Local use of geographic information systems to improve data utilisation and health services: mapping caesarean section coverage in rural Rwanda

Local use of geographic information systems to improve data utilisation and health services: mapping caesarean section coverage in rural Rwanda
of 9
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Similar Documents
  Local use of geographic information systems to improve datautilisation and health services: mapping caesarean sectioncoverage in rural Rwanda Leanna Sudhof  1 , Cheryl Amoroso 2 , Peter Barebwanuwe 2 , Fabien Munyaneza 2 , Adolphe Karamaga 3 ,Giovanni Zambotti 4 , Peter Drobac 2,5 and Lisa R. Hirschhorn 6,7 1  Women and Infants Hospital, Providence, RI, USA 2  Partners In Health, Rwinkwavu, Rwanda 3  Ministry of Health, Rwinkwavu, Rwanda 4  Harvard Center for Geographic Analysis, Cambridge, MA, USA 5  Division of Global Health Equity, Brigham and Women’s Hospital, Boston, MA, USA 6  Department of Global Health and Social Medicine, Harvard Medical School, Cambridge, MA, USA 7  Partners In Health, Boston, MA, USA Abstract  objectives  To show the utility of combining routinely collected data with geographic location usinga Geographic Information System (GIS) in order to facilitate a data-driven approach to identifyingpotential gaps in access to emergency obstetric care within a rural Rwandan health district. methods  Total expected births in 2009 at sub-district levels were estimated using community healthworker collected population data. Clinical data were extracted from birth registries at eight healthcentres (HCs) and the district hospital (DH). C-section rates as a proportion of total expectedbirths were mapped by cell. Peri-partum foetal mortality rates per facility-based births, as well as therate of uterine rupture as an indication for C-section, were compared between areas of low and highC-section rates. results  The lowest C-section rates were found in the more remote part of the hospital catchmentarea. The sector with significantly lower C-section rates had significantly higher facility-based peri-partum foetal mortality and incidence of uterine rupture than the sector with the highest C-sectionrates ( P  <  0.034). conclusions  This simple approach for geographic monitoring and evaluation leveraging existinghealth service and GIS data facilitated evidence-based decision making and represents a feasibleapproach to further strengthen local data-driven decisions for resource allocation and qualityimprovement. keywords  geographic information systems, maternal mortality, regional health planning, access tohealth care, Rwanda Introduction Most maternal morbidity and mortality are preventable,and yet in 2008, nearly 350 000 women died in preg-nancy or childbirth worldwide. Although progress hasbeen made in reducing maternal mortality, the world willnot achieve the target of Millennium Development Goal(MDG) 5: a 75% reduction in Maternal Mortality Ratio(MMR) by 2015 (Hogan  et al.  2010). Caesarean sectionrates are a commonly used indicator to monitor accessto, and use of, emergency obstetric care, one of the criti-cal components in reducing maternal mortality (WHO et al.  2009). WHO estimates that population rates of Caesarean sections (C-sections) between 5 and 15%reflect appropriate access and utilisation (2009). Popula-tions with lower rates potentially represent compromisedaccess and utilisation, raising the risk of preventablematernal death.In Rwanda, where 85% of the population live in ruralareas, significant progress has been made towards achiev-ing MDG 5 (Hogan  et al.  2010). The MMR hasdecreased from an estimated 1300 deaths per 100 000live births during the period 2000  –  2004, to 487 for theperiod 2004  –  2010 (Hill  et al.  2007; National Institute of  18  © 2012 Blackwell Publishing Ltd Tropical Medicine and International Health doi:10.1111/tmi.12016 volume 18 no 1 pp 18  –  26 january 2013  Statistics of Rwanda 2011). Despite this progress, theRwandan Health Sector Strategic Plan-II noted thatMDG 5 was ‘proving the most difficult to achieve’(2009), and in response, the Rwandan Ministry of Health(MOH) has prioritised maternal health.One priority for countries in sub-Saharan Africa work-ing towards maternal mortality reduction is improvingthe use of existing data to improve health systems(Gething  et al.  2007). Local decision makers often lackthe tools to efficiently and effectively use data to identifygaps and ensure evidence-based decision making andequitable allocation of limited resources. One tool thathas been increasingly used in sub-Saharan African set-tings to address this challenge is Geographical Informa-tion System(s) (GIS). This work has included thedocumentation of the negative impact of distance on ser-vice utilisation and identification of areas with low accessto services (Noor  et al.  2003; Heard  et al.  2004; Feikin et al.  2009; Cooke  et al.  2010). Tanser  et al.  (2001)analysed the geographic variability in usage rates of clin-ics in rural South Africa to identify areas of under-perfor-mance. GIS was also used for program monitoring andvisualising progress in coverage of insecticide-treated netsin malaria-endemic regions between 2000 and 2007(Noor  et al.  2009).Most GIS use reported in the literature was by groupswith good data management infrastructure and expertisein the context of research institutions or large initiatives.In this article, we describe the implementation of a rela-tively low-cost, community-based approach to integrateGIS analysis and data use into district-level monitoringand evaluation in the catchment area of a single districthospital (DH) in southern Kayonza District. Local decisionmakers then used the geographic disaggregation and visu-alisation of locally available data to identify potential gapsin access to and utilisation of emergency obstetric care,providing guidance for evidence-based health resourceallocation to improve equity at a sub-district level. Methods Study setting Since 2005, Partners In Health (PIH) has collaboratedwith the Rwandan Ministry of Health to strengthen thehealth system in three rural districts. Joint interventionsat the DH, health centre and community level have beenemployed to improve financial and social access to healthcare, including community health worker initiatives toaccompany pregnant women in obtaining prenatal careand incentivised grants to the health centres (HCs) tominimise barriers to care as well as improve the qualityof care. In 2009, PIH introduced GIS mapping and analy-sis as an adjunct to existing monitoring and evaluationefforts in the uniformly hilly southern region of KayonzaDistrict to examine geographic access.The health system in the southern part of Kayonza Dis-trict includes one DH and eight HCs serving 7.5 adminis-trative sectors, which make up the catchment area of theDH (see Figure 1). The names of the DH, HCs and sec-tors were replaced with letter designations, with the HCand the sector they serve given the same letter. Sectorsincluded a population of 15 000  –  25 000 people, coveringan average area of 48 km 2 . Each sector is divided by theRwandan government into four or five administrativecells, each containing 5  –  16 villages. Each HC serves onesector, with the exception of the HC in Sector C, whichserves a single cell in the DH catchment. Clinically, HCsare staffed by nurses alone, and standard obstetrical ser-vices include normal labour and vaginal delivery. Compli-cated labour and high-risk cases are referred to the DH,and C-sections are only performed at the DH. Operativevaginal deliveries (forceps and vacuum) are not per-formed at the HCs or the DH. One HC in the catchment,HC (H), does not provide childbirth services becausewomen deliver at the adjoining DH. Geographic data collection To support integration of GIS into routine monitoring andevaluation, PIH employed a Rwandan Bachelor-level GISstudies graduate to provide training and support data usethrough analysis, interpretation and feedback of results toprograms. The GIS team also included a local high schoolgraduate program associate who acted as coordinator of data collection activities, which included mapping at thevillage level with community health workers. District-widegeographic data encoding sector boundaries and roadswere obtained from the Centre for GIS at the NationalUniversity of Rwanda, and the National Institute forStatistics in Rwanda provided geographic data encodingcell boundaries. Sector and district maps were created withArcGIS 9.3.1 software, which was obtained through anacademic partnership with Harvard University. Allanalysis was carried out by the local GIS team. Data collection for facility-based delivery and caesareansections C-section and facility delivery data were manually col-lected over two months by a medical student and dataofficer. Maternity data in paper-based labour registries atthe eight HCs and the DH for the 4583 women whodelivered between 1 January 2009 and 31 December © 2012 Blackwell Publishing Ltd  19 Tropical Medicine and International Health  volume 18 no 1 pp 18  –  26 january 2013 L. Sudhof   et al.  Local use of GIS mapping to improve rural C-section access  2009 were extracted and entered in a secure Excel data-base. The 104 patients from outside of the DH catchmentwere excluded. Data extracted included patient name (foridentification of duplicated entries), age, address, admis-sion date, discharge date, delivery date, APGAR score,referral status and reason for referral. Indications for pre-partum referral of labouring patients documented in HCregistries and indications for C-section documented at theDH were categorised according to first recorded diagno-sis: failure to progress, cephalopelvic disproportion, foetaldistress, repeat Caesarean, malpresentation, uterine rup-ture, failed induction, placental abnormality and pre-eclampsia. C-sections with documented reasons that arenot recognised indications for C-section were coded aselective (‘voluntary’, ‘tubal ligation’ or ‘normal labour’). Estimating total births Total expected births were calculated using total popula-tion at the cell and sector levels as estimated below andpublished crude national birth rates. We used the esti-mate of Rwanda birth rate of 38.1 births per 1000 popu-lation from the US Census Bureau 2009 estimate, whichwas the most recent and also consistent with the trend inbirth rates reported in the 2005 and 2007 Rwanda DHS(US Census Bureau 2009, Institut National de la Statis-tique du Rwanda (INSR), 2006, Ministry of Health(Rwanda), Macro International, Inc 2007  –  2008, 2010).The last Census in Rwanda was conducted in 2002, withresults only made available at the provincial level to theauthors (Minnesota Population Center 2011). The rapidincrease in Rwanda’s population since 2002 made theapplicability of that data to estimate births in 2009 aconcern (World Bank 2009). We thus turned to Rwan-da’s robust CHW system, which first began collectingand reporting population data on households in mid-2009. The CHWs (2  –  5 per village) submitted monthlyreports, which included the number of inhabitants. HCsroutinely collected monthly CHW reports and aggregatedthe data at the sector level. We used CHW-level paperreports for up to 4 months (August, September, andDecember 2009 and January 2010) and aggregated thetotal population numbers in these reports by village bymonth, before calculating the median monthly villagepopulation for each village. The village-level estimateswere then aggregated to give cell and sector populationnumber estimates. These population estimates, along withthe estimated national crude birth rate, were used to cal-culate expected births by cell and sector. To test the vari-ability in the CHW population data, we redid thecalculations using the lowest and highest observations foreach village, and the effect on the cell- and sector-levelC-section rates was marginal, with a less than 5% differ-ence for all but two cells, where the difference was less MukarangeNyamiramaRwinkwavuRuramiraKabarondoMuramaKabare HC Inside DH CatchmentCells in DH CatchmentCells Outside CatchmentMajor Unpaved RdPaved RdSector Boundaries Legend Ndego 510 km N 0 Rwinkwavu Hospital Figure 1  Map of study area. 20  © 2012 Blackwell Publishing Ltd Tropical Medicine and International Health  volume 18 no 1 pp 18  –  26 january 2013 L. Sudhof   et al.  Local use of GIS mapping to improve rural C-section access  than 7%. Once the DHS 2010 was available, we againtested the reliability of the CHW data by calculatingexpected births using AfriPop 2010 estimates of womenof child-bearing age, and population distribution andrural age-specific fertility rates averaged across the 2007  –  2008 and 2010 DHSs. A paired t-test showed no statisti-cal difference between the two methods ( P  =  0.5098). Outcome indicators The main outcome indicator, C-section rate, was mea-sured using two different denominators: estimated totalbirths in the HC catchment areas (population-basedC-section coverage rate) and total health facility births(health facility-based C-section rate) (Table 1). Otherindicators included the hospital-based delivery rate andfacility-based peri-partum foetal mortality rate. Both of these rates were calculated as a proportion of total docu-mented births by patient cell or sector of residence.Health facility refers to both the HCs and the DH. Analyses Maps were created in ArcGIS 9.3.1, linking C-section cov-erage rates and facility-based peri-partum foetal mortalityrates to cells, the government-determined sub-sectoradministrative boundaries. Distances by road betweeneach HC and the DH were calculated using the networkanalysis extension. Ambulance travel time was estimatedusing the calculated distances and an assigned averagespeed on the two different types of road, paved road andmajor unpaved road. Health facility and population-basedC-section rates, facility-based peri-partum foetal mortalityrates and the rates of C-section for uterine rupture werecompared across sectors. Chi-squared test or Fisher exacttest were used to compare the sectors with the lowest andhighest C-section rates to the other sectors.Workshops led by the GIS team were held with Minis-try of Health and PIH staff at the district, hospital andlocal levels to discuss how GIS data can be used in pro-gram monitoring and evaluation as well as qualityimprovement activities. Ethical approval This project was reviewed by the institutional reviewboards of Partners Health care in Boston, USA and theRwanda National Ethics Committee. Patient data wereaggregated at the village or cell level and were anony-mised for analysis and reporting. Results HC catchments and distances to DH Documentation of patient addresses at the HCs allowedlinkage of over 97% of women presenting for delivery ata health facility to their cell of residence, although oneHC (G) had lower rates of documentation (Table 2).Most women presenting in labour resided within theHC’s catchment, with the exception of HC (C), whereabout 50% of presenting patients lived in neighbouringsectors. The average age of women that delivered in theDH catchment ranged between 25.8 and 27.7 by sector.As shown in Figure 2, distances by road between theHCs and the DH ranged from 12 km to 32 km, and theestimated travel times varied from a half hour to twohours. Table 1  Definitions of outcome indicatorsPopulation-based C-section coverage rate(per 100 estimated total births)Total C-sections at the DH for women residing in a cell/Expected total births forthat cell. This rate is a UN indicator for access to emergency obstetric care(WHO  et al.  2009).C-section rate in health-facility based deliveries(per 100 health-facility based deliveries)Total C-sections at the DH in women residing in a sector/Total documentedhealth-facility based deliveries at any health facility in S. Kayonza for womenresiding in that sector. This rate shows the percentage of women presenting forhealth-facility based deliveries who received a C-section.DH delivery rate(per 100 health-facility based deliveries)Total deliveries that occurred at the DH for women residing in a sector/Totaldocumented health-facility based deliveries at any health facility in S. Kayonzafor women residing in that sector. This measured the proportion of documentedhealth-facility based deliveries that occurred at the DH, as a measure of healthfacility to health facility access.Facility-based peri-partum fetal mortality rate(per 100 health-facility based deliveries)Total neonates that were documented to be dead at delivery or to have died inthe first ten minutes of life whose mothers lived in a cell/Total documentedhealth-facility based deliveries at any health facility in S. Kayonza for womenresiding in that cell. © 2012 Blackwell Publishing Ltd  21 Tropical Medicine and International Health  volume 18 no 1 pp 18  –  26 january 2013 L. Sudhof   et al.  Local use of GIS mapping to improve rural C-section access  Based on the CHW data, cell populations ranged from2 300 to 8 200 people, and sector populations rangedfrom 14 000 to 31 000. The calculated expected births in2009 ranged from 88 to 313 births by cell, and from 551to 1183 births by sector.The number of patients presenting in labour fromwithin the DH catchment ranged from 236 at HC (C) to1340 at the DH (Table 2). The DH delivery rate was sig-nificantly lower for women residing in Sector E( P  <  0.016) and significantly higher for women residingin Sector B ( P  <  0.001) than for women living in othersectors (Table 3). Population-based C-section coverage rates by cell The map of population-based C-section coverage rates(Figure 3) varied from 3.0 to 29.6%. The highest rates 120Travel time (min)Distance (km)100806040200Sector BSector FSector ASector GSector CSector DSector E Figure 2  Travel time and distance by road from each HC to theDH. Table 2  Geographic characteristics of women registered in maternity ward registries in S. KayonzaHealth facilityof presentation,represented byname of sectorTotal patientsin labor registeredat each health facility*% Patients linkedto cell of residenceTotal patientspresenting from assignedHC catchmentSector A 429 428 (99.8) 421 (98.1)Sector B 797 795 (99.7) 777 (97.5)Sector C 236 235 (99.6) 131 (55.5)Sector D 384 384 (100.0) 381 (99.2)Sector E 347 344 (99.1) 331 (95.4)Sector F 527 525 (99.6) 520 (98.7)Sector G 419 369 (88.1) 412 (98.3)Sector H (HC + DH) †  1340 1302 (97.2) NA*Including only patients srcinating from within DH catchment. † Patients identified as having been referred from one of the seven HCs were excluded. Table 3  Distribution by sector of residence of health-facility (HF) based deliveries and C-sections in S. KayonzaSector of residenceEstimatedtotalpopulationEstimatedtotal birthsTotaldocumentedHF-baseddeliveriesDeliveriesat the DH(% of HF-baseddeliveries)Total C-sections(% of HF-baseddeliveries) † C-sectioncoveragerates (per 100estimatedtotal births)Range inby-cellC-sectioncoverage rateC-sectionsfor UterineRupture(% C-sections)Sector A 31 038 1183 544 149 (27.4) 59 (10.8) 4.99 3.0  –  8.7 1 (1.6)Sector B 24 953 951 961 330 (34.3) 112 (11.7) 11.78 10.1  –  16.7 0Sector C* 4437 169 147 24 (16.3) 13 (8.8) 7.69 7.7 0Sector D 17 408 663 502 144 (28.7) 49 (9.8) 7.39 4.9  –  10.2 2 (3.8)Sector E †  14 459 551 351 77 (21.9) 24 (6.8) 4.36 3.0  –  7.0 3 (12.5)Sector F 26 213 999 712 188 (26.4) 91 (12.8) 9.11 8.5  –  9.6 3 (3.3)Sector G 14 998 571 468 118 (25.2) 56 (12.0) 9.81 8.8  –  13.0 2 (3.6)Sector H †  23 642 901 790 NA 145 (18.4) 16.09 11.7  –  29.6 2 (1.4) * Only one cell in Sector C. † Sectors E and H significantly different from the other sectors ( P  <  0.001). 22  © 2012 Blackwell Publishing Ltd Tropical Medicine and International Health  volume 18 no 1 pp 18  –  26 january 2013 L. Sudhof   et al.  Local use of GIS mapping to improve rural C-section access
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!