The household was the ultimate samplig uit. I compouds that comprised oe to three households, oe household was radomly chose for iterviews; i compouds with four to six households, two were surveyed; i compouds with seve or more households, three were surveyed. Withi each radomly selected household, i the baselie survey, all ever-married wome of childbearig age (15-49 years) were iterviewed, whereas i the mid-term survey oly oe ever-married wome with at least oe child bor i the last 5 years was selected for iterview. I the BHS there were 6,842 wome with successfully completed iterviews, while i the MHS there were 2,310 completed iterviews. Aalysis The depedet variables are the key health behaviors pertaiig to ewbor care ad care of sick childre. Baselie ad mid-term data were separately aalyzed usig appropriate samplig weights. The two sets of survey data were separately aalyzed usig appropriate samplig weights, ad bivariate tests (Chi-square, t-tests ad z-scores) were used to test for sigificat differeces betwee the key idicators measured by the BHS ad the MHS. I additio to the compariso betwee the two periods of time, the aalyses also test for sigificat differeces betwee the BHS ad the MHS results by itervetio status (itervetio or cotrol). Aalyses were performed usig Stata 11.0 (Statacorp, College Statio, TX) ad SPSS versio 19.0 (SPSS Ic. Chicago, Ill). RESULTS Respodet Characteristics

4 The wome iterviewed with the MHS are youger ad of lower social status, characteristics ofte associated with poor access health care workers or services. I both rouds, most wome iterviewed were betwee the ages of 20 ad 34, but there was a shift toward slightly youger ages i the MHS. (See Table 1). Virtually all wome iterviewed were curretly married, ad about 80% were moogamously married. The proportio polygamous rose slightly from 19.3% to 23.3% i the MHS. Over 80% of wome had o formal schoolig, ad amog those with some schoolig, there were fewer wome at mid-term who had atteded more tha primary school (27.9% at midterm, versus 47.0% at baselie.) The majority of wome i both surveys could ot read or write ay laguage (Hausa, Arabic, Eglish). Most families eared their livig by subsistece farmig, with 57.0% raisig millet ad 29.0% raisig cor. Two-thirds of the families were self-sufficiet from their ow productio i the year prior to the survey. Two-thirds of wome work without cash icome o the family farm, maitaiig the household, ad raisig childre. Roughly oe-third cosidered themselves fulltime housewives (30.3% vs. 43.6%, BHS vs. MHS). Oly 44.3% at the baselie vs. 33.0% at midterm worked for cash or i-kid earigs doig farmig, food processig or agricultural processig. The most commo alterative sources of icome were tradig (17.5% vs. 26.4%, BHS vs. MHS). Wome iterviewed i the MHS also were more likely to have access to a cell phoe (7.9% vs. 31.7%, BHS vs. MHS). [Table 1 about here] The majority of the households (about 80% i both surveys) icluded oly oe family, but if the household has more tha oe family i the compoud, there were a average of 2.5 families livig together. There were a average of four wome livig i each household, ad of these wome, o average 1.5 had give birth i the past year. Newbor ad Child Health Outcomes I 2011, more ifats were protected from tetaus. The proportio of wome who had received atitetaus vacciatios had icreased from 69.0% to 85.0%, with the icreases equal i the cotrol ad the itervetio areas. There was a sigificat icrease i the proportio ewbors first breastfed withi 24 hours from birth, from 42.9% to 57.5%, with sigificatly more (60.5%) i the itervetio areas. (See Table 2) Fewer ifats had a postatal check by a health worker withi 48 hours of birth, dow from 39.2% at baselie to 27.5% i the itervetio ad 18.9% i the cotrol areas. However, there was a large chage i who checked o the ewbor. At baselie, the majority of ewbors were checked at home by the traditioal birth attedats (TBAs) (40.8%), while at the mid-term most ewbors were checked by a urse/midwife at the health facility, 51.3%% i the cotrol areas ad 38.6%% i the itervetio areas. More ewbors were checked by CHWs, with eve more i the itervetio (46.3%) tha cotrol areas (35.8%). There was a sigificat icrease i ewbor care provided to the ifat: cord care from 7.0% to 26.4%, washig the baby i warm water from 39.1% to 52.9%, kagaroo care from 16.8% to 17.5%, breastfeedig withi eight hours from 15.6% to 39.3%, ad ewbor vacciatios from 3.1% to 22.2%. (See Figure 1) Cosistetly, the provisio of ewbor care elemets was sigificatly greater i the itervetio tha i the cotrol zoe, the sole exceptio beig ewbor vacciatios. At the midterm i 2011, sigificatly more wome were told about how to care for their ewbor, up from 68.1% to 71.6%, with eve more iformed i the itervetio tha cotrol areas. The other major chage was a shift from relyig o the TBA for iformatio about ewbor care (from 48.4% to 11.0%) to CHWs (from 6.8% to 11.7%, ad further icrease to 13.9% i the itervetio commuities). The impact of the commuity discussio groups is see i the large share of wome learig about ewbor care from wome s groups, frieds, ad family. [Table 2 about here] At the midterm follow-up, most wome kew at least oe of the ewbor dager sigs, with the most commoly kow dager sig beig high fever, kow by 82.7% i the cotrol ad 84.2% i the itervetio commuities. (See Table 3) May wome kew other critical dager sigs that idicated the eed for the baby to be see by a health worker. I the itervetio areas, 31.0% kew to worry about diarrhea, dehydratio ad suke fotael ad about fittig or covulsios, sigificatly more tha i the cotrol areas. Wome i the itervetio areas were also more likely to kow about breathig problems ad ot beig able

5 to suckle or refusig to feed. About half (47.2%) of the ewbors experieced at least oe of these dager sigs durig the first six weeks of life, with o differece i the frequecy betwee cotrol ad itervetio areas. Of those with oe of the dager sigs, the most commo dager sig was fever, with sigificatly more of the ewbors havig fever i the cotrol (31.8%) tha itervetio commuities (27.8%). Oe i five ewbors i the itervetio commuities (20.6%) cried excessively, oted by sigificatly more mothers i the itervetio tha cotrol commuities (16.6%). More ewbors i the itervetio versus cotrol commuities were also oted to have breathig problems (18.0% vs. 14.9%). The ext most commo dager sigs were diarrhea ad swolle stomach, each experieced by about 16% of all ewbors i the cotrol areas ad 19% i the itervetio areas. [Table 3 about here] Oe-third (32.6%) of the households reported a sick child (uder age 5) i the moth prior to the iterview. The average duratio of illess was 7.8 days. The most commo illesses were fever (presumed to be malaria) (26.5%), diarrhea (17.4%), cough (13.3%), ad malutritio or weight loss (13.7%). Oe i te (11.2%) childre had both fever ad cough, as did aother 10.1% who had diarrhea ad malutritio (See Table 3). The reported illess prevalece rates were slightly higher i the itervetio tha cotrol commuities. Betwee the BSH ad MHS i 2011, Table 2 shows there was a shift i the source of advice about the care of sick childre. More wome i the itervetio commuities kew about the care of their sick childre, with oly 22.5% i the itervetio areas ad 28.3% i the cotrol areas havig o oe to teach them about the care of their sick childre. More wome leared how to care for sick childre from CHWs, both at the health post ad i the commuity, with CHWs providig this iformatio to 14.5% i the cotrol commuities ad 23.8% i the itervetio commuities. Fewer relied o family ad frieds i the itervetio commuities, 27.1% vs. 32.2% i the cotrol commuities. I the itervetio commuities, fewer wome wet to a TBA or drug vedor/ chemist for advice o treatig a sick child. I both time periods ad regardless of the child s symptoms, about oe-third of all mothers with sick childre i the past moth reported seekig o advice ad providig o special care to the sick child. (See bottom row, Table 4) Approximately oe i twelve (8%) mothers i the itervetio commuities gave additioal fluids, icludig breastfeedig more, to their childre with fever, cough, fever ad cough, ad diarrhea, compared to slightly fewer givig fluids i the cotrol commuities. Although the reported rate of givig ORS declied betwee the BSH ad MSH, i the itervetio commuities over oe i te (9.2% to 13.8%) gave oral rehydratio solutio (ORS), compared to a ORS usage rate of 5.0% or below i the cotrol commuities. [Table 4 about here] Over oe-third used a aalgesic (paracetamol) to reduce fever at both baselie ad mid-term followup, with more usig aalgesics for diarrhea at mid-term tha at baselie. (See Table 4) Approximately oethird of all mothers also used atibiotics to treat their childre s fever, cough, or diarrhea. Use of atibiotics geerally was less at the mid-term tha baselie, but there were differet directios of chage i the cotrol ad itervetio commuities. Atibiotic treatmet of fever dropped to 26.4% i the cotrol commuities, compared to 32.2% i the itervetio commuities. A similar patter was see for the treatmet of fever ad cough. The reverse was see for cough, for which oly 31.4% of the mothers i the itervetio commuities gave atibiotics, compared to 37.7% i the cotrol commuities. This alterate patter was also see for atibiotic treatmet of diarrhea, higher i the cotrol (37.7%) tha itervetio (34.6%) commuities. Use of ati-malarials dropped precipitously betwee the baselie, whe 57.5% used them for fever ad/or cough, dow to 20% or less i both the itervetio ad cotrol commuities at the mid-term follow-up. About a quarter of mothers i both the cotrol ad itervetio commuities also reported usig cough medicie or other patet medicatios for cough, with or without fever. Betwee the BSH ad MSH, there was also a sigificat icrease i the reported use of traditioal medicie or herbs, particularly i the itervetio commuities, where the use of traditioal medicie or herbs was almost double that reported i the cotrol commuities for the treatmet of fever ad/or cough.

6 Betwee 2009 ad 2011, there was a declie i both the ifat ad the uder-five mortality rates. (See Figure 2) I 2009 the ifat mortality rate (IMR) was 90.0 deaths per 1000 live births, ad this fell to 79.0 i the cotrol commuities ad 50.5 i the itervetio commuities, averagig 58.5 i both. Durig this same period, the child mortality rate declied from 160 to 104 i the cotrol commuities ad 84 i the itervetio commuities. DISCUSSION Although the PRRINN-MNCH Programme had bee uderway for just over two years whe the MSH was coducted, there already was evidece of sigificat improvemet i several of the key ewbor, ad child health behaviors ad outcomes. The level of ewbor ad child morbidity remais high i these commuities, with almost half of all ewbors reported to show oe of the dager sigs ad oe-third of youg childre experiecig a illess episode i the previous moth. What is chagig is what their mothers are doig about these health crises. At the baselie i 2009, there were may more wome who did ot seek advice ad who did ot respod as quickly or as thoroughly as at the follow-up i Fairly cosistetly, the itervetio commuities display improved eoatal ad ewbor care practices. More wome i the itervetio tha cotrol commuities started breastfeedig immediately ad exclusively, had their ewbors checked by a health worker, washed their baby i warm water, kept the ifat warm with kagaroo care, ad kew to watch the ifat for fever or other dager sigs. These all were topics addressed i the commuity discussio groups or dialogues orgaized by CVs recruited ad traied by the project, ad the chages i care of the ewbor immediately after delivery reflect this icreased commuity dialogue, awareess ad support for the eed to keep a eye o the mother ad ewbor durig this critical period. The dialogues also ecouraged wome to go to the CHW or the health worker at the closest health post istead of the TBA, because these traied idividuals ca do more to help the mother recogize ad respod to ay problems that might develop. A secod patter see i these results is the icreased reliace o CHW for iformatio about care of the ewbor or sick child. At baselie, almost half of all ewbors were checked by a TBA, whose activities geerally cosisted of assistace with cord care ad cleaig up the baby. At the midterm follow-up two years later, hardly ay wome took their ewbors to be see by a TBA. Istead, they had their babies check by the midwife or CHW at the health post or by a CHW makig a outreach visit to the commuity. The availability of these traied health workers was the direct result of the program s efforts to place midwives i the itervetio clusters ad to trai CHWs o ewbor care. The commuity dialogues geerated the awareess ad cofidece i goig to these wome for advice o their ewbors. A similar patter is see for seekig advice o the care of sick childre, with the itervetio group showig icreased reliace o the midwives ad CHWs ad less o the TBA or family ad frieds. Aother critical chage i ewbor care is the icreased uderstadig ad ability to observe ewbor dager sigs. I the itervetio commuities, more wome kew ewbor dager sigs ad they also kew more of them. Mothers i the itervetio commuities reported higher icidece of some of the ewbor dager sigs, but this likely reflects their ew foud uderstadig that these sigs are ot ormal ad require the mother to have her child treated quickly. While we were ot able to moitor what happeed whe these dager sigs were observed, wome reported that almost half of their ewbors had show at least oe of these dager sigs, with the most commo beig a high fever. It is likely that this level of observatio was coected to takig steps to seekig urget care, as that is what the dager sig message is all about. The higher quality ad availability of care at the primary health ceters would have give wome cofidece that arrivig their with their ewbor could ideed be a life-savig trip. Although the recogitio ad treatmet of sick childre (uder age 5) was less a priority i the iitial implemetatio of the program, there was some evidece that mothers were respodig with more home care

7 for their sick childre. Overall, mothers reported lower utilizatio of ORS packets, but that chage is likely related to a shift from use of the pre-packaged ORS to the istructios o mixig the sugar-salt solutio at home. I the cotrol commuities, there was hardly ay use of ORS at the midterm poit i 2011, whereas i the itervetio areas, over oe i te mothers were mixig up sugar-salt solutios to give their childre at home for fever, with ad without cough, ad diarrhea. They also reported givig more fluids to their childre, though the differeces were ot sigificat betwee the cotrol ad itervetio areas. Both chages reflect the ifluece of the commuity discussios i the itervetio areas, where wome i the commuity leared about the importace of rehydratio ad how to mix the sugar-salt solutios. This study has several limitatios. First, we did ot combie the datasets for a itegrated aalysis of the behavior chage betwee baselie ad midterm, which limits the aalysis to a compariso of meas ad proportios. The pre-post compariso betwee the BHS ad the plaed Ed-of-Project Survey (May 2013) will iclude this merger of data, eablig regressio aalyses to be used to idetify the predictors of behavior chage ad health outcomes. Secod, all behaviors ad health outcomes are by self-report, with o medical verificatio of the health evets or deaths. As i ay retrospective self-report, particularly of ifat ad child deaths, there is likely to be uder-reportig. We assume that the level of uder-reportig for these evets is comparable across both surveys, but with the icreased push for birth registratio by the programme, it is possible that the reportig of births ad deaths is higher i the MHS, which would ted to upwardly bias the mortality estimates relative to the BHS. Third, ad most importatly, the CHW itervetio compoets were oly operatioal for approximately oe year prior to the MHS, ad hece the period of exposure is more limited for these elemets of the itervetio. The lack of chage for some idicators betwee the BHS ad MHS may therefore be due to limited exposure to the itervetio. The Ed-of-Project Survey will permit a loger duratio of exposure to be assessed, ad it will iclude a more detailed set of measures of program participatio. These limitatios otwithstadig, the et result of the chages i uderstadig about ewbor ad sick child care are evidet i the observed declies i ifat ad child mortality rates durig this short time period. Declies i both rates were observed i both the cotrol ad itervetio areas, but they were sigificatly greater i the itervetio commuities. A umber of chages i the itervetio commuities ad i the primary care system are likely to have cotributed to the declie i ifat mortality from baselie to mid-term. The marked icrease i ati-tetaus vacciatio rates likely reduced eo-atal deaths from this disease, ad the icreases i early ad exclusive breastfeedig likely further protected the ewbor from exposure to bacteria. Most importatly, it is very likely that the chages i uderstadig about ewbor careboth the importace of a prompt check by a qualified perso ad the eed to be vigilat for dager sigs i the ewbor- cotributed to the sigificat declie i ifat mortality which has bee observed i the itervetio commuities.

8 Refereces 1. Alkema L, You D: Child mortality estimatio: a compariso of UN IGME ad IHME estimates of levels ad treds i uder-five mortality rates ad deaths. PLoS Med 2012, 9(8):e Joes G, Steketee RW, Black RE, Bhutta ZA, Morris SS: How may child deaths ca we prevet this year? Lacet 2003, 362(9377): UNICEF: The State of the World's Childre-Special Editio: Celebratig 20 Years of the Covetio o The Rights of A Child. I.: UNICEF; Bhutta ZA, Ali S, Couses S, Ali TM, Haider BA, Rizvi A, Okog P, Bhutta SZ, Black RE: Alma-Ata: Rebirth ad Revisio 6 Itervetios to address materal, ewbor, ad child survival: what differece ca itegrated primary health care strategies make? Lacet 2008, 372(9642): Bhutta ZA, Chopra M, Axelso H, Berma P, Boerma T, Bryce J, Bustreo F, Cavagero E, Cometto G, Daelmas B et al: Coutdow to 2015 decade report ( ): takig stock of materal, ewbor, ad child survival. Lacet 2010, 375(9730): Bhutta ZA, Darmstadt GL, Hasa BS, Haws RA: Commuity-based itervetios for improvig periatal ad eoatal health outcomes i developig coutries: a review of the evidece. Pediatrics 2005, 115(2 Suppl): Kiey MV, Kerber KJ, Black RE, Cohe B, Nkrumah F, Coovadia H, Nampala PM, Law JE, Axelso H, Bergh AM et al: Sub-Sahara Africa's mothers, ewbors, ad childre: where ad why do they die? PLoS Med, 7(6):e Guilbert JJ: The World Health Report 2006: workig together for health. Educ Health (Abigdo) 2006, 19(3): Wirth ME, Balk D, Delamoica E, Storeygard A, Sacks E, Miuji A: Settig the stage for equity-sesitive moitorig of the materal ad child health Milleium Developmet Goals. Bulleti of the World Health Orgaizatio 2006, 84(7): Christopher JB, Le May A, Lewi S, Ross DA: Thirty years after Alma-Ata: a systematic review of the impact of commuity health workers deliverig curative itervetios agaist malaria, peumoia ad diarrhoea o child mortality ad morbidity i sub-sahara Africa. Huma resources for health 2011, 9(1): Chopra M, Sharkey A, Dalmiya N, Athoy D, Biki N: Strategies to improve health coverage ad arrow the equity gap i child survival, health, ad utritio. Lacet 2012, 380(9850): Stregtheig the performace of commuity health workers i primary health care. Report of a WHO Study Group. World Health Orgaizatio techical report series 1989, 780: Lewi S, Lavis JN, Oxma AD, Bastias G, Chopra M, Ciappoi A, Flottorp S, Marti SG, Patoja T, Rada G et al: Supportig the delivery of cost-effective itervetios i primary health-care systems i low-icome ad middle-icome coutries: a overview of systematic reviews. Lacet 2008, 372(9642): Natioal Populatio Commissio [Nigeria] ad ICF Macro. Nigeria Demographic ad Health Survey 2008 Report. Abuja, Nigeria ad Calverto, Marylad: Natioal Populatio Commissio ad ICF Macro, 2009.

9 Table 1: Backgroud characteristics of respodets, Norther Nigeria, 2009 vs Backgroud characteristics BHS 2009 Number (%) MHS 2011 Number (%) Age group (years) (4.7) 179 (7.8) (11.1) 529 (22.9) ,417 (19.2) 608 (26.3) ,618 (22.0) 526 (22.8) ,176 (16.0) 281 (12.2) ,015 (13.8) 147 (6.4) (13.3) 29 (1.3) Marital status Married 6,664 (97.2) 1581 (99.0) Widowed 78 (1.1) 5 (0.3) Divorced or separated 115 (1.7) 10 (0.6) Rak of wife 1 5,401 (80.6) 1182 (76.2) 2 1,128 (16.8) 301 (19.4) (2.7) 68 (4.4) Formal educatio Yes 1,293 (18.8) 219 (13.9) No 5,593 (81.2) 1358 (86.1) Level of formal educatio Primary 713 (53.0) 155 (72.1) Secodary 466 (34.7) 34 (15.8) Post-secodary 166 (12.3) 26 (12.1) Readig ad writig i Hausa Not at all 5,373 (78.1) 1392 (89.5) With difficulty 644 (9.4) 82 (5.3) Easily 864 (12.6) 81 (5.2) Readig ad writig i Arabic Not at all 4,174 (60.5) 1352 (85.9) With difficulty 1,635 (23.7) 167 (10.6) Easily 1,093 (15.8) 55 (3.5) Readig ad writig i Eglish Not at all 6,076 (88.3) 1511 (96.0) With difficulty 410 (6.0) 36 (2.3) Easily 399 (5.8) 27 (1.7) Occupatio Food processig 2,480 (36.2) 449 (28.5) Agricultural processig 310 (4.5) 70 (4.5) Farmig 249 (3.6) 0 (0.0) Tradig/Sellig 1,200 (17.5) 410 (26.4) Housewife 2,079 (30.3) 682 (43.6) Other 534 (7.8) 105 (6.7) Cell owership Yes 543 (7.9) 500 (31.7) No 6,363 (92.1) 1077 (68.3) Number 6,906 2,310 Notes: Some percetages may ot add up to 100 due to roudig of decimals. Some umbers for sub-categories may ot add up to the total due to (1) icosistecies across related variables ad (2) referece to specific category.

10 Table 2: Differeces i ewbor care for the most recet birth i past five years, by itervetio area, Norther Nigeria, 2009 vs BHS 2009 MHS 2011 Cotrol Itervetio p-value p- value Characteristic (%) (%) 2011 (%) 2011 (%) BHS vs Ctl vs It It Mother had ati-tetaus vaccie , First breastfeedig withi 24 hours ,335 2, First postatal check withi 48 hour Perso checkig ewbor Nurse/midwife CHW- health post CHW- outreach TBA Other Care provided to the ewbor No special ewbor care Cord care Wash baby Keep baby warm (kagaroo) Breastfeed immediately Watched for dager sigs Register the birth Newbor vacciatio Weigh baby Watch for high fever Source of iformatio about ewbor care* No oe Nurse/midwife CHW i health post CHW i outreach TBA Family/ frieds Drug vedor/ Chemist Other , , , , , Number of wome 6,208 2, ,576 Notes: *Source of iformatio about ewbor care is the perso checkig the ewbor ad couselig the mother after delivery; CHW Commuity Health Worker; TBA Traditioal Birth Attedat; Not applicable , ,

11 Table 3: Differeces i kowledge of ad respose to ewbor dager sigs ad illess episodes of childre uder age 5 by itervetio area, Norther Nigeria, 2011 Cotrol 2011 Itervetio 2011 p-value (Cotrol vs. (%) (%) Itervetio) Newbor dager sig kowledge ad respose type Kowledge of ewbor dager sigs Noe Kow Kow: High fever Stiff eck, fittig or covulsios Jaudice Difficult/fast breathig Not able to suckle/refuse to feed Diarrhea/dehydratio/suke soft spot Observatio of dager sigs i ewbor (<6 weeks old) High fever Stiff eck, fittig or covulsios Swolle stomach Diarrhea Difficult/fast breathig Not able to suckle / refuse to feed , Source of advice o care of sick childre Nurse/midwife CHEW i health post CHEW i outreach TBA Family/ frieds Drug vedor/ chemist Traditioal healer/ other No oe metioed Acute illess episodes i past moth, childre <5 years Fever Diarrhea Cough Malutritio (weight loss) Fever ad cough Diarrhea ad malutritio , ,

12 Table 4: Type of care give to sick child i the moth precedig the survey by itervetio area, Norther Nigeria, 2009 ad Type of care (%) ( with ay care) Fever/ cough =1, Fever oly = Cough oly = Fever ad Cough = Diarrhea = 1, Diarrhea =417 Cotrol vs. Itervetio C I C I C I C I Homecare Gave more fluids Gave ORS Medicatio use Aalgesics Atibiotics Ati-malarial Other drug Gave cough medicie Traditioal remedies Traditioal medicie/ herbs Did othig Total Sick 2, , Notes: Bold idicates a p-value of <0.05 betwee itervetio ad cotrol; Percetages do ot sum to 100% because multiple care activities may have bee used per episode; C ad I refer to cotrol ad itervetio areas, respectively.

13 Figure 1: Newbor Care Activities by Itervetio Area, Norther Nigeria 2009 & Cotrol 2011-Itervetio

14 Figure 2: Ifat ad Child Mortality Rates by Itervetio Area, Norther Nigeria, 2009 versus cotrol 2011-Itervetio CMR IMR