No Woman Should Die Giving Life

Umeå University, Medical DissertationsNew Series No 1705, ISSN 0346-6612, ISBN 978-91-7601-239-0 Department of Public Health and Clinical Medicine Epidemiology and Global Health Umeå University, SE-901 87 Umeå, Sweden No woman should die while giving life Does the Health Extension Program improve access to maternal health services in Tigray, Ethiopia? Tesfay Gebregzabher Gebrehiwet Department of Public Health and Clinical Medicine Epidemiology and Global Health Umeå University Umeå 2015  Department of Public Health and Clinical Medicine Epidemiology and Global Health Umeå University SE-901 87 Umeå, Sweden Responsible publisher under Swedish law: the Dean of the Medical Faculty This work is protected by the Swedish Copyright Legislation (Act 1960:729) ISSN: 0346-6612 ISBN: 978-91-7601-239-0 Copyright © Tesfay Gebregzabher Gebrehiwet 2015 Cover photo and other photos: Tesfay Gebregzabher Gebrehiwet Electronic version available at http://umu.diva-portal.org/ Printed by: Print & Media, Umeå University, Umeå, Sweden 2015 This thesis is dedicated to my late (young) brother Mussie Gebregzabher Gebrehiwet, let his soul rest in peace. 0 Table of Contents Abstract in Tigringa 3 Abstract 9 Acronyms 12 Terminology 13 Prologue 15 Introduction 17 The global maternal health and health care situation 19 Maternal health care initiatives globally 21 Background 25 Profile of Ethiopia 25 History and geography 25 Socio-economic situation 26 Gender relations in Ethiopia 27 Health status 28 The Ethiopian health system and its organizational 29 structure … The Health Extension Programme 31 Maternal health care in Ethiopia 35 The Tigray region 37 Regional context 37 Maternal health care delivery system in Tigray 39 The maternal health care situation 41 36 Rationale for this thesis 43 Aim of the thesis 45 The overall aim 45 Specific objectives 45 The conceptual framework 47 1 Methods 51 Study area 51 The research design 52 Quantitative retrospective longitudinal study 53 Quantitative cross sectional study 55 Qualitative, Grounded theory approach 57 Qualitative, Thematic analysis approach 59 Ethical considerations 60 Main findings 61 The association of HEP with the levels and trends of 62 antenatal, delivery and postnatal care utilization from year 2003-2012 Regression analysis 64 Factors associated with antenatal care and delivery services …60 65 Women’s experiences in giving birth 68 Institutional delivery from health workers perspective 71 Discussion 75 Availability 76 Accessibility 77 Acceptability 78 Methodological considerations 81 The role of the researcher 84 Conclusion 85 Implications for further intervention 87 Implications for further research 89 Acknowledgements 91 References 95 2 Abstract in Tigrigna ሓፂር መግለፂ ፅንዓት ሽፋን መሰረታዊ ክንክን ጥዕና ምዕባይ ድሕንነት ኣዴታትን ህፃናትን ንኽረጋገፅ ዝለዓለ ተራ ከምዘለዎ ይፍለጥ፡፡ ይኹንደኣምበር ግልጋሎት ጥዕና ኣዴታት ብማዕረ ኣብ ምብፃሕን ብምዕሩይ ኣገባብ ኣብ ምሃብን ብዙሓት ሃገራት እናተፀገማ እየን፡፡ ነዚ ዘይምዕሩይን ማዕረ ዘይኾነን ኣዋህባ ግልጋሎት ጥዕና ንምምሕያሽ ኣብ ሃገርና (ኢትዮጵያ) ብ1994-1995 (ብአቆፃፅራ ግእዝ) ዝተኣታተወ ፕሮግራም ምግፋሕ ጥሙር ጥዕና ቤተሰብ (Health Extension Program/HEP) 10 ዓመታት ኣሕሊፉ ይርከብ፡፡ ፕሮግራም ምግፋሕ ጥሙር ጥዕና (HEP)፣ ሕብረተሰብ መሰረት ዝገበረ-ምክልኻል ሕማማትን ምስጓም ጥዕናን - ምትእትታው ግልጋሎት ሕክምና ቀለልቲ ሕማማትን ብፍላይ ድማ ንኣዴታትን ህፃናትን ትኹረት ብምግባር ስድራ ቤት ብምሕቋፍ ዝንቀሳቐስ ፕሮግራም እዩ፡፡ እዚ ፕሮግራም ንምትግባር ኣብ ክልል ትግራይ ልዕሊ 600 ኬላታት ጥዕና ተሃኒፀን፡፡ ልዕሊ 1200 ሞያተኛታት ጥሙር ጥዕና እውን ሰልጢነን እየን፡፡ ክልተ ሞያተኛታት ኣብ ሓደ ኬላ ጥዕና ተመዲበን ካብ 5000-7000 በዝሒ ንዘለዎ ሕብረተሰብ ግልጋሎት እናሃባ ይርከባ፡፡ እንተኾነ ግን ዝተፈላለዩ መፅናዕታታት ከምዘመላኽትዎ ግልጋሎት ኣዴታት ጥዕና (ወሊድን ድሕሪ ወሊድን) ትሑት ሽፋን ከምዘለዎ ይሕበር፡፡ ዕላማ እዚ ዝገበርናዮ መፅናዕቲ እውን ኩነታት እቲ ግልጋሎት ብኣሃዝ ንምዕቃንን ከምኡ እውን ምኽንያታት እቲ ትሑት ግልጋሎት ንምድህሳስን እዩ፡፡ 3 ኣብዚ መፅናዕቲ እዚ ምኽንያታት ትሑት ሽፋን ግልጋሎት ክንክን ጥንሲ ወሊድን ድሕረ ወሊድን እንታይ ምዃኑ ካብ ሕሉፍ ልምዲ (ተሞክሮ) ኣዴታትን ሰብ ሞያ ጥዕናን ብዝርዝር ብምድህሳስ - ፕሮግራም ምግፋሕ ጥሙር ጥዕና ንቐረብ ግልጋሎት ኣዴታት ጥዕና ንምምሕያሽ ዘለዎ ኣስተዋፅኦ እንታይ ከምዝመስል ተተንቲኑ ቐሪቡ ኣሎ፡፡ እቲ ቀረብ ግልጋሎት ጥዕና ሓሙሽተ መዐቀኒታት ብዘለዎ ፍሬም ዎርክ እዩ ተዳህሲሱ እቶም ሓሙሽተ መመዘኒታት 1. ቅርበት ትካል ጥዕናን ሰብ ሞያ ጥዕናን ንተገልገልቲ 2. ህልውነት ሰብ ሞያ ጥዕናን ናውትን ድሌት ተገልገልትን 3. ሰብ ሞያ ጥዕና ብተገልገልቲ ዘለዎም ተቐባልነት 4. ትካል ጥዕና ንድሌት ተገልገልቲ ንምዕጋብ ዘለዎ ድልውነት 5. ተገልገልቲ ወፃኢታት ሕክምና ንምሽፋን/ንምኽፋል ዘለዎም ድሌትን ዓቕምን እዮም፡፡ እዞም ሓሙሽተ መዐቀኒታት መሰረት ዝገበሩ ኣርባዕተ ዓይነታዊን አሃዛዊን ሜላታት ብምጥቃም ዝተኻየዱ መፅናዕትታት ኣብ ኣብ 4ተ ወረዳታት ትግራይ እዮም ተኻይዶም፡፡ ኣብቲ ቀዳማይ መፅናዕቲ ፕሮግራም ምግፋሕ ጥሙር ጥዕና ቤተሰብ (HEP) ኣብ ምምሕያሽ ክንክን ጥንሲ-ወሊድን-ድሕረ ወሊድን ግልጋሎት ዘምፅኦ ለውጢ ኣብ ሰለስተ ወረዳታት (ጋንታአፈሹም፤ ክልተ ኣውላዕሎ፣ ሕንጣሎ ዋጅራት) ንዝሓለፉ 10 ዓመታት ኣብ ዝተዋህበ ግልጋሎት ብምድራኽ እዩ ዳህሰሳ ተኻይዱ፡፡ ትኽክለኛነት እቲ ፀብፃብ እውን ካብ ትካላት ጥዕና ዝተልኣኸ ወርሓዊ ኣብ ወረዳ ምስ ዘሎ ፀብፃብ ብምንፅፃር ንኽረጋግፅ ተገይሩ እዩ፡፡ እቲ ካልኣይ መፅናዕቲ ኣብ ወረዳ ሰሓርቲ ሳምረ ካብ 19 ጣብያታት ካብ ዝተመረፃ 30 ቑሸታት ዕድሚአን ካብ 15-49 ዓመት ምስ ዝኾና 1115 ደቂኣንስትዮ ኣስታት 30 ደቓይቕ ዝወደአ ቃለ መሕትት ብምኽያድ እዩ ዳህሰሳ ተኻይዱ፡፡ ኣብ ወሊድ ግልጋሎት ዘሎ ልምድን ተሞክሮን ንምድህሳስ 51 ኣዴታት ዝተሳተፋሉ ሽዱሽተ ጉጅላዊ ምይይጥ በቲ ሳልሳይ መፅናዕቲ ምርምር ዝተፈፀመ 4 እንትኾን ምስ ሸሞንተ ሞያተኛታት ጥሙር ጥዕናን ኣርባዕተ ነርስ መዋልዳንን ቃለ መሕትት ብምኽያድ እቲ 4ይ ምርምር/መፅናዕቲ ተፈፂሙ እዩ፡፡ ውፅኢት ቀዳማይ መፅናዕታዊ ፅሑፍ ኣብዚ ቐዳማይ መፅናዕቲ እቶም ዝተአከቡ መረዳእታት ኣብ ሰለስተ ደረጃታት - ቅድመ ፕሮግራም-ፕሮግራም-ድሕረ ፕሮግራም ብዝብል ዝተመቐሉ እዮም፡፡ እቲ ቐንዲ ዕላማ ፕሮግራም ምግፋሕ ጥሙር ጥዕና ድሕሪ ምጅማር ዝተራእየ ለውጢ ንምፍታሽ እዩ፡፡ ኣብ ድሕሪ እቲ ፕሮግራም ካብ 2001-2004 ዓ/ም ብዝተኻየዱ ንጥፈታት ጥዕና ግልጋሎት ወሊድን ድሕረ ወሊድን ኣብ ኩለን ትካላት ጥዕና ካብ ዓመት ናብ ዓመት ልዑል ኣዝማሚያ እናርኣየ ከምዝኸደ ብስታቲስቲካዊ መረዳእታ ንምርግጋፅ ተኻኢሉ እዩ፡፡ ብተመሳሳሊ ኣብዚ ወቕቲ እዚ ኣዝማሚያ ግልጋሎት ክንክን ጥንሲ ኣብ ጥዕና ጣቢያታት ጥራሕ እናለዓለ ከምዝኸደ ውፅኢት እቲ መፅናዕቲ የረድእ፡፡ እቲ መፅናዕቲ ከም ዘረድኦ ሽፋን ክንክን ጥንሲ ብ1995 ካብ ዝነበሮ 28.2% ብ2004 ናብ 46.7 ክምዝለዓለ፤ ግልጋሎት ወሊድ ብ1995 ካብ ዝነበሮ 5% ዝነበረ ናብ 23% ከምዝደየበ ድሕረ ወሊድ ግልጋሎት እውን 11% ዝነበረ ናብ 41% ከምዝለዓለ ንምርዳእ ተኻኢሉ ኣሎ፡፡ ውፅኢት ካልኣይ መፅናዕታዊ ፅሑፍ ኣዴታት ግልጋሎት ክንክን ጥንሲ ኣብ ጥዕና ጣቢያ ንኽጥቀማ ቅርበት ትካላት ጥዕና ንመንበሪ ገዛውቲ፣ ሓዳር ምግባር፣ ልዕሊ 5 ዓመት ስሩዕ ትምህርቲ ምምሃርን ካብ ሕርሻ ወፃኢ ኣብ ካልእ ስራሕቲ ዝተዋፈሩ ሰብ ሓዳር ምህላው ወሰንቲ ኣካላት ምዃኖም በቲ ዝተገብረ ካልኣይ መፅናዕቲ ተረጋጊፁ፡፡ ብተመሳሳሊ ኣብ እዋን ክንክን ጥንሲ ንኣዴታት ምኽሪ ግልጋሎት ምሃብ ቅድመ ታሪኽ ዝንጉዕ ሕርሲ ወይ ሃልኪ ምንባር እውን ኣብ ትካላት ጥዕና ወሊድ ግልጋሎት ንኽመሓየሽ ወሰንቲ ኩነታት ከምዝኾኑ በቲ መፅናዕቲ ተረጋጊፁ፡፡ 5 ብሓፈሻ ፕሮግራም ምግፋሕ ጥሙር ጥዕና ቤተሰብ (HEP) ቀረብ ግልጋሎት ጥዕናን ሽፋን ክንክን ጥንሲ ወሊድን ድሕረ ወሊድ ግልጋሎትን ኣብ ምምሕያሽ ዝተፃወቶ ተራ ትርጉም ዘለዎ ምዃኑ በቲ ዝገበርናዮ መፅናዕቲ ንምርዳእ ተኻኢሉ ኣሎ፡፡ እንተኾነ ግን ባህላዊን ልማዳዊን ኩነታት (ኣብ ገዛ ክትወልድ ምድላይ- ነፍሰፁር ዓይኒሰብ ከይረኽባ ኢልካ ምእማን) - ተበቲኖም ዝሰፈሩ ነበርቲ ምህላው - ኣፀገምቲን ዓቐብ ቁልቁልን ጎቦታትን ዝበዝሖም መንገድታትን እኹል መጉዓዝያ ዘይምህላውን ቀንዲ ሃልኪታትን ዕንቅፋታትን እቲ ዝወሃብ ግልጋሎት ምዃኖም ኣብዚ መፅናዕቲ ተገሊፁ እዩ፡፡ ብተወሳኺ ኣብ ላዕለዎት ትካላት ጥዕና ዘይብሩህ ገፅን ሰሓባይ ኣቀራርባ ሰብ ሞያ ጥዕና ዘይምህላውን ተቐባልነት ዘይብሎም ባህሪያት ምንፅብራቕን ነቲ ግልጋሎት ዝዓዘዘ ዕንቅፋት ከምዘለዎ በቲ መፅናዕቲ ንምርዳእ ተኻኢሉ እዩ፡፡ ማይን መብራህትን ዝኣመሰሉ ትሕቲ ቕርፂ ኣብ ኬላታት ጥዕና ዘይምህላዉ ኣዴታት ኣብ ቀረበአን ዘሎ ትካል ጥዕና ንኽወልዳ ዘየተባብዕ ከምዝኾነ እውን ተሓቢሩ እዩ፡፡ ውፅኢት ሳልሳይን ራብዓይን መፅናዕታዊ ፅሑፍ ኣብ ሳልሳይ መፅናዕቲ ምስ ኣዴታት ብዝተገበረ ምይይጥ - ኣደ እትወልደሉ ቦታ ባዕላ ንኽትውስን ከምእነሓጎታት ዝመሰላ ዕድመ ዝደፍኣ ኣዴታትን ፀቕጢ (ተፅእኖ) ከምዝግበረላ እቶም መፅናዕቲታት ይሕብሩ፡፡ ዋላ አኳ ኣብ ትካል ጥዕና ብዛዕባ ምውላድ ኣዎንታዊ ኣረኣእያ ኣዴታት ዝዓዘዘ እንተኾነ ብዛዕባ ድኹም ኣዋህባ አገልግሎት ጥዕና ኣዝዩ ከምዘተሓሳስበን እቲ መፅናዕቲ ይገልፅ፡፡ ትካል ጥዕና ናብ መንበሪ ኣዴታት ዘለዎ ርሕቐትን መጓዓዓዚ ዘይምርካብ ዝኣመሰሉ ፀገማት ከምዘገድስወን እውን እቲ መፅናዕቲ ይሕብር፡፡ ኣብቲ ራብዓይ መፅናዕቲ ብወገን ሞያተኛታት ጥሙር ጥዕናን ነርስ መዋልዳንን ብዝተገበረ ምይይጥ - ኣብ ሆስፒታል ዝዋሃቡ ዝሐሹ ግልጋሎት ኣዴታት ካብ ታሕተዋይ ትካል ጥዕና ናብ ሆስፒታላት ሪፈር እንትበሃላ ኣብ ምጉዕዓዝ ብሕብረተሰብ ዝግበር ምትሕግጋዝ ከምኡ እውን እናዓበየ ዝኸይድ ዘሎ ግንዛበ 6 ሕብረተሰብ ኣዴታት ኣብ ትካል ጥዕና ንኽወልዳ መሳለጢ ከምዝኾነ እቲ ፅንዓት የረድእ፡፡ ዓቕሚ ምንኣስ ሞያተኛታት ጥሙር ሞያተኛታት ጥሙር ጥዕናን ነርስ መዋልዳንን - ድኹም ኩነታት ትካል ጥዕናን (ሕፅረት ናውቲ ጥዕና ምህላዉ - ትሕቲ ቅርፂ ዘይምምላእ) ፀገም መጉዓዝያን እውን ኣዴታት ኣብ ትካል ጥዕና ንኸይወልዳ ዕንቅፋታት ከምዝኾነ እቲ ፅንዓት የመላኽት፡፡ መጠቓለሊ እቶም ዝተጠቐሱ ዕንቅፋታትን ፀገማትን ብምንካይ ቀረብ ወሊድ ግልጋሎት ጥዕና ንምዕባይ እዞም ዝስዕቡ ፃዕሪታት ምክያድ ይግባእ፡፡ ኣብ ጎቦታትን ኣዝዩ ርሑቕን ኣፀገምትን ዝሰፈረ ሕብረተሰብ ካብቲ ልሙድ ዝተፈለየ ቀረብ ግልጋሎት ጥዕና ምሃብ (ንኣብነት ካብ ጥዕና ጣቢያ ኣዝየን ዝርሕቓ ኬላታት ጥዕና ክእለት ዘለወን ነርስ መዋልዳን (midwives) ምምዳብ - ኣደ ማእኸል ዝገበረ ግልጋሎት ንኽወሃብ ሰብ ሞያ ጥዕና ብዓቕሚ ንኽዓብዩ ምግባርን ነቶም ሓሙሽተ መዐቀኒታት ቀረብ ግልጋሎት ብምምላእ ኣብቲ ሕብረተሰብ ተቐባልነት እቲ ግልጋሎት ክዓቢ ምግባርን፡፡ 7 8 Abstract Introduction Ensuring access to universal primary health care is essential to secure a safe and pleasant motherhood and to provide compassionate care for mothers and newborns. However, inequalities in the access to maternal health services still remain a prominent problem in many countries. As part of reducing inequalities, Ethiopia launched the Health Extension Program (HEP) in 2003. The HEP is a community based program designed with a defined package of essential promotive, preventive and basic curative services targeting households, particularly mothers and children. Despite the construction of over 600 health posts and deploying more than 1200 Health Extension Workers (HEWs), preliminary data suggests a low utilization of maternal health care services. This thesis explores the HEP contribution in improving women’s access to maternal health care, and the reasons for the low use of maternal health care services from the perspectives of the involved actors in the Tigray region in Ethiopia. The five dimensions of access were used as a framework to explore the access to maternal health care utilization in this setting. Methods A total of four districts were included in the study. Both quantitative and qualitative methods were applied. In the first sub-study, we assessed the HEP and its association with change in the utilization of antenatal, delivery and postnatal care services. Retrospective longitudinal data for 10 years was extracted from three selected districts and checked for accuracy. Segmented linear regression technique was used to control the secular trends adjusted for correlation of the data. For the second sub-study, we conducted a cross sectional survey with 1115 women (aged 15-49 years who had given birth within five years prior to the survey period) to determine the prevalence of antenatal care and institutional delivery utilization and explore their determinant factors of low utilization. For the third sub-study, we conducted six focus group discussions (FGDs) with a total of 51 women to explore women’s experiences of childbirth and maternal care. An interview with eight HEWs and four midwives were carried out to capture health workers’ perspective on access to maternal health care services in the fourth sub-study. 9 Grounded theory for the former, and thematic analysis for the latter were used for the analysis. Main findings The finding of the first sub-study showed a statistically significant upward trend for delivery care (DC) and postnatal care (PNC) in all facilities during the HEP late implementation period (July 2008-June 2012). In addition, a substantial trend of antenatal care (ANC) service use was observed at health centres after the intervention. In the second sub-study, the determinant predictors for ANC utilization were: proximity to health facilities, to be married, ≥5 years of education and having non-farming husbands. The last three factors were also significantly associated with institutional delivery, but also lower parity, previous history of obstructed/prolonged labour and ANC counselling. Findings from the qualitative studies pointed out that elderly women influenced women’s decision making about where to give birth. Women were mostly positive about giving birth at health facilities, but were concerned about the poor quality of care, inaccessibility and unavailability of transport. From the health workers’ perspective: specialized performance of hospital services, community assistance during referral and an increased awareness among women regarding the benefits of giving birth at a health facility were perceived as facilitators for institutional deliveries. Poor perceived competence of HEWs, poor conditions of health care facilities and inaccessibility of transportation, among others, were perceived as barriers for giving birth at health facilities. Conclusion Overall, this research revealed a considerable contribution of the HEP in improving the access and coverage of maternal health services (ANC, DC and PNC). However, cultural traditions, scattered localities, mountainous roads without adequate transportation and low quality of care are still the major obstacles to accessing the services. Mechanisms need to be designed to enable health facility access of safe delivery for women in hard to reach areas, improving the proficiency of health workers and introducing a women centered approach that enhances acceptability of the services. 10 Original papers This thesis is based on the following papers: I. Gebrehiwot T, San Sebastian M, Edin K, Goicolea I: Health Extension Program and its association with change in utilization of selected maternal health services in Tigray region, Ethiopia: a segmented linear regression analysis. Submitted to PLOS one. II. Tsegay Y, Gebrehiwot T, Goicolea I, Edin K, Lemma H, San Sebastian M: Determinants of antenatal and delivery care utilization in Tigray region, Ethiopia: a cross- sectional study. International Journal for Equity in Health 2013, 12:30. III. Gebrehiwot T, Goicolea I, Edin K, San Sebastian M: Making pragmatic choices: women's experiences of delivery care in Northern Ethiopia. BMC Pregnancy Childbirth 2012, 12:113. IV. Gebrehiwot T, San Sebastian M, Edin K, Goicolea I: Health workers’ perceptions on facilitators and barriers for institutional delivery in Tigray, Northern Ethiopia. BMC Pregnancy and Childbirth 2014, 14:137. Reprints were made with permission from the respective publishers. 11 Acronyms ACRMMA Accelerated campaign to reduce maternal mortality in Africa ANC Antenatal care BEmNOC Basic emergency and neonatal obstetric care CBHC Community based health services CBPHC Community based primary health care CEmNOC Comprehensive emergency and neonatal obstetric care DC Delivery care DHO District health office EDHS Ethiopian demographic and health survey FGD Focus group discussion GTP Growth and transformational planning HEP Health extension program HEW Health extension worker HF Health facility HSDP Health sector development program ICPD International conference on population and development IMF International monetary fund MDG5 Millennium development goal 5 MHTF Maternal health thematic fund MMR Maternal mortality ratio MoH Ministry of health NHP National health policy PASDEP Plan for accelerated sustained development to end poverty PHC Primary health care PHCU Primary health care unit PMNCH Partnership of maternal, neonatal and child health PNC Postnatal care PPC Postpartum care PPH Postpartum hemorrhage RH Reproductive health SMI Safe motherhood initiative SSA Sub-Saharan African countries TBA Traditional birth attendant THB Tigray health bureau UN United nations VCHW Volunteer community health workers WHO World health organization WDA Women development army WHDA Women health development army 12 Terminology Antenatal care coverage: Number of pregnant women examined at least once during the current pregnancy, by a skilled health worker (nurse, midwife, health officer, doctor) and by HEWs for reasons related to pregnancy divided by the total number of live births. Skilled attendance delivery coverage: Number of pregnant women assisted by a skilled attendant/accredited health professional (nurse, midwife, health officer, doctor) who is trained to be proficient in the skills necessary to manage normal pregnancies, childbirth and the immediate postnatal period including the identification, management and referral of complications in women and newborn divided by the total number of live births. Postnatal care coverage: Number of women who seek care, at least once during the postpartum period (42 days after delivery) from skilled health workers (nurse, midwife, health officer, doctor) and by HEWs, for postpartum related reasons, divided by the total number of live births. Clean and safe delivery coverage: Deliveries assisted by HEWs divided by the total number of live births. This singles out the activities of the national HEP in which clean and safe delivery service by HEWs is included as one of their tasks. Basic emergency obstetric and neonatal care (BEmONC): The signal functions of BEmONC include: a) remove retained products after miscarriage/abortion (e.g. manual dilation, extraction and curettage), b) administer parenteral antibiotics, c) administer parenteral anticonvulsants for pre-eclampsia and eclampsia (such as magnesium sulphate), d) administer uterotonic drugs (such as oxytocin), e) perform assisted vaginal delivery (including vacuum extraction and forceps delivery), f) perform basic neonatal resuscitation (with ambu bag and mask) and g) manually remove the placenta. Comprehensive emergency obstetric and neonatal care: Perform all BEmONC signal functions (a-g) plus: perform surgery (cesarean section) and blood transfusion. Health Sector Development Program (HSDP): The strategic plan of the country designed to improve the health status of the Ethiopian society implemented through a series of five years programs since 1997/98. It emanated from the national health policy and is aligned with international commitments such as the MDGs. 13 Volunteer community health workers (VCHWs): Members of the community who are selected as the best persons for implementing HEP tasks subsequent to the training by HEWs. They are role models by being fast learners and acceptors to help HEWs to diffuse health messages leading to the adoption of the desired practices and behaviors by the community. Women development army/group (WDA): An approach of mobilization within a cluster of thirty households which is further subdivided into one to five networks of women. These women are trained and engaged in their localities to mobilize their peers (women of child bearing age) in the nearby community to be enhanced for increased health care seeking, particularly for ANC and institutional delivery. Service hours: Hospitals and health centres are open to provide health services 24 hours a day and seven days a week, whereas health posts are open only working days (from Monday to Friday) from 8:00-12:30 in the morning and 1:30-5:30 in the afternoon. HEWs are on call for any emergency situations such as referrals or for any first aid measures. 14 Prologue My years of work at health centres and hospitals as a registered nurse has given me the experience of providing skilled attendance to birthing women and resuscitating newborns. During my internship, in the delivery unit, I couldn’t believe my eyes when women were pinched and slapped by health workers during delivery. On top of the excruciating labor pain, the humiliation and grief that those women were subjected to by health workers was unbearable. Since then I have been touched and affected by the dehumanizing nature of treatment and as a professional I promised myself to provide compassionate care. I graduated from Asmara School of Nursing, Eritrea (former Ethiopia) as a registered nurse with a diploma certificate. After working a few years there I came to Tigray, Ethiopia. This was 1991 and after the military regime was defeated by Ethiopian Proletarian Revolutionary Democratic Front (EPRDF) and the leadership had been taken over by the transitional Government of Ethiopia. The freedom fighters who struggled 17 years for peace, accountability, justice and democracy were on their way to implement the strategies and policies which had been designed and developed during the period of the armed struggle. I was extremely glad to apply my knowledge and skills with people who were thirsty for health service access and I also felt like I was contributing to health sector development. To be honest, while I was working in Northern Tigray (Zalanbesa), I saw that the commitments of the health centre teams in dealing with various communicable diseases (such as malaria, measles and whooping cough) and provision of pregnancy and delivery care resulted in improving the wellbeing of the community. There were hard moments during outreach services in the villages. It was an unforgettable event when I struggled to save my own life during a flooding after sensitizing a community about the importance of institutional delivery. I suffered only minor cuts and bruises and the community’s respect towards professionals was encouraging and motivated me to want to build a strong connection with public health and research. Being from a low income family may also have motivated me to face challenges and struggle for better attainment in life. As part of improving my future carrier, I graduated as a BSc in public health 15 (health officer), from Gondar University and was assigned at Adi- remets, Wolkayt which is one of the most remote districts in Tigray. This was a time where I was challenged to deal with three main roles of the health service delivery. Providing clinical activities as a non- physician clinician, coordinating health promotion and disease prevention of the district and dealing with administrative issues of the health centre. However, the collaborative effort of internal and external stakeholders and the love and care of the community taught me how to harmonize the tasks. I was very inspired by the coordination and organized movement of the health centre team. We were following the principle of many as one and one as many in solving emergency problems. These memories are all shined brightly into my mind. On the other hand, falling down from a mule, travelling a long journey (48 kilometres) to reach a place where a vehicle is available and escaping a furious camel were some of the accidental events also in my memory. In the period of 2002-2003, I accomplished a MPH at Umeå University Sweden and soon after I was re-assigned to be head for the district health office (DHO) and then transferred to Tigray Regional Health Bureau as a HEP coordinator. My connection with Umeå ignited new ideas about research and community service in my mind. Due to my interest of research and community service, I applied to work in the University. In the meantime, I also worked for the World Health Organization (WHO) and United Nations International Children’s Fund (UNICEF) through contractual agreement for some months. After working for one and a half years as a lecturer in the University (2008-2010), I was appointed the head for the department of public health. It was a nice coincidence when Dr. Miguel San Sebastian from Umeå University came to visit his former PhD student in Tigray which also became an opportunity to discuss doctoral studies. To be honest, my position as HEP coordinator and the evident connections between my nursing and public health profession and the field work at the community level, and my engagement in the University motivated me to pursue a PhD and choose the topic of my research. 16 Introduction In 2000, the United Nations addressed the concern of maternal mortality reduction with the slogan “No woman should die while giving life”. In Ethiopia, the slogan captured the interest of the late Prime Minister Meles Zenawi who constantly repeated it in his speeches. Moreover, guided by this motto, great efforts towards the reduction of maternal mortality and morbidity were taken by the country, the Health Extension Program being the flagship of these efforts. This thesis aims to explore in which ways the efforts made by Ethiopia aimed at realizing “No woman should die while giving life” resulted in better accessibility to maternal health care services for women. Ensuring access to universal primary health care is essential to secure a safe and pleasant motherhood and to provide compassionate care for mothers and new-borns. However, inequalities in the access to maternal health services still remain prominent in many countries [1- 2]. The implementation of community based primary health care programs (CBPHC) is considered one of the best strategies to improve the access of populations to health services. CBPHC is an approach to identify health care problems and subsequently to design, implement and evaluate health intervention activities aimed to address such problems. The process is led by the community, and involves key actors, such as government agencies, non-governmental organizations, donors and private stakeholders [3-6]. CBPHC is believed to address the major health problems of the society providing promotive, preventive and curative services based on the socio-cultural, political and economic context of the country. The CBPHC strategy is enhanced through principles such as community participation, self-reliance, social awareness and decentralization of the health system to lower administrative units. The long term goal of CBPHC is to achieve universal access to health care to the larger community, through the provision of acceptable, accessible, available, well organized and appropriate and affordable health care, including maternal and child health care [7-9]. Ethiopia was one of the four African signatory countries for implementation of the primary health care (PHC) strategy in the late 1970’s [10]. However, poor governance, a centralized top down management and a fragmented health service delivery affecting the 17 country for decades, among other factors, hindered the real implementation of PHC until 1993 [11-13]. That year, the government of Ethiopia developed the National Health Policy (NHP) emphasizing the democratization and decentralization of the health structures and aiming for universal access and equity of health care to the population. To achieve these goals, the government formulated a 20 year plan (1997/8-2014/5) under the name of Health Sector Development Program (HSDP). Along these years, the HSDP has been performing continuous interventions in policy and programs to strengthen the country’s health care system. In 2003, as part of the community based primary health care intervention, the Health Extension Program (HEP) was established as the flagship of the HSDP. The HEP is a new program designed with a defined package of essential promotive, preventive and basic curative services targeting households, particularly mothers and children in a community. Improving access to maternal and child health services was among the major objectives of the HEP [13-15]. Despite the empowerment and involvement of the community in implementing the HEP, the access to curative services remains low. The limitations in fulfilling the communities’ expectations in terms of access to curative services might be a challenge to the HEP strategy which focuses mainly on promotive and preventive services. The northern state of Tigray has been the leading region in the country in creating a solid platform for delivering the HEP. In 2005, 600 health posts, staffed with 1200 Health Extension Workers (HEWs) were working in the region. This was possible due to the coordinated effort between the leadership at the regional level and a strong commitment of the district level officials together with the rural communities. However, many challenges were encountered during the initial stage of the program implementation. Some of these main challenges included: 1) incompleteness of equipment, supply and furniture at health posts, 2) absence of clear guidelines on career structure of HEWs and their relation with other health workers, 3) weak transport and communication services and 4) absence of reading and reference materials in the local language [13, 16]. When this thesis was planned, despite the government’s effort in implementation of the HEP, maternal health care utilization 18 remained poor and the reasons for the low coverage of services were not sufficiently clear. This thesis consists of ten chapters and gives a comprehensive description regarding the endeavor of the HEP in improving the access of the maternal health care services to the rural population of Tigray. The global maternal health situation, maternal health initiatives including the safe motherhood initiative and other global strategies to improve maternal health are presented in chapter 1. The socioeconomic context, the health system and its organizational structure, both at national and regional level, with a focus in the health extension program are described in the second chapter. Then, the rationale of the thesis is included in chapter 3. The next chapters of the thesis focus on the four studies that constitute the core of this thesis. The aim and the four specific objectives are presented in chapter 4 while the conceptual framework linking the four objectives with the concept of access is elaborated in chapter 5. The research design, consisting of a combination of different methods, and the ethical considerations are described in chapter 6. The major findings of this thesis are then presented in chapter 7. The discussion in chapter 8, the conclusion in chapter 9 and the implications of the results for practice and future research in chapter 10 are considered in the last section of the thesis. The global maternal health and health care situation The global community started to pay significant attention to women’s health when the United Nations (UN) charter was signed during the establishment of the World Health Organization (WHO) in 1948 [17]. Later, maternal health was recognized as one of the core elements of primary health care (PHC) during the international conference of Alma Ata in 1978. Consequently, improving access to maternal health services in low and middle income countries continued to be a concern of governments and international agencies [7, 18]. This can be illustrated by the growing attention that safe motherhood has received at international conferences and summits over the last two decades. Despite being prioritized on the international agenda, more than 289,000 women around the globe still die because of pregnancy and delivery related complications. Maternal mortality has continued to decrease in south and east Asian countries and southern and central Latin America [19-20]. Almost all (99%) maternal deaths occur in 19 developing countries. The overwhelming tragedy is particularly evident in sub-Saharan African (SSA) countries, accounting for around two thirds of the global burden (62%) (Figure 1). The estimated risk of women dying due to pregnancy and childbirth is 1 in 38 in SSA in comparison to high income countries where the risk is 1 in 3400 [19-24]. Due mainly to the burden of communicable diseases, such as HIV, the number of maternal deaths in SSA from 1990-2000 showed an increased trend hence, an evidence of decline was observed later (till 2013) due to enhanced anti-retroviral uptake [19- 20]. These statistics might underestimate the number of maternal deaths in the region due to the absence of vital registration system and we most certainly have unrecordable cases which were supposed to capture maternal deaths. Figure 1. Maternal mortality ratio by countries, trends 1990-2013 [19]. Several factors related to culture, context and health systems contribute to the overall maternal mortality situation and include: poverty, gender inequality, low utilization of services, absence of social pressure to improve access, lack of coordination, weak information and fragile health systems. At the individual level, women’s socio-demographic characteristics such as: older age, parity, not being married (single, divorced, separated), poor access to health services and low economic and educational status - have been commonly identified as major risk factors for the low utilization of maternal health care that leads to increased maternal mortality [25- 31]. 20 Besides death due to unsafe abortion, most maternal deaths occur during labor, delivery, or 24 hours after birth because often serious complications during these stages cannot be reliably predicted. The medical causes of maternal mortality are mainly hemorrhage, obstructed labor, eclampsia, severe pre-eclampsia and sepsis. Beyond the preventable medical causes, three types of delays, which mainly focus on the social determinants of maternal deaths, have been mentioned as explanatory causes throughout the literature for many decades [22]. The first delay refers to delay in deciding to seek appropriate medical care and socio-cultural norms and traditions are influential in this delay. These may include religion, influence from elderly, the consideration of delivery as a natural event, fear of the health facility setting, poor decision making power of women and low levels of education among women. Besides, poor access to information for women about the place of delivery, failure to identify different barriers (cost and geographical challenges) and poor community involvement are also aggravating factors [22, 32-33]. In addition, there are delays in reaching appropriate obstetric facilities (second delay) and delays due to lack of adequate care after reaching the facility (third delay). Both delays are connected to the supply factors which are mainly related to health policies and health system organization. The most frequently mentioned factors in the literature regarding the second delay include the unavailability of transport, poor quality and delay of referral. Long waiting time at health facilities, shortage of supplies, absence of skilled health workers, unfriendly approach of health workers, poor responsiveness and poor quality of care are usually part of the third delay [22, 32, 34- 35]. Health care providers’ lack of sensibility towards traditional norms of the society on the process of childbirth – as promoted in the CBPHC approach is also reported as a key barrier for institutional delivery [36-37]. Overall, the primary bottleneck of the supply factors in poorly resourced countries are due to the overwhelming failure of health systems, lack of good governance, poor political commitment, lack of attention to the ethos of CBPHC, lack of donor coordination and lack of harmonization between stakeholders [28, 31-34, 38]. Maternal health care initiatives globally As part of the ongoing effort to improve maternal health and reduce maternal mortality, numerous global initiatives have been implemented in the past decades. The main aim of the global Safe 21 Motherhood Initiative (SMI) conference in Kenya (1987) was to cut down maternal mortality by 50% by the year 2000. The SMI blended the major components considered to be key for maternal mortality reduction - family planning, access to quality care during pregnancy and throughout the process of child birth and early postnatal care [39-42]. The SMI with its successive initiatives has been influenced by different emerging evidence and competing interests over the years [43-44]. In the 1970s-1980s, screening for risk factors during pregnancy and training of traditional birth attendants (TBAs) were considered useful approaches to improve maternal health services and were implemented in low income countries as core strategies. Despite the fact that donors dispensed their cash on TBA training and screening pregnant women (for high risk factors such as pregnancy induced hypertension or gestational diabetes), such strategies were not effective enough to curb maternal mortality [45-46]. As a result, community-based actions and training programs for TBAs were criticized and often discontinued. Immediately, the focus of the programs shifted towards increasing access to skilled delivery attendance and emergency obstetric care. The assumption was that since it was not possible to predict obstetric complications during child birth, women should deliver close enough to health care facilities that could manage these complications. This meant focusing on implementing Basic Emergency Obstetric and Neonatal Care “BEmONC” and Comprehensive Emergency Obstetric and Neonatal Care “CEmONC”, (see pages 13-14 for definitions) [42, 47-48]. The availability of these services is measured by the number of facilities that perform the complete signal functions in relation to the size of the population. According to the WHO, the minimum acceptable level for a population of 500,000 is five BEmONC facilities with at least one providing comprehensive care [49]. Parallel to these events, a series of international conferences and summits focusing on sexual and reproductive health and gender equality were held where the issue of enhancing safe motherhood and reducing maternal mortality were also addressed [50-53]. For instance, the International Conference on Population and Development (ICPD) in 1994 appealed to the concerned global bodies to focus their attention on reducing maternal mortality and improving women’s health through ensuring access to: family 22 planning and reproductive health services, maternal health-care services, skilled attendance at delivery and emergency obstetric care [54]. In 1995, the Beijing World Conference on women reaffirmed that the universal access of health care, women’s empowerment and their right to health needs to be ensured as a priority action. It further emphasized the position of women to be considered beyond motherhood, the importance of gender equality for improving women’s health and education, and the contribution of women to the welfare of the family and to the development of the society [51]. Lastly, improving maternal health and reducing maternal mortality was adopted by all United Nations (UN) countries in 2000 as one of the eight Millennium Development Goals (MDGs) aiming to reduce the maternal mortality ratio (MMR) by three quarters between 1990 and 2015 [55]. The most recent global Partnership of Maternal Neonatal and Child Health (PMNCH) was established in 2005. It advocates for an integrative and life cycle approach to maternal health care, focusing on ensuring access to maternal health care at the institutional level, but coordinated with care provided at the household and community levels [56, 57]. As part of strengthening PMNCH, the 194 UN member states and several partners renewed their commitment to save women and newborns lives during the 67th WHO assembly in 2014 [42, 57- 61]. In Africa, various programs are currently being introduced aiming to strengthen women’s sexual and reproductive health rights through empowering and enabling women to decide whether and when to get pregnant and improving access to quality care through pregnancy and child birth. Since 2009, the Maternal Health Thematic Fund (MHTF) under The African Union Commission (CARMMA), an accelerated campaign to reduce maternal mortality in Africa, supports more than 43 African countries in reducing maternal mortality through building the national capacity for strengthening and sustaining emergency obstetric care. The share of expenditures allocated to African countries by MHTF in 2012 amounted to 62% compared to 57% in 2011 [62]. Despite the efforts made in SSA countries, including Ethiopia, to improve access of maternal health services utilization, the reduction 23 of maternal mortality remains insufficient - from 990/100,000 live births in 1990 to 510/100,000 live births; a 49% reduction in 2013 compared to the MDG5 goal that aims for a reduction of 75% [19]. 24 Background Country profile of Ethiopia History and geography Ethiopia is an ancient country recognized as Abyssinia, the initial origin of human being [63]. The Axumite Empire during 1st year after death of Christ (AD) in Ethiopia was predominantly Christian. It was one of the great powers of his time along with Persia, Rome and China. The Islam religion was introduced around 615 AD when the country offered shelter to followers of the prophet Mohammed [64- 65]. The country has a rich cultural diversity and is home to more than 80 ethnic groups and 83 different languages. The country has maintained the status as an independent nation free from colonization and preserves its unique calendar, timing and script. Ge’ez is an exclusive script which is still used by the Ethiopian Orthodox Tewahedo Church at present. Amharic and Tigrigna languages are derived from the Ge’ez script, where the former is the official national language of the country [66]. Ethiopia is, after Nigeria, the second most populous country in Africa with 84,320,987 inhabitants. It is situated in the horn of Africa and is bordered by Djibouti in the east, Sudan in the west, Somalia in the southeast, Eritrea in the north and northeast, and Kenya to the south [67-68]. Ethiopia’s total surface area is about 1.1 million square kilometres (Figure 2). Around 84% of the country’s population lives in rural areas and it is recognized as one of the least urbanized countries in the world. The average annual growth rate is 2.6%. Nearly one quarter of the population (23.4%) are women of reproductive age (15-49 years). The average lifetime fertility has declined from 6.4 births per woman (1990) to 5.4 births in 2005 and 4.8 births in 2011 [69-70]. 25  Nine regional states - Oromia - Southern nations and nationalities - Amhara - Tigray - Afar - Somali - Benshangul - Gambella - Harari Two admin. cities - Addis Ababa - Diredawa Figure 2. Ethiopian map with nine regions and two administrative cities [69]. Socio-economic situation Prior to 1993, the centralized nature of the political structure led to Ethiopia having low socio-economic and technological development. Since 1993, the transitional government of the country introduced a decentralization system of administration where decision making power is shared at all levels -federal government, regional states and district administrative units [13, 70]. At present, the country is a federal state with nine regions and two administrative cities (Figure 2). The regions are divided into 80 zones, 551 districts (woredas) and about 12,000 villages (kebeles/tabias) [67-69]. A district is the basic decentralized administrative unit whereas the village is the smallest administrative unit in the governance structure. Farming is the major source of livelihood for the population in Ethiopia and the country’s economy is agriculture centred. Agri- business accounts for 43% of the gross domestic product and 80% of exports. 26 Coffee, livestock, gold, leather products, dried legumes, vegetables, flowers and oil seeds are the major export items of the country. The plan for accelerated and sustainable development to end poverty (PASDEP) and the Growth and Transformational Planning (GTP) have been strategies designed to attain the MDGs and to maintain rapid economic growth and ultimately to end poverty [69, 71-74]. The country has achieved an effective and efficient utilization of resources by introducing sector wide approach programs. These programs have aimed at improving health care planning and implementation through harmonization and alignment -one budget, one plan, and one report and all donors are obliged to align their budget program along with the governmental goals. Poverty head count index in the country has reduced from 49.5% in 1994/95 to the level of 29.2% in 2009/2010. Recently, the World Bank announced Ethiopia as among the five African countries with the fastest growth of economy [75]. However, socio-economic inequality is still of major concern in the country. Ethiopia is ranked at 173 of 188 countries on the human development index (HDI) based on the last human development report [76-77]. The initiatives that have been implemented in the economic sector only contributed to increase the HDI value from 0.284 in 2000 to o.435 in 2013. Despite the efforts, Ethiopia couldn’t cross the HDI value beyond 0.550 which is the cut- off point for low human development [75, 78]. Gender relations in Ethiopia Gender can be described as a set of expected and incorporated features, characters, and behavior patterns that differentiate women from men socially and culturally and relations of power between them. The power relationship and decision making in the Ethiopian context have been dominated by men [79-81]. Traditionally, the patriarchal and cultural norms of the country have considered men as breadwinners and responsible for satisfying the needs of the household members. Women have only been considered responsible for domestic work and as caretakers of their husbands and children. Despite women’s involvement in all aspects of society, their position has not always been recognized. Instead, their poor decision making power, lack of economic stability and low level of education has been contributing to low social status. In response to the existing gender inequalities, a women’s national policy was formulated in 1993 in order to promote gender equity. 27 Later on, in 2005, the government re-established the women’s affairs office at the Ministry level. The main aim of the Ministry is to strengthen gender development and to reduce gender inequality, to set up strategies for the improvement of women’s access to health and education, and to empower their socio-economic positions. Partly and probably due to these policies, a significant number of women, mostly from urban areas (some but fewer from rural areas) are taking a major part in all kinds of employment and men are also starting to take a greater part in domestic life mainly in towns and cities [79, 82]. One Ministerial and three state Minister positions are currently headed by women [79-81]. The proportion of women who shared the labor participation increased to 78.2%, and 25.5% of the seats in the parliament are held by women as reported in the gender gap development report in 2014 [82]. However, the country still needs to work more on sharing the power relations between men and women to reduce the gender inequality, for instance: the prevalence of life-time intimate partner physical violence against women is reported to be 39%, and 12.5% during the last three months, in 2013 [79]. Ethiopia is ranked 121 and 126 for gender inequality index and gender development index respectively of 178 countries. Health status Tuberculosis, malaria, HIV/AIDS and non-communicable diseases such as cardiovascular diseases, diabetes mellitus and cancers along with injuries are among the major contributories of high morbidity and mortality in Ethiopia [71, 83]. The country has made good progress towards children’s health. Infant mortality rate (IMR) has been reduced from 88/1000 in 1990 to 59/1000 live births in 2012, respectively. Child mortality has been decreased from 184/1000 live births to 88/1000 in the same period. More than 90% of child deaths are due to pneumonia, diarrhea, malaria, neonatal problems, malnutrition and HIV/AIDS, or combinations of these conditions [71]. In terms of sexual and reproductive health, the prevalence of new HIV infections has dropped from 12.4% in 2003 to 1.3% in 2012. Maternal mortality has also been reduced from 871 in 2000 to 471 per 100,000 live births in 2012 [71, 84]. More than 80% of maternal deaths are due to direct obstetric causes: infections/sepsis 47.1%, hemorrhage 29.4%, severe pre-eclampsia/eclampsia 7.6%, obstructed /prolonged labor and ruptured uterus 2.9%; and complications from 28 unsafe abortions 2.9%. Indirect obstetric causes such as malaria, HIV, anemia and cardiac problems account for the remaining proportion [71, 84]. Selected health indicators of the country are presented in Table 1. Table 1. Basic health indicators of Ethiopia for years 2011/2012 [69, 84]. Health indicators Value Total fertility rate 4.8 per woman Contraceptive prevalence use rate 40.4 % Antenatal coverage first visit 89.1 % Deliveries attended by skilled attendants 20.4 % Postnatal coverage 44.5 % Maternal mortality ratio 471/100,000 live births Infant mortality rate 59/1000 live births Under five child mortality 88/1000 live births Pentavalent three immunization coverage 84.9 % Full immunization coverage 71.4 % Life expectancy for males 60 years Life expectancy for females 62 years The Ethiopian health system and its organizational structure Due to a fragile health policy and shortage of resources, the health access for the majority of the population has been very limited for decades. The few health institutions have been poorly equipped and budgeted with a health system dominated by curative services and with more than 50% of health facilities located in urban areas. A critical and prolonged shortage of human resources has been another major challenge [13, 85-87]. In 1993, the transitional government of Ethiopia developed a National Health Policy (NHP) aimed at addressing the health rights, values and the needs of the less advantaged population. The NHP emphasized the principles of democratization and decentralization to strengthen the health system and to ensure health promotion, disease 29 prevention and equitable, efficient and effective health care delivery. Health system strengthening was considered as an instrument of social justice and equity as well as central for social and economic development. As part of the NHP implementation, the HSDP was developed during 1997/1998-2014/2015 rolling with five-year program periods. The HSDP gives attention to harmonizing the goals, objectives and targets set by all stakeholders (governmental and non- governmental organizations, donors, private partners and the community). Thus, the objective of HSDP focuses on attaining universal access to PHC for the whole population and eventually achieving the MDGs [12-13]. The successive implementation conducted to achieve MDG5 is illustrated in Figure 3. - Decentralization of the health care system at all levels (1993 and beyond). Changing health care tier system (1996, 2012), establishing new programs and designs (2003 and beyond) - Provision of transportation facilities, strengthening referral and outreach services, strengthening the national medical supplies and equipment and maintenance system (1998-2014). - Accelerated expansion of PHC service from 2003- 2014 (construction of new health posts, health centres and upgrading of health facilities). - Accelerated training of health officers, midwives, health extension workers, integrated emergency surgery and obstetrics graduates and other health professionals from 2003-2014. Figure 3. Successive implementation of strategies for achieving MDG5. As part of the ongoing efforts to improve the package of PHC services across the country, the structure of the health delivery system was reorganized from six to four and then to a three-tiered system [12-13, 88]. The current system involves three levels comprised of primary health care units (PHCU), general hospitals and specialized hospitals. A PHCU consists of five satellite health posts, one health centre and one primary hospital serving populations of 5,000, 25,000 and 100,000 respectively. The secondary level -general hospital- serves 30 1,000,000 and the tertiary hospital serves 5,000,000. The health centre is the immediate referral health facility for the health posts and it is usually staffed with health officers, midwives, nurses, medical laboratory technologists, pharmacy professionals, and administrative staff. Nurses and midwives are trained for two and four years respectively after finishing high school and are granted diploma and degree certificates. To promote decentralization and ensure community participation, decision making power is shared among regions and districts [12-13]. The Federal Ministry of Health plays a role on designing policy matters, guidelines, setting forth frameworks and high level capacity building issues. The Regional Health Bureaus are proactively involved in establishing the policies and operating guidelines and play facilitating, coordinating and mentoring roles. The fundamental roles of managing, coordinating and decision making activities at the PHCU level are made by the district health offices (DHOs). They are in charge of supervising and coordinating the primary health care units. The authority and decision making power of the districts is up to the level of recruiting and employing health workers, building infrastructures, procuring essential supplies for health facilities through fair allocation of resources to the lowest administrative units. This practical implementation is aligned with community based PHC principles. The Health Extension Program In an effort to increase the access to the health care system and improve the health status of the most vulnerable rural population, the government of the country started the implementation of the health extension program (HEP) in 2004 [15, 71]. The fundamental philosophy of the HEP is to transfer ownership and responsibility for preserving people’s own health to households by conveying health knowledge and skills to them [71, 89]. Model households assist to diffuse information aimed to encourage adaptation of desired practices and behaviors in the community. HEP assumes that health behavior can be enhanced in communities by creating model families that share their best experiences and encourage their neighbors to learn and practice similar behavior [71, 90]. It is designed to improve equitable access to preventive essential health interventions through community-based health services. It is also intended to achieve significant basic health care coverage through the provision of a staffed health post to serve an area of approximately 3,000 to 5,000 31 people - a kebele. Each kebele has one health post and two health extension workers (HEWs). Health posts receive technical support from a closer health center and they are located nearest to other public institutions to facilitate coordination among public service providers. The HEWs are the community members’ first point of contact with the health system providing integrated promotive, preventive and basic curative health services. They aim to ensure the continuity of care throughout pregnancy, child birth and postnatal care [14, 91]. Influential members (such as women associations, civil society organizations and religious and clan leaders) of the community and district offices representatives (health, education, agriculture, women’s affair and capacity building) select the HEWs. The criteria for recruitment include: being female, to be known to respect societal norms and values, to be known for community participation, knowledge of the local language and enthusiasm to serve the community after training. Three reasons can be mentioned for selecting females to be HEWs: 1) most of the HEP packages are related to issues affecting mothers and children in the setting, 2) contextually, in the absence of husbands at the household, communication between female HEWs and the women was thought to be easier and culturally acceptable, 3) employing salaried female health workers by the government at village level was also considered to empower women. The HEWs are recruited from their own village and are trained for one year after they completed high school. Soon after their graduation, HEWs are usually further mentored about midwifery services for one month at health centres and hospitals under the supervision of nurses and midwives. They also receive refresh training once a year. The HEP has been implemented throughout Ethiopia, with more than 34,000 HEWs already trained and deployed in over 14,000 health posts since 2004 (Figures 4 and 5). Immediately after HEWs are deployed, they provide training to members of the community for 96 hours during a period of six months. During the training HEWs are engaged in selecting model families who are considered to have best performance. Afterwards, the selected model families are recruited as volunteer community health workers to collaborate with HEWs to mobilize the community [13, 71, 89]. 32 Figure 4. Trends of health posts construction by consecutive years (2004-2010) (Source HSDP III, document of 2010/2011). Figure 5. Trends of graduated and employed HEWs by consecutive years (2004-2010) (Source HSDP III document of 2010/2011). Initially the HEP was developed for agrarian inhabitants, and then subsequently, it has been tailored and scaled up into urban and pastoral communities. In rural areas, the HEP is composed of 16 intervention packages categorized into four areas: “hygiene and environmental sanitation" (seven packages), "family health" (five packages), "disease prevention and control" (three packages), and "health education and communication" (one package), for details please refer to Figure 8 [14-15, 71]. The role of HEWs and operational definition of health post are described in Boxes 1 and 2. 33  Image 1. A standard health post. Box 1. Health post organization - Operational centre of the HEP. It functions under the supervision of the Woreda Health Office, kebele administration with technical support from the nearest health centre. - Built and owned by the community with support from the regional government. - One of the HEWs is elected as a cabinee member of the local community (kebelle), for empowering the decision making of the health unit. Box 2. Selected tasks of HEWs - Conduct home visits and outreach Image 2. HEWs at a health post. services. - Provide referral services to health centres. - Identify, train and collaborate with VCHWs. - Provide reports to Woreda Health Offices. - Provide health promotion and preventive services related to the 16 components of the health packages. - Provide preventive services such as immunization, antenatal care and family planning distribution. - Provide basic curative services, treat for malaria after rapid diagnostic tests, provide oral rehydration salts for children with diarrhoea and perform first aid measures. Despite the fact that stretched efforts have been made through effective implementation of HEP at the grass root level, it is still hard to reach rural families. In response to this, the government established the Women Health Development Army (WHDA) network in 2011. This network was required to create community ownership by coordinating various sectors and mainly involving the community from the planning up to the evaluation stages of the health sector and other divisions. A total of 257,153 WHDA have been formed all over 34 the country. Prior to this network, volunteer community workers have been recruited and trained to support HEWs in mobilizing the community for preventive health programs such as immunization and delivering key health information messages designed by the HEWs. The volunteer community health workers (VCHW) and TBAs are combined with the WHDA and some of the active members of VCHWs hold a leading role in the WHDA. Currently the majority of women in the villages are accessing information about birth preparedness and benefits of institutional delivery from WDAs at their locality [92-94]. Maternal health care in Ethiopia Antenatal, delivery and postnatal care activities are provided at all levels of health facilities. Antenatal care (ANC) is provided by nurses and midwives (skilled professionals) at health centres and hospitals and by HEWs at health posts. Women are assisted by skilled health attendants for labor and delivery only at health centres, hospitals and private clinics, whereas, clean and safe delivery is provided by HEWs. Similarly, women receive postnatal care at home by HEWs, and by midwives and nurses at health centres and hospitals. The maternal health components of a PHC unit include service performed at both health centres and health posts: immunization, hygiene and sanitation, family planning, antenatal care, health information provision and counselling service and referral linkage to higher health facilities. Apart from these services, health centres provide skilled delivery attendance, basic emergency obstetric and neonatal care (BEmONC), safe abortion and post abortion care services, whereas, health posts occasionally deal with clean and safe delivery services. HEWs also provide postnatal care services at the household level. The primary hospital is the higher level referral centre of the primary health care unit which deals with all preventive and curative type of services including basic and comprehensive emergency obstetric and neonatal care [13-71]. As indicated in Figure 6, the Annual Ethiopian Health Indicator Bulletin shows an increasing trend of users flow for the three types of services in the last eight years, although not very notorious for delivery care [69, 84]. 35 Figure 6. Trends of ANC, DC and PNC at national level year 2005- 2012. According to the national surveys, the trends of ANC, DC and PNC have increased from 46%, 13% and 15% in 2005 to 89%, 20% and 45% in 2012 respectively. Similar figures regarding ANC visits (85%) have been reported in different studies. In terms of the prevalence of institutional delivery, studies provide different figures: i.e. Abera et al. in 2011 reports a 16% prevalence of institutional delivery, whilst the Bulletin and UNDP report it to be 20% (20.4%) [26-27, 30-31, 95-100]. Despite this increase, the prevalence of institutional delivery is still low when compared to the country’s target of 62% for 2010- 2015. The reasons reported for the low coverage of this services by the annual HSDP bulletin of 2011/2012 include: a) absence of services 24 hours a day and seven days a week in most health facilities, b) inadequate numbers and high turnover of experienced and senior midwives, c) shortage of drugs, supplies and equipment forcing mothers to buy gloves, drugs and intravenous fluids, d) poor delivery room environment, e) absence of neonatal unit and f) weak referral system [93, 101]. According to the Ethiopian Demographic Health Survey (EDHS) 2011, lack of transport (71%), lack of money (68%) and distance (66%) were the major reasons mentioned by women for not accessing postnatal care services [95]. 36 The majority of women still utilize the traditional model of child birth at home where they are assisted by traditional birth attendants (TBAs), grandmothers, mother in-laws, or neighbors. TBAs continue to assist home deliveries due to the fact that they are culturally accepted and trusted by the community. However, nowadays, TBAs formally collaborate with the WDAs and work under supervision of HEWs to counsel and refer women to nearby health centres for delivery [12, 13, 71, 93, 101-102]. The Tigray region Regional context Tigray is one of the nine regional states in Ethiopia. It is bordered by Eritrea to the North, Sudan to the west, Afar region to the East and Amhara region to the south and south west. The total area of the region is about 54,569 square kilometers. The total population is 4,806,843, and almost 80% are rural inhabitants. The proportion of males and females is 49.2% and 50.8% respectively and the majority of the inhabitants are Christians. The Tigray region is divided into seven zones and 47 districts of which 35 are rural and 12 urban. There is one specialized referral hospital as well as six zonal hospitals, seven district hospitals, 214 health centres and more than 613 health posts in the region [67, 94]. Figure 7 shows the map of Tigray and the regions where the different sub-studies in this thesis were conducted. 37  Study districts for paper I, III and IV Study district Study district for paper I for paper II Figure 7. Map of Tigray region. Like the national health status, the most common health problems of the region are communicable diseases, such as pneumonia, diarrhea, malaria, tuberculosis and HIV/AIDS. Childhood problems are mainly aggravated due to malnutrition. Low awareness of communicable diseases, poor personal hygiene and lack of environmental sanitation are among the major problems identified by the TRHB annual report. A recent study from the demographic surveillance site in Tigray has also reported the emergence of non-communicable diseases in the region [83, 94]. Although some studies indicate a positive impact from HEP in some of its sub-components such as community participation, community health awareness, immunization coverage, and hygiene and sanitation, still the impact in access to maternal health services is low. Studies reveal the existing poor infrastructures of health posts in the region: electricity, tap water supply, and fixed telephone line (all HEWs have mobile phones but the connection is poor) were only available in 5%, 8%, and 21% of the health posts respectively [90, 99, 103-105]. 38 The basic health indicators of the region are indicated in Table 2. Table 2. Basic health indicators of the Tigray region year 2011/2012 [69, 84]. Health indicators Value Total fertility rate 4.6 per woman Contraceptive prevalence rate 28.6 % Antenatal coverage first visit 50.1 % Deliveries attended by skilled attendants 26.3 % Postnatal coverage 49.6 % Maternal mortality ratio 266/100,000 live births Infant mortality rate 64/1000 live births Under five child mortality 85/1000 live births Pentavalent three immunization coverage 86.3 % Full immunization coverage 74.6 % Life expectancy for males 60 years Life expectancy for females 62 years Maternal health care delivery system in Tigray The organizational health system in Tigray is more decentralized compared to the other regions of Ethiopia. The health bureau communicates directly with the district health offices, after the zonal health structure was eliminated in 2002. Assisting and supervising the PHC units (health posts, health centres and hospitals) is the primary role of the regional health bureau staff. The health posts are staffed with two HEWs enrolled in providing health care services both at the health post and in the community, focusing on the 16 HEP components that include maternal and child health. The Tigray Health Bureau (THB) designed the sub-divisions of the HEP in a star form which is thought to be easily understandable (Figure 8). 39 Figure 8. Rural HEP’s components simplified around a star structure. Regarding maternal and child health, HEWs provide family planning, immunization for children, antenatal, and postnatal care and may attend deliveries. They also give health education on maternal nutrition, hygiene and sanitation. HEWs work in collaboration with the WDAs, TBAs and other voluntary community health workers are expected to support health education activities in the communities through house to house visits [14, 46]. HEWs use a family folder which helps them identify the maternal health service needs of the family when they visit their households [106]. Though, HEWs were supposed to assist labor either at health posts or at home during the initial phase of HEP, their work pattern now is shifting towards referring to health centers and hospitals. The reasons for the change include: 1) increased community awareness regarding benefits of health facility delivery, 2) improved accessibility through organized community effort in transportation, and 3) enhanced acceptability of services at the health center and hospital due to the implementation of humanized approach to childbirth in these settings. They use the family folder to trace women who reached their 40 expected date of delivery and refer them to nearby health centres. WDAs work closely in collaboration with HEWs in mobilizing and convincing the community on the health facility benefits in general and institutional delivery in particular. The health centres and primary hospitals need to have well equipped and staffed facilities to provide BEmONC and CEmONC for saving women with obstetric complications. Currently, the staffing pattern of nurses and midwives at health centres is increasing in number and BEmONC has been introduced in the majority of the health centres. Some remote health centres in Tigray are staffed with integrated emergency obstetrics/surgery clinicians for provision of CEmONC services. Hospitals are facilities where BEmONC and CEmONC services are provided and they are staffed by skilled health attendants (physicians, midwives and nurses) and integrated emergency surgery clinicians. Maternal and child health services are free of charge through the health care financing strategy; however the costs such as transport and other charges are expected to be covered by users. Despite the significant effort of the Tigray regional health bureau and the government to improve maternal health care, a high proportion of home deliveries are still practiced in the region which is a barrier to maternal mortality reduction and a major reason for not achieving MDG5 [94]. The maternal health care situation The last national survey report indicates that 50.1% of all pregnant women in Tigray received at least one ANC visit at health centres and hospitals (45.6% by nurse or midwife and 4.3% by a doctor) whereas 14.4% were at health posts by HEWs [69]. During ANC visits, health professionals’ will check the position and presentation of the baby and fetal heartbeat. The components of ANC services also include: 1) blood pressure measured (86%), 2) blood sample taken for hemoglobin, HIV testing, blood group and RH factor (66.9%), 3) urine sample taken for bilirubin, sugar and other microscopic investigations (48%), 4) iron tablets provided (33.6%) and 5) information conveyed on complications and danger signs during pregnancy (27.5%) [93]. Different figures for institutional delivery have been reported for Tigray, ranging between 11.6% [84] (estimates from the national 41 report) and 32.2% [94] (estimates from the regional report). The reason for the variation might be different sampling methods and periods; EDHS collected data from the 2006-2010, whereas, the regional estimate included the recent progress of the implementation from the years 2010-2012. The majority of rural women (83.3%) do not receive postnatal care in the region. In 2011, the Ethiopian Demographic Health Survey (EDHS) reported that only 16.7% of women received one PNC visit of which, only 12.3% were seen by doctor/midwife or nurse and 0.8% by HEWs in the first two days after birth. In 2012, the THB bulletin reported an increased proportion of PNC use by women, from 5% in 2000 to 49.1% in 2012 [94]. The reason for the increased proportion might be the cumulative effort of the health work force, mainly HEWs, women development groups and the community. 42 Rationale for this thesis The HEP is a community based health care approach that has been implemented since 2004. The THB has been investing in building over 600 health posts and deploying more than 1200 HEWs. UN agencies and other international stakeholders have also been committed to support this process. However, the mere presence of health posts and availability of HEWs within the rural areas is not enough to ensure the fulfilment of community expectations and their access to health services. In 2010 (when this research project began), few studies assessing the HEP had been conducted. The findings of existing studies indicated a positive experience of the program by the population, but also poor working conditions [107, 108]. Despite the continuous effort of the THB in improving access to health care services for the population, especially since 2004, preliminary data pointed out a low utilization of maternal health care services. This thesis therefore emerged to understand on the one hand, the impact of the HEP regarding women’s access to maternal health care, and on the other hand, the reasons for the low use of maternal health care services from the perspectives of the involved actors. A final goal of the thesis was to be able to suggest possible approaches for policy makers to develop appropriate responses to increase the access to these services. 43 44 Aim of the thesis The overall aim The overall aim of this thesis is to assess if the Health Extension Program is improving maternal health care services’ utilization in Tigray region, Northern Ethiopia. Specific objectives 1) To evaluate the association of the HEP with the levels and trends in utilization of antenatal, delivery and postnatal care services during the period 2003-2012 (sub-study I). 2) To determine the prevalence of maternal health care utilization, particularly antenatal and delivery care, and to explore its determinants among rural women aged 15-49 years (sub-study II). 3) To explore women’s experiences and perceptions regarding delivery care (sub-study III). 4) To explore health service providers’ perceptions on factors influencing utilization of institutional delivery (sub-study IV). 45 46 The conceptual framework: Access to Community Based Primary Health Care In 1978, the Alma-Ata declaration on PHC recognized that the world’s health issues required more than just hospital-based and physician centered policies. The declaration called for a paradigm change that would allow governments to provide essential care to their population in a universally acceptable manner [7]. One key focus of the strategy was how to increase the access to the health services for the most vulnerable groups in the society in an affordable and effective way. The issue of access was considered again as a key dimension by the 2000 WHO Health Report on health systems performance when discussing how to achieve the goals of a health system: to improve the health of the population, to establish a fair share of financing and to ensure the responsiveness of the system. Penchansky & Thomas have defined access as “the fit between the user and the health care system” [cited in 109]. They proposed five dimensions that the concept of access should include: a) availability, namely -the relationship between the existing services in volume and type, and the clients demand, b) accessibility refers to the physical location of providers and health facilities versus setting of the users, c) accommodation relates to the extent the resources of the health care system are organized in a manner to meet users expectation, d) affordability is measured by the users’ ability and willingness to pay for the services and e) acceptability, how the users feel satisfied receiving health service from the providers and vice versa [109]. This framework has been used in different studies to assess access to health care in low-income settings such as in the case of malaria treatment in Tanzania and Ethiopia or the use of maternity waiting homes to promote institutional delivery in Liberia [109-110]. Figure 9 represents the application of the access model of Penchansky & Thomas as have been applied in this thesis (Figure 9). 47  Health Extension Programme 1. Excellence in health care Core values of the health sector delivery development program: community first, 2. Excellence in leadership and collaboration, commitment, change, governance om trust, professional development 3. Excellence in health infrastructure and resources Primary health care unit - Women Hospital - Family - Volunteer community Access health workers (VCHWs) Health centre - Women development Army - Communities Health post Increased utilization and coverage Improved status of women’s health Figure 9. Conceptual framework of access modified from Thomas and Pechansky [109]. The HEP and its association in utilization of ANC, DC and PNC was assessed in the first sub-study capturing mainly the dimensions of accessibility and availability. Prevalence of ANC and institutional delivery and its determinants, analyzed in the second sub-study, focused on accessibility, availability and acceptability. The experience of users’ regarding institutional delivery (sub-study III) explored three dimensions of access (accessibility, availability and acceptability) from a qualitative perspective. The perceptions of health workers in relation to women’s place of delivery (sub-study IV) mainly reflected the accessibility, and acceptability domains of the access concept. A summary of the relations between the objectives of this research and the three dimensions of access is presented in Table 3. 48  Table 3. Sub-studies, level of health care and framework of access to maternal health service utilization. Sub-studies I-IV Level of Research Dimensions primary question of access in health focus care unit I. The association of How does the trend of Accessibility Health post HEP with the levels ANC, delivery and Availability and trends in postnatal care look? Health utilization of selected centre What is the impact of HEP maternal health in improving access to services from year maternal health care utilization? 2003-2012 II. Determinants of Health post What is the prevalence of Accessibility antenatal and delivery Health antenatal and delivery Availability care utilization centre care? Acceptability What are the determinants of institutional delivery? III. Women’s Health post Why women give birth at Accessibility experiences and Health home or at a health Availability perceptions regarding centre facility? Acceptability delivery care What factors influence Hospital women when choosing the place of delivery? IV. Health workers Health post What is the perception of Accessibility perceptions of Health health workers on Availability facilitators and centre institutional delivery? Acceptability barriers to What are the facilitators Hospital institutional delivery and barriers to institutional delivery? 49 50 Methods Study area The contextual description of Tigray can be found under Regional context (page 37-38). Of the 35 rural districts in Tigray, ten were first selected purposively from different geographic areas. Then, three districts were excluded for inaccessibility reasons and another three were not included due to large missing data. Finally, a total of four districts (Saharti-Samre and Hintalo-wajirat from south east zone and Kilte-Awlaelo and Ganta-Afeshum from eastern zone) were included in the study. The longitudinal data was extracted from Kilte- Awlaelo, Ganta-Afeshum and Hintalo Wajirat districts. Saharti-Samre is a rural and remote district where the second study was conducted. Kilte-Awlaelo and Ganta-Afeshum are districts where the voices of women and health workers were captured. These four districts have an estimated population of 526,130 of which 50.2% are females. According to the year 2012 THB report, women of reproductive age (15-49 years) constitute approximately 121,010 (23%) of the population and the number of deliveries in 2011-2012 was estimated to be 19,993. The first visit of ANC coverage was reported as 76.4%. The skilled delivery attendance, clean delivery and postnatal care (PNC) proportions in the districts were reported 23%, 12.7% and 46.7% respectively (Table 4). (See page 13 for definitions of skilled delivery and clean delivery). Table 4. Maternal health services coverage in the study districts, 2011-2012 (annual report). Name of Distance Population Eligible* ANC Skilled % of PNC district from 1st visit delivery clean visit Mekelle in % in % delivery in % Ganta 120 Kms 100,261 3810 57.2 12.3 13.6 30.3 Afeshum Hintalo 35 Kms 172,222 6544 94 29.2 10.2 69.4 Wajirat Kilte 45 Kms 112,788 4286 60.2 34.7 14.7 40.3 Awlaelo Samre 55 Kms 140,859 5353 94 15.9 12.2 47.9 Saharti Total 526,130 19,993 76.4 23 12.7 46.7 * Eligible: Number of women expected to receive the services based on the Federal Ministry of Health data 51 The total number of health posts and health centres in the four districts is 71 and 24 respectively. In average, the ratio of nurses and midwives (diploma level) to the population ranges from 1:4,200 (nurse), 1:21,527 (midwife) of Hintalo Wajirat to 1:3,342, 1:5,276 of Ganta-Afeshum, respectively. All health posts are staffed with two HEWs, the details of human resources in each district is illustrated in Table 5. Table 5. Number of health facilities in the districts with type and number of health workers (for level of diploma-12+2 and BSc degree 12+4, please read page 31), [12, 13]. District HPs HEWs HCs Type and number of health workers at health No. of centres HCs with BEmONC Nurses Midwives Health officers 12+2 12+4 12+2 12+4 Ganta 16 40 5 30 3 19 0 3 5 Afeshum Hintalo 18 40 7 41 2 8 3 4 7 Wajirat Kilte 17 32 5 34 0 15 0 5 5 Awlaelo Samre 18 28 5 35 1 10 2 5 5 Saharti Total 69 140 22 140 6 52 5 17 22 The research design For this thesis, quantitative and qualitative methods were applied in an attempt to answer each of the four objectives using the most appropriate methodology [111]. The combination of both methods enhanced the trustworthiness of the findings. Longitudinal retrospective and cross sectional study designs were employed for sub-studies I and II respectively. The qualitative approaches utilizing grounded theory and thematic analysis were used for sub-studies III and IV (Table 6). 52 Table 6. Study objectives, methodological approaches and analytical methods used. Specific Research Data collection Analytical objective design techniques approach I. To assess the Quantitative Documents Segmented linear change in retrospective review carried out regression utilization of longitudinal study in three districts; analysis selected maternal ANC, DC and health services PNC data and its retrieved from association with 2002-2012 the HEP II. To determine Quantitative cross 1115 women Multivariate the prevalence of sectional study interviewed logistic regression maternal health through a analysis care utilization structured particularly questionnaire antenatal and delivery care and explore its determinants among rural women aged 15- 49 years III. To explore Qualitative study Six focus group Grounded theory women’s discussions with experiences and 51 women in two perceptions districts regarding delivery care IV. To explore Qualitative study In depth Thematic analysis health service interviews with 12 providers’ health workers (4 perceptions on midwives and 8 facilitators and HEWs) in two barriers for districts utilization of institutional delivery Quantitative retrospective longitudinal study For sub-study I, a rapid assessment to check the availability of the necessary information in district health offices was first carried out. Eight out of 34 districts of the region were visited. Health profiles and 53 reports were reviewed for checking data availability. A checklist which answers the research question was prepared. Then, three districts which fulfilled the requirement of the check list were chosen. The type of services provided (AND, DC and PNC) and variables such as: name of the district, health facility, date, facility code, and total population of the catchment area were included in the checklist. Due to absence of electronic monitoring recording system and poor health management information system in the archive office of the districts, the process of data collection was lengthy. Data collection was carried out by two pairs of people at each district. From the district health office, the person in charge of the archive office and the maternal health expert were fully engaged in sorting the relevant documents and handing over to the researchers. Then, the data was labelled and recorded after a proper check for its completeness and accuracy. Retrospective longitudinal data from 10 years was extracted from the three selected districts (61 health facilities of which 16 were health centres and 45 health posts) and checked for accuracy. Then it was aggregated on a quarterly basis and exported to the Stata software for analysis after being cleaned in an excel spreadsheet. The analysis was performed independently (by HC, HP and as total) under the three phases of the intervention periods named as pre (the period before the program started, July 2002-June 2005), immediate (initiation of HEP, July 2005-June 2008) and late intervention (when the program runs smooth, after July 2008). Because July 2005 was the time where HEP implementation in the study districts started with minimum resources, it was selected as the starting point of the HEP. Since the nature of the longitudinal data is prone to bias, segmented linear regression technique was used to control the secular trends adjusted for correlation of the data [112-114]. The linear regression employed in the analysis was: Yt = βo+β1*time+β2*intervention+β3*postslope+et Yt indicates ANC, DC or PNC variables at time t where time is a continuous variable which was subdivided in three periods of the intervention named as pre-intervention (July 2002–June 2005), early intervention (July 2005–June 2008) and post-intervention (after July 2008). Two models were designed for running the analysis 54 in Stata. To estimate the level and trend of changes related to the duration after July 2005, Model 1 was used. The changes during the post intervention period were estimated by Model 2. Autocorrelation was considered using the Prais Winston estimator. In this model, the base line level of the outcome variable and the estimated value of the structural trend in utilization are represented by βo and by β1 respectively. Whereas, the estimated level of the immediate impact after intervention is denoted by β2 and the change in trend after the intervention by β3. Quantitative cross sectional survey study To determine the prevalence of ANC and institutional delivery and to explore its determinants a cross-sectional study was conducted. Women aged 15-49 years (n=1115) who had given birth within five years prior to the survey period participated in the study. The sample size was determined using a single population proportion formula. Multi-stage cluster sampling technique was employed. Of a total of 23 tabias (lowest administrative unit), four were excluded due to difficult roads (Images 3 and 4 illustrate the mountainous and hard to reach localities). Image 3. Hard to reach areas. 55 Image 4. Mountaineous landscape. Random sampling was carried out among the remaining 19 tabias for sorting the respected villages. Population proportion allocation to size sampling was employed to select 30 clusters. A random selection method was used to identify the first house for beginning the data collection. The questionnaire was developed based on national and international literature, national guidelines, field observations and common local practices in order to better respond to the research question of the study. The data was collected by 15 HEWs through administered interview and supervision was performed by four experienced nurses. The two outcome variables were ANC use and place of delivery. ANC use was defined as at least one ANC visit to a health facility (HF) by a pregnant woman. For place of delivery, the home and health facility options were available. Socio-demographic and obstetric characteristics such as respondent’s age, marital status, education, occupation, presence of health facility in a village, gravidity, previous ANC use, history of obstetric problems and counselling and information provision on benefits of institutional delivery during ANC were included as predictor variables. The quality of the data was ensured by training the enumerators, pre-testing the questionnaire and performing close follow up and supervision. Data collection was carried out from August to September 2009. 56 Data was entered into EPI-info after meticulous revision by the supervisors and the principal investigator. Then it was coded, cleaned and imported into the software Stata version 10 [115]. Cross tabulations were performed to assess the relationship and the significance between the outcome and predictor variables. Then, after executing univariable analysis, the significant predictor variables were included in a multivariable logistic regression analysis. Odds ratio was used to explain the association between the outcome variable and the risk factors. The presence of collinearity among predictor variables was also checked. Qualitative, Grounded theory approach Women’s experiences and perceptions about giving birth at home and at a health facility were explored in two rural districts. Initially, two FGDs were conducted and a preliminary analysis carried out. Then, six additional FGDs were carried out and analyzed. The data collection was carried out from September 2010 to January 2011. Potential participants were identified by HEWs and invited to the health posts. A total of 51 married women with age range of 15-40 years were invited, 27 who experienced giving birth at home and 24 who were assisted by health workers during delivery at health facilities. To obtain adequate disclosure of opinions and to include maximum number of eligible participants’ women who gave birth three years prior to the research participated in the FGDs. For capturing various experiences, women who lived close to urban centres and those residing in remote areas were equally invited. Only women who had never been serving as part of VCHWs were selected in the FGDs to avoid social desirability bias. The FGDs were conducted in a setting chosen by the participants, and the moderator tried to ensure a friendly environment. An interview guide to capture reasons for women seeking and not seeking ANC and delivery care was developed, and additional issues emerging during the first discussions were further explored during the subsequent FGDs following an emergent design. (Image 5 illustrates the FGD conducted with women). 57 Image 5: FGD with women. An explanation on the purpose of the study was given to all participants by the researcher. Afterwards oral informed consent was obtained from women to participate in the FGDs. The local language (Tigrigna) was used for carrying out the FGDs. The discussions were digitally recorded and transcribed verbatim. Notes were also taken for complimenting the data. All FGDs lasted between 90 and 120 minutes. The researchers continued to undertake the FGDs until saturation was reached, that is new information relevant to our research questions was no longer emerging [111]. The process of transcription was done verbatim, word by word, which contributed to capturing the pronounced feelings, opinions and expressions of the participants. The transcribed records were translated from Tigrigna to English and the texts were read very carefully several times for grasping the meaning of the respondents. The translated transcriptions were imported into the Open Code software [111] for managing the data during the analytical process. Texts were separated line by line and open coding was conducted. Codes were generated in two ways: translating ordinary terms into concepts, where expressions were chosen up directly from the data and used as open codes (“in-vivo”) and also by constructing codes from the texts (“in-vitro”) [111]. The team of researchers reviewed the data repeatedly and negotiated to select the codes, meanwhile, 58 moving back and forth to the data was executed to advance the coding, then similar codes were clustered together to form subcategories and categories that were refined, until a core category emerged. A constant comparison method was applied during the analysis. Qualitative, Thematic analysis approach To capture the health workers’ perspective on access to maternal health care services by women in Tigray, HEWs and midwives were invited to participate in interviews. All potential participants were selected by district health officials from various categories of health facilities (health posts, health centres and hospitals) in order to enrich the findings. Informants with different training backgrounds (ranging from 1-4 years after high school) were included aiming to explore a variety of experiences from different perspectives. Four midwives and eight HEWs participated in the interviews. This study shared similar study setting, duration of the study, steps and procedures of participants’ selection as described in the previous section about FGDs. An interview conducted with a midwife is shown in Image 6. Image 6: An interview with a midwife. The researcher requested permission from the participants for the interview as well as for digitally recording it. All participants agreed to 59 the request. Then, for each interview comfortable places that encouraged interactive discussions were chosen based on informant preferences. The research interview guide for these informants included semi-structured open ended questions covering various issues focusing on enabling and hampering factors for institutional delivery. These included: the role of decision making processes (by elderly women, family, husband) during delivery and what factors encouraged women to give birth at home or at health facility. Careful verbatim transcription of the recorded interviews in Tigrigna was conducted by the principal researcher and then translated into English and organized with the help of the Open Code software [111]. The analysis was guided by the framework of the three delays model [22, 116]. Firstly, the parts of the interviews linked to each delay were identified and coded. Secondly, for each delay codes referring to facilitators and barriers were identified and grouped in two different clusters. Thirdly, through revising the emerging codes, the initial three delays themes were fine tuned in the light of the new information. Ethical considerations This study was approved by the University of Mekelle, Research Ethical Review Committee of the College of Health Sciences, Northern Ethiopia. A permission letter was received from THB and district health authorities. Participants were informed about the purpose and procedures of the study. Written or verbal informed consents were obtained from all participants. Confidentiality and privacy of participants were guaranteed and they were assured that they had the right to withdraw from the interview or survey at any time. 60 Main findings This chapter is structured according to the four objectives of the thesis and a summary of the main results is presented in Table 7. Table 7. Summarized findings of the four objectives. Objective Summarized findings Association of the - A statistically significant upward trend found for DC Health Extension and PNC in all facilities during the HEP late Program with the implementation period (July 2008-June 2012). levels and trends - The flow of first ANC visit increased by 6.4 fold in of antenatal, health posts during the late implementation period but delivery and not in health centres. postnatal care utilization Prevalence and - 54% of the respondents visited ANC once. Women determinants of close to health posts used the services more frequently. ANC and - 4.1 % of women delivered institutionally. institutional - 3.6% births were delivered by skilled delivery delivery attendants. - 7% of women of those who gave birth at home were assisted by HEWs. - Women who received information about institutional delivery used more often health facilities. Women’s - Women are in favor of home delivery due to cultural experiences on tradition and faith. delivery care - Elderly women influenced the decision making process, preferring home delivery. - Positive opinion about delivery at health facility, but concerned about the poor quality of care, inaccessibility and unavailability of transport. - HEWs, TBAs and husbands supported women on their struggle for a safe delivery. Perception of - Women development group efforts have a positive health workers on effect on increased community awareness (F). facilitators (F) and - Community willingness to solve transportation barriers (B) of problems (F). institutional - Poor decision making power for women to choose delivery institutional delivery (B). - Poor competence of HEWs and some midwives in delivery (B). - Poor conditions of health facilities, lack of electricity and water facility in the delivery room (B). - Inaccessibility of transportation (B). 61 The association of HEP with the levels and trends of antenatal, delivery and postnatal care utilization from year 2003-2012 During the ten year period (July 2002-June 2012), the consultations on the three types of services - ANC, DC and PNC- increased constantly, particularly after the late-intervention period (July 2008- June 2012). Increases were higher for ANC and PNC at health post level whereas for DC at health centers. The trend of ANC, DC and PNC increased from 28% to 47%, 5% to 23% and 11% to 41%, respectively (Table 8). Table 8. Frequencies of ANC, DC and PNC users in both HFs (HCs and HPs) July 2002-June 2012 Year Eligible % ANC % DC % PNC July 02-June 03 22,501 28.2 5 11 July 03-June 04 23,096 25.5 3 5 July 04-June 05 21,219 29.5 4 10 July 05-June 06 (early 20,145 28.3 4 7 intervention) July 06-June 07 17,670 46.1 7 18 July 07-June 08 18,112 40.9 6 14 July 08-June 09 (late 18,564 49.5 9 19 intervention starts) July 09-June 10 19,029 52.0 13 29 July 10-June 11 19,505 57.5 16 27 July 11-June 12 19,993 46.7 23 41 NB: Population: Total population of three study districts based on population census (1994 to 2006 and then census of 2007), *ANC=antenatal care, DC=delivery care, PNC=postnatal care Eligible: Number of women expected to receive the services based on the Federal Ministry of Health data. The flow of ANC visits in health posts increased by 2.85 and 6.4 fold immediately after the intervention (July 2005-June 2008) and during the late intervention period (July 2008-June 2012) respectively compared to the pre-intervention period (July 2002-June 2005). This pattern however was not observed at the health centres (Figure 11). 62 Figure 11. ANC visits of women in both health facilities, Years 2002- 2012. The number of women assisted at health centres and those assisted by HEWs at health posts for delivery were three times higher and 11.9 fold respectively in the late-intervention period when compared to the pre- and early intervention stages (Figure 12). Figure 12. Women assisted for DC in both health facilities, Years 2002-2012. 63 Attendance to PNC services in health posts increased by 2.7 and 6.4 fold immediately after the intervention and during the late intervention period respectively compared to the pre-intervention period. But that pattern was not seen at health centres (Figure 13). Figure 13. PNC visits of women in both health facilities, Years 2002-2012. Regression analysis During the early intervention, the number of ANC consultations increased by 116.88 visits per quartile (period of three months) at health posts which was not significant, hence, a decrement was observed at health centres. A substantial change in trend was observed at health centres after the intervention (p=0.009), but this pattern was not seen in health posts (p=0.76) (Model 1, it can be found on Paper 1, Table 3). Regarding delivery care, Model 1 indicates a decrement of consultations per quartile in both health facilities (non-significant trend), whereas for postnatal care an increased number of consultations in both health facilities was observed, though this trend was also non-significant. Model 2 was built to detect the changes related to the late intervention period. Overall, a positive statistically significant upward trend was found for DC and PNC in both (health centres and health posts) facilities (p=<0.01). An increase by 122.75 and 326.47 of ANC consultation visits per quartile of the year (period of three months) at health centres and health posts was observed respectively. A positive 64 trend was shown in ANC at health centres (p=0.04), but not at health posts (Table 9). Table 9. Model 2 regression analysis (corrected for auto correlation) of attendants for ANC, DC, PNC at health centres, health posts and both, July 2002−June 2012. Service B0 B1 (Time) B2 B3 ANC Total 1412.00 19.50 (0.11) 456.78 (0.07) -2.74 (0.90) HC 1400.04 -18.70 (0.04) 122.75 (0.05) 37.18 (0.04) HP 9.85 38.41 (0.00) 326.47 (0.006) -39.54 (0.01) Delivery Total 321.62 -15.82 (0.52) -160.97 (0.21) 54.50 (0.038) care HC 208.56 -3.67 (0.46) -139.03 (0.11) 49.60 (0.000) HP -6.83 6.18 (0.001) 91.55 (0.01) 11.46 (0.001) Post Total 315.37 14.08 (0.28) -41.19 (0.84) 84.38 (0.002) natal HC 333.10 -4.03 (6.0) -148.88 (0.21) 45.65 (0.003) care HP -21.92 18.50 (0.01) 107.78 (0.37) 37.26 (0.007) NB: for comparison the analysis with Model 1, please refer to Paper IV in the appendix. For an explanation of Bo, B1, B2 and B3 please refer to page 54-55. The number of consultations for delivery care at health posts during the early intervention period increased by 91.55 visits per quartile (p=0.01), but decreased at health centres, though not being statistically significant. The trend (post slope) for both (health centres and health posts) in the late intervention period, was positively significant (p=<0.01) (Table 9). Finally, in the early intervention, the model showed a non-significant decrease of consultations to PNC services in both health facilities. However, a positive significant change in an upward trend was observed in health centres and health posts during the late intervention period (HC: p=0.003), (HP: p=0.007). Factors associated with antenatal care and delivery services Of 1115 women, 1113 (99%) responded to the survey. The mean age of the participants was 30.4 years. Only 3.3% of the participants had completed secondary education. Fifty four percent of women reported 65 at least one ANC visit during last pregnancy (Table 10). The main reasons for women not using ANC were: being apparently healthy (32.7%), lack of awareness of the benefits (28.2%), feeling shame (16.7%), work load (13.4%) and health facility too far (12.5%). Regarding ANC services, women mentioned that abdominal examination and checking blood pressure and fetal heart beat were the most valuable components of pregnancy check-ups. Table 10. Socio-demographic characteristics of women in the sample - Saharti-Samre district (n=1113). Individual variables Number of women (%) Age groups 16-29 473 43 30-39 482 43 40-50 158 14 Marital status Single 31 2.8 Widowed 29 2.6 Married 986 89 Divorced 67 6 Education Illiterate 880 79 1-4 Grade 126 11 5-12 Grade 107 10 Parity 1-4 593 53 5-7 371 33 8-11 149 13 Husbands’ Occupation Farmer 981 88 Government employee 31 3 Daily laborer 50 4 Merchant 51 5 Health facility in village No 687 62 Yes 426 38 Total deliveries Home 1067 96 Health facility 46 4 Health center 41 3.6 Health post 5 0.5 ANC use No 511 46 Yes 602 54 66 Table 11. Women’s characteristics and its association with ANC and DC for their recent birth in Saharti-Samre district (n=1113). Individual Received Adjusted Place of Adjusted variables ANC Odds Ratio delivery Odds Ratio n (%) (95% CI) Home n (95% CI) (%) Marital status Single+Wido 19 (1.7) 1 55 (4.9) wed Married 546 (49.0) 2.57 945 (84.9) (1.44-4.58) Divorced 37 (3.3) 2.78 67 (6.0) (1.31-5.89) Education No education 452 (40.7) 1 854 (76.7) 1 1-4 Grade 64 (5.8) 0.96 123 (11.1) 0.72 (0.65-1.42) (0.20-2.55) 5-12 Grade 86(7.7) 3.18 90 (8.1) 2.56 (1.85-5.47) (1.08-6.01) Parity 1-4 320 (28.8) 1 560 (50.3) 1 5-7 202 (18.1) 1.16 361 (32.4) 0.39 (0.88-1.55) (0.17-0.93) 8-11 80 (7.2) 1.28 146 (13.1) 0.40 (0.87-1.88) (0.11-1.48) Health facility in village No 325 (29.2) 1 666 (59.8) 1 Yes 277 (24.9) 1.83 401 (36.0) 1.59 (1.41-2.38) (0.77-3.27) Husbands Occupation Farmer 499 (44.8) 1 957 (86) 1 Others 103 (9.2) 2.26 109 (9.8) 3.84 (1.43-3.58) (1.78-8.29) Attended ANC No 504 (45.3) 1 Yes 563 (50.6) 1.45 (0.48-4.34) ANC advice No 634 (57.0) 1 Yes 433 (38.9) 3.08 (1.21-7.84) Difficult/prolonged labor No 837 (77.4) 1 Yes 200 (18.5) 10.0 (4.87-20.6) 67 The factors associated with a first ANC visit after running a multivariable logistic regression were: to be married (OR=2.57, 95% CI: 1.44-4.58) and divorced status (OR=2.78, 95% CI: 1.31-5.89), having accomplished 5-12 years of education (OR=3.18, 95% CI: 1.85- 5.47), closeness of the health facility to women’s village (OR=1.83, 95% CI: 1.41-2.38) and having husbands with a non-farming occupation (OR=2.26, 95% CI: 1.43-3.58) (Table 11). Of the whole sample, only 46 (4.1%) gave birth at health facilities, whereas, the remaining reported giving birth at home. The most frequent reasons for home birth were: easy labor (64%), health facility too far (4.7%), and transport problems (4%). Of those women who gave birth at home, 53.2%, 40% and 6.8% were assisted by non-qualified persons (elderly women, mother-in-law, relatives, neighbors), TBAs and HEWs, respectively. Among a few women who gave birth at health facilities, their reasons were: saves mothers life (31.2%), health facility is clean (30.6%), bleeding will not occur (21%), retained placenta will not be encountered (12.5%) and health facility supports labor (10%). Women’s experiences in giving birth From the FGD’s with women, “making pragmatic choices” emerged as a core category linked to three other categories. Women in this region were aware that complications might emerge during delivery and they “aim for safer deliveries” (category one). In their struggle to deliver as safe as possible, they have to choose between home delivery - exemplified by the category “embedded in tradition”- and the modern institutional delivery - represented by the category “medical knowledge under constrained circumstances”. The latest was considered more useful for handling complications, but the poor quality of the medical services provided hindered women’s access to them for delivery. In this struggle, TBAs, husbands and HEWs were supportive to women and facilitate their access to certain services (TBAs advocated for home delivery, whilst husbands and HEWs advocated for institutional deliveries). The model is illustrated in Figure 14. 68  Making pragmatic choices Figure 14. Model showing the core category, categories and sub categories emerged from the analysis on the experiences of delivery care in Tigray, Ethiopia. Category one reflects how women were aware of the risks of giving birth while recognizing the benefits and drawbacks of home and institutional deliveries, as one participant conveyed: ‘Unless, the labor is soon, it is not good to give birth at home. Because consequences like much bleeding cannot be treated on time. I get afraid of bleeding and narrow pelvis if the labor is at home. There is a good care in the hospital, when there is bleeding they give injections’. (FGD2) For services such as antenatal check-ups, participants visited health facilities because they wanted to know the fetal health status as well as their own. Usually the positive feedback given by health workers during ANC created a sense of safety that often prevented women for going back to the health facility for delivery. Another factor that hindered women’s access to health care facilities for delivery was their lack of control in the decision making process, both in terms of delivery being an event that can happen any time and because of women’s subordinated position. However, when complications such as bleeding, prolonged labor and retained placenta occurred, care from health care facilities was sought, usually with the support from relatives, and especially from their husbands. 69 Religious and cultural factors acted as barriers for institutional delivery. Religious rituals, the ascendancy of elderly women, and considering home as a familiar environment were perceived as pulling factors towards home delivery, as one participant described: ‘Women were not used to going to the HF before getting sick, it was not taken as a tradition, the more you get sick you go to HF. Elderly women have been giving birth at home in past times and they insist their daughters to give birth at home, It’s because there is a faith that St. Mary and God supports women during labor at home’. (FGD5) However, as another participant put it, faith did not need to conflict with institutional delivery: ‘Hikiminaa [health facility] is good with God’s help. Health professionals cure you from pain, exhaustion delay of placental expulsion and bleeding. Women die due to bleeding. So I think health professionals could help a lot with God’s help’. (FGD4) Women were well aware of the benefits of institutional delivery. They valued the information provided by health workers and were satisfied with the health care. Various procedures performed at the ANC visits such as checking fetal heartbeat, measurement of blood pressure and abdominal examinations were admired. The care provided by HEWs and the information they provide during ANC visits, were especially appreciated by women, as one participant expressed: ‘HEWs have a good hospitality. Whatever, how we behave, they treat and follow us during ANC at health post. They visit our houses, advise us to check our self during pregnancy and to give birth at HF. They inform us about issues about pregnancy, birth and after birth situations’. (FGD4) However, when it came to delivery, unlike pregnancy, the difficulty to forecast exact timing prevented them from planning their transportation in advance. In addition, despite women acknowledging the benefits of institutional delivery they also pointed out that the poor quality of care provided further prevented them from seeking institutional delivery. Long waiting times, unavailability/delay of health workers during night’s weekends, delay in referral to higher 70 health institution and offering incomplete of health services were among the critics mentioned by women. Women considered that health professionals lacked the skills, were not always available and had disrespectful attitudes towards them. Women were disappointed when relatives were not allowed to enter into the delivery room. The poor treatment, insults and embarrassment during delivery discouraged women to seek institutional delivery. Institutional delivery from health workers perspective Three themes emerged from the analysis of interviews with midwives and HEWs: “struggling between the tradition and newly acquired knowledge”, “community willingness to deal with geographical barriers” and “striving to do a good job with insufficient resources” (Figure 15). Both groups agreed on the important role of the HEP to increase communities’ awareness regarding the benefits of maternal health services utilization. Midwives acknowledged that the proximity of health posts and HEWs to the communities, the relentless effort of the volunteer community health workers and the newly created women development groups were major factors to increase community awareness on the benefits of institutional delivery. At the same time, they considered that the main reasons for poor institutional delivery included cultural traditions - such as the belief on catching evil eye - and the strong influence of elderly women on decision making. In the second theme, geographic inaccessibility emerged as a key barrier. HEWs expressed how the difficulties of mountainous topography and scattered localities challenged the communities endeavour. As a HEW mentioned: ‘Most of the localities of this kebele are with topography of terrains and mountainous. The shortest walking distance to reach the main road is 2-3 hours. Women prefer to give birth at home rather than troubling with the difficult roads’. (HEW 6) Despite the poor network connection of the rural localities, the joint commitment of the community members and HEWs in searching for transportation for referrals was considered as remarkable by the midwives. 71 In the final theme, midwives admired the dedication of HEWs in striving to provide health promotion and preventive services such as immunization, ANC and family planning with the scarce available resources. However, midwives were concerned regarding the poor capacity of HEWs to detect women who needed early referral. At the same time, HEWs recognized their own limitations on assisting women during labor, and considered the one month midwifery training they receive as inadequate. In general, HEWs mentioned the availability of competent health professionals, and the provision of services without fees for 24 hours a day and seven days a week in hospitals as promoters of institutional delivery. Both, HEWs and midwives acknowledged the hospital setting as the best place for deliveries and for dealing with obstetric emergencies. However, they also expressed their criticisms regarding the poor quality of care provided by certain professionals at the hospitals. Most of the HEWs commended the contribution of the HEP for improving preventive activities at the household level. However, they were concerned regarding communities’ increased demands for essential curative services. The large amount of time spent by HEWs for household visits prevented them from engaging in the provision of curative services within the health post. In addition, the poor infrastructure of the health posts in terms of electricity, water facilities and shortage of infection prevention materials (masks, goggles and shoes for the delivery room) were mentioned by HEWs as major barriers to provide adequate care. ‘An electric power is also necessary to offer services during the night time. When we assist delivery during night time we have to prepare kerosene or candle’. (HEW 2) 72 Figure 15. Health workers perspective on facilitators and barriers for institutional delivery. 73 74 Discussion This section is structured around the dimensions of access previously described in the conceptual framework (pages 47-48), namely availability, accessibility, acceptability, accommodation and affordability. We mainly focused on the first three, however the dimensions of access are interconnected and it is difficult to consider them as separate entities. For instance, accommodation was reflected in the data alongside with acceptability, and some aspects that we consider refer to acceptability might be perceived as referring to accommodation by other researchers. The findings of this thesis indicate an overall positive outcome of the HEP in improving different domains of access to maternal health care services but revealed also numerous challenges ahead. The increased trend of health services utilization after the intervention period (ANC from 28% to 47%, delivery care 5% to 23% and postnatal care 11% to 41%) indicates an improvement of the accessibility and availability dimensions of access. Findings from sub- study II showed a moderate coverage of ANC, but very low institutional delivery utilization by women in rural Tigray. ANC and delivery care use were associated to higher education of women (> 5 years) which is related to acceptability to the services. The provision of health information during ANC was associated with the use of institutional delivery that might be mediated through enhancing the acceptability of the services. In sub-studies III and IV, women and health workers referred to the influence of cultural factors, ascendancy of elderly women, unreliable transport, and poor quality of care as promoters of home delivery. Participants acknowledged the efforts that were being carried out in the region to improve the access dimensions of health care (Figure 9). Aspects that were mentioned included the presence of human resources, drugs and supplies at health facilities (availability), the community efforts to solve transportation problems in case of obstetric complications (accessibility), as well as the increased awareness on benefits of institutional delivery (acceptability). In this thesis, issues of affordability –users’ ability and willingness to pay for the services, and accommodation - the extent the resources of the health care system are organized in a manner to meet users expectation- did not emerge strongly. On one hand, the provision of free maternal health care services which was introduced by the HEP, 75 has contributed to make health services more affordable. On the other, the economic growth that the country has experienced in the last decade allowing to increase the health budget and the effort of the government to improve the organizational set up of the health service delivery system might have contributed to enhance accommodation [84]. Availability Availability refers to “the relationship between the existing services in volume and type, and the users’ demand” [109]. In the last decade, both national and regional governments, through the HEP, have invested in building infrastructure, expanding trained human resources and providing an adequate stock of supplies at health facilities to enhance availability. The availability of various medical equipment (i.e. sphygmomanometers, weighing scales and stethoscopes) and the presence of HEWs at health posts might have been perceived as a precondition for providing credible health care by users. The improved flow of ANC users was observed during the late intervention period of HEP which might be the effect of increased availability of health services. It seemed that the availability of midwives/nurses at health centres might have been more important for women than the availability of HEWs in health posts, since the trend of ANC increased significantly only at HC level. The preference of ANC offered in HC might be due to the fact that such facilities also offered other services valued by women such as laboratory testing investigations, diagnostic and curative services for common diseases and outpatient therapeutic feeding for malnourished children. Other studies reveal similar findings where maternal health care utilization is linked to availability of infrastructure, supplies, and human resources, and not merely to the presence of the health facilities [117- 118]. Such strong associations between availability of supplies and human resources and utilization of ANC has been reported for instance in one study conducted in Tanzania [119]. Conversely, the availability of HEWs in HPs did not improve access to delivery care services, since the health posts were mainly focused on promotive and preventive services and were not supposed to provide skilled delivery services after 2011-2012. In line with the country’s development and economic growth the Ethiopian news agency have broadcasted increased per capita income of some communities in the 76 rural population: as a result; women might skip the HPs to overcome the deficiencies and seek delivery services directly at health centres where nurses and midwives are available. This thesis also shows the perceived women’s benefits for maternal and newborn care derived from the availability of skilled professionals, equipment and supplies in the surrounding health care facilities (sub-study III). Another study conducted in western Ethiopia also reported a significant association between respondents’ awareness on benefits of skilled attendance and institutional delivery [120]. However, women were also aware that the availability of such resources in health posts and health centres was erratic. This uncertainty regarding which resources would be actually available when needed, discouraged women from seeking care from such facilities. Despite the continuous efforts to improve the services, poor availability of essential supplies and human resources remains as a major gap in remote localities in Tigray. The sub-optimal supply chain of some essential drugs such as iron sulphate (to treat anemia) and misoprostol (for prevention of post-partum hemorrhage) has contributed to HEWs frustration and created mistrust of HEWs’ capabilities. Furthermore, the absence of infection prevention materials (masks, eye goggles and shoes) was also another major barrier for HEWs to practice clean and safe delivery. Consistent with our findings, a study in Tanzania revealed the reduced motivation of health workers when the curative services are insufficient or sub- optimal due to the unavailability of drugs [119]. Accessibility Accessibility can be defined as: when the location of health facilities and supply of services is close to the users’ locality [109]. Sub-study I showed an increased trend on ANC, delivery care, and PNC services’ utilization. Such an increase might be partly attributed to the proximity of over 600 health posts (including the placement of HEWs and resources) and the upgrading number of clinics to health centres in the region under the HEP program [26-27, 46, 104]. Accessibility was closely linked with ANC utilization: living in the close proximity to a health post was a major predictor of ANC use. Consistent with our results, local and national studies have indicated 77 the benefit of living nearby to both ANC and institutional delivery services [26-27, 46, 104, 121-127]. Despite the moderately high rate of ANC coverage, institutional delivery in rural Tigray remained quite low (sub-study II) which is similar to findings from other national studies [19-22]. This result points out that even though accessibility is necessary, other relevant factors strongly influence the place of childbirth. Acceptability Acceptability refers to “how the users feel satisfied receiving health service from the providers and vice versa”. The users level of education, to receive advice and counseling, the degree of communication (by health workers) and the clarity of the messages conveyed, the respect received from health professionals and their knowledge and skills, are among the factors which are important for enhancing the acceptability of the services [109]. Besides, how health facilities are organized to accommodate users and their families with a friendly and comfortable approach plus cleanliness of health facilities, ensuring ease of use of drugs, and modern equipment and the overall quality of care are also determinant factors for acceptability [33, 128]. The role of HEWs, VCHWs and women development groups is central to enhance services’ acceptability. HEWs at health posts play a major role in establishing the linkage between the community and modern health care practice. The HEWs also serve to give voice for the demands of women in order to make health services more responsive to their needs. The WDA which mobilizes women for pregnancy check-ups and institutional delivery also promotes awareness of the community on the importance of the services. Moreover, the effort of skilled health workers for saving women and newborns lives; and the provision of 24 hours services in these health facilities have improved the acceptability of the services (sub-study IV). The Lancet series on midwifery highlights that effective practices related to maternal and child health along with respectful and responsive care optimizes acceptability of the services [129-130]. Sub-study III pointed out that women in Tigray aspire for safer child birth and were conscious of the possible complications during delivery. However, the decision of where to give birth was influenced by many aspects including cultural beliefs and traditions. The 78 ascendancy of elderly women, religious beliefs and socio-cultural aspects influenced women’s decision-making, presenting home delivery more acceptable than institutional delivery. This finding is in line with other studies [57, 96, 131]. On the other hand, HEWs and husbands, were advocates for institutional delivery (sub-study III) and they might try to convince the relatives and elderly women that it is the best option. When women get an opportunity to reach the health facility, the perceived low quality of services might be a strong factor hindering the acceptability of services. Among the issues found to diminish health care service acceptability for childbirth (sub-studies III and IV) were: 1) poor and slow referral system, 2) unavailability of curative services due to the focus on promotion, 3) communities’ perception that their services have been downgraded and 4) verbal abuse and arrogance of health workers during the service provision. These findings align with results from other studies where health care facilities that disregard of the traditional norms of the community is shown to reduce the acceptability of the services [132-137]. The child birth humanization approach is one that enhances the acceptability of institutional delivery. Such an approach is currently being introduced by THB and some changes on the way child birth is assisted in the facilities have been implemented; i.e. women can choose the position to give birth, porridge is served after birth and husbands are allowed to enter the delivery room. These interventions may contribute to improving the acceptability of the health facilities by the community. In sub-study II factors that enhance acceptability were shown to increase the use of maternal health care services. Educated women had higher use of ANC and institutional delivery compared to non- educated women. Extensive evidence shows how increasing women’s education might enhance their acceptance of maternal health care services [26-27, 121, 131, 136-137]. On the other hand women with higher education are more empowered to give voice to their needs and might be reluctant to accept the service when professionals treat them without empathy and respect. Apart from literacy, women receiving proper counseling service were found to be more likely to use institutional delivery. The positive effect of counseling on acceptability of ANC and institutional delivery has also been reported by other studies [126, 135-137]. A possible 79 explanation for this is that: 1) informed women are more likely to break misconceptions and opt for institutional delivery and 2) women might develop a sense of trust on the health workers capacity when they receive proper counseling service and physical examination during ANC. In addition, village councils have involved HEWs in their boards which may enable HEWs to better reach the community with their messages, which in turn can enhance service acceptability [123]. 80 Methodological considerations gling ween Various efforts were made to strengthen the research regarding both the quantitative and qualitative criteria considering the truth value, applicability, consistency, and neutrality [111, 138]. Truth value refers to the ability of the study to measure what it aimed to measure. In quantitative studies is defined as internal validity, and refers to minimizing bias [139-140]. Selection bias could have occurred in the cross sectional sub-study, for instance: due to flooding some remote communities might have been excluded from the study. In the longitudinal sub-study, only the districts with complete data on their record were included, so selection bias could have occurred due to the poor data handling of the districts. Regarding qualitative standards, truth value is defined as “credibility” and refers to how well the findings had captured the reality being explored [138]. In the two qualitative studies triangulation of researchers with different knowledge on the setting enhanced credibility. Triangulation of participants was done by involving both users and providers perspectives (sub-studies III and IV). In order to have a deep understanding and stay closer to the participants’ perceptions, the local (Tigrigna) version was used along with the translated English version until the end of the write up. The PI being originally from Tigray enabled prolonged engagement. The risk of naivety was minimized by discussion with the other researchers who were not familiar with the area. Applicability is important in public health research since it refers to what extent results can be generalized to other settings. Quantitative studies seek external validity by appropriate sample procedures [139- 140]. In the cross sectional survey, we ensured that communities and participants were randomly selected, thus enhancing their likelihood of representing the situation in Tigray in terms of socio-demographic characteristics. However, due to the diversified ethnic and cultural differences representativeness needs to be carefully handled. Qualitative studies seek transferability, since generalizations from those studies are theoretical and not statistical [138]. In both our qualitative studies, transferability was enhanced by purposely selecting participants based on their ability to contribute to the study question. We also made an effort to contextualize the results in order to help readers evaluate to what extent our results were applicable in other settings. 81 Research consistency, according to quantitative criteria, is defined as reliability and refers to the likelihood of getting similar results if the study is repeated [140]. For enhancing reliability the cross sectional study used questionnaires that had been validated in the country and they were piloted prior to application. Interviewers were well trained and supervised, and they were recruited from a similar background to that of the participants. In qualitative research, consistency has a different meaning and is defined as “dependability” or the ability to respond to issues emerging from the data [138]. In both the qualitative studies an emergent design was pursued. The interview guides allowed the incorporation of issues emerging from previous interviews, and we used preliminary emerging results to refine the design of the studies. The final issue for methodological considerations refers to neutrality [111]. Quantitative studies claim neutrality and objectivity by maintaining distance from the observed phenomenon (for instance, the data of the survey was collected by others), while qualitative studies deny the possibility of researcher’s neutrality and seek neutrality in the data (confirmability). In the quantitative studies, neutrality was applied by accepting the results of the analysis and not hiding the ones that did not support our working hypothesis. In the qualitative studies we strived to capture the ideas and experiences of informants as we tried to be open-minded and to put our pre- understanding into “brackets” [141]. For instance: despite the rich and positive experience of the researcher on HEP, the data was allowed to emerge by itself to reduce the bias in assumption. That meant not going in-depth into existing theories until findings were analyzed, and not excluding contradictory or negative findings. The main limitations of each sub-study are presented in Table 12. 82 Table 12. Limitations of sub-studies I-IV. Sub-study I  Use of retrospective data might be questionable for its validity.  Validity of data might have been hindered by poor handling of the archives and an absence of electronic medical recording system in the facilities which could have affected the interpretation of the results.  The study design could not detect the causal determinants of the impact on utilization.  Because of unavailability of data, the external factors that might have influenced the outcomes were not controlled. Such as: increase in per capita income, building roads and infrastructure, strengthening micro and small scale enterprises.  Due to the fact that our sample was taken from only a small amount of districts, the issue of generalizability needs to be carefully handled. Sub-study II  The data were collected by HEWs who were familiar to the community, which could have resulted in reporting bias but on the other hand also increased the response rate.  The questionnaire assessed the use of maternal health services in the five years preceding the survey, which might have introduced recall bias in the use of services.  The reasons given by women for attending/not attending might have been longstanding ideas that could introduce a recall bias, with the outcome often stated rather than the actual reason.  The methodological nature of the cross sectional study design limited the causality inference of the study variables. Sub-study III  My position as a former staff member of the THB may have affected participants’ opinions through social desirability bias.  The fact that the first author was a man, interviewing women could have affected the response of respondents (social desirability bias) and the results. However, the respondents reflected their opinion freely.  The process of translation of the interviews to English might have led to some of the original meanings (narrated by the participants) being lost. Sub-study IV  The setting and my position as a member of the university’s medical faculty interviewing health workers, could have affected the results (social desirability bias).  Purposive selection of the participants might have affected the results through social desirability bias, however, much work was done to connect the results with the actual circumstances of the setting and to detail the analytic process. 83 The role of the researcher During the development of this thesis, I have had both an insider and outsider role in investigating the HEP. I considered myself partly as an insider due to providing health care services in health facilities and working in the regional health bureau as a health extension program coordinator. This position allowed me to gain access to participants. The participants in the interviews and FGDs were women, and me (the interviewer) being a man could have influenced their responses. However, the objective of the study was clearly oriented and participants were approached to discuss their opinion freely. I also consider myself as an outsider because when I started this thesis, I was already working at the University, assessing the HEP from that position. This position allowed me to be more critical to the HEP and also to keep my pre-understandings due to my previous working experience with the HEP “in brackets” during the analysis of the data.nd 84 Conclusion Our findings reveal that there has been an overall increase in ANC, DC and PNC use after the introduction of the HEP, particularly in ANC and PNC at health post level and DC at health centres. However, despite the proportion of women who received ANC for their recent births (54%), those who gave birth at a health facility was only 4.1%. Factors associated with ANC utilization were marital status, high education, proximity of health facility to the village and the non- farming occupation of the husbands. Use of institutional delivery was mainly associated with parity, high education, having received ANC advice, to have a history of difficult labor and husbands with a non- farming occupation. The results point out that despite the effort of the THB to facilitate the women access to health facilities several challenges remain, particularly for child delivery. The women who participated in this study were actively making choices for a safer delivery. Women had to decide between (1) giving birth at home, in a known place, sanctioned by tradition, faith and the ascendancy of the elderly, but where little could be done to manage complications, or (2) delivering at health facilities, where medical knowledge could (theoretically) manage complications, but where services’ availability was on occasion erratic, the quality of care poor and staff’s attitudes disrespectful. In a similar line, health workers emphasized three major barriers for institutional delivery: 1) the perception of delivery as a natural event by the community which was linked to, cultural, ritual and traditional values of home delivery, 2) geographical barriers, such as long distances, mountainous localities and poor referral infrastructure and, 3) lack of adequate training, poor quality of care and shortage of supplies at health facilities. 85 86 Implications for further intervention In terms of enhancing availability  THB is acknowledged for effectively organizing HEWs and WDA together towards striving for better service acceptability. The direction of this movement needs to be reinforced and sustained in collaboration with all partners and with possible integrated supportive supervision.  Research from colleagues in Tigray has shown the feasibility of mobile health in improving maternal health services [142] so, enhancing access of the services through mobile health needs to be strengthened.  To break the misconceptions on benefits of home delivery and to create mutual understanding of the different actors involved, meetings and forums with elderly women, religious leaders, husbands, women development groups, TBAs and HEWs need to be conducted frequently.  The current efforts and commitments of HEWs need to be recognized and praised in an open forum. Performance based awards and a certificate that recognizes individual and team efforts should be prepared on annual or bi-annual bases.  The integrity and commitment of the regional government towards advancement of women’s education and their empowerment; and the willingness to improve access through organized community effort needs to be enhanced to ensure its sustainability.  The recent introduction of humanization approach for delivery deserves to be evaluated, in terms of its effect on enhancing acceptability of services. 87 88 Implications for further research  Action based research that enhances positive provider-user interaction and strengthens women centered approach for ensuring effective health service delivery is needed.  This study only focused on users and health providers’ perspective to explore the issues. It would be useful to conduct research exploring the perspective of community actors (such as WDGs) and local policy makers regarding issues of maternal health services.  PNC was not well assessed except for levels and trends of its utilization. It would be very interesting to fill the gap by conducting operational research on the coverage of PNC and its determinants from the perspective of various actors using quantitative and qualitative methodologies.  The access framework approach would be useful to explore other issues of maternal and reproductive health related problems, such as: maternal malnutrition, anemia in pregnancy, gender based violence, contraceptive use and HIV/AIDS. Finally, a strong collaboration between THB, Mekelle University, and Tigray Science and Technology Commission is needed to:  Enhance and facilitate the dissemination of PhD thesis and other research outputs to policy makers and to the community.  Coordinate and organize in designing community based projects that can be implemented based on the main findings of this PhD as well as other research. 89 90 Acknowledgements I am very grateful to the many people and organizations who contributed to this PhD thesis. First and foremost, I sincerely express my deep appreciation to all women who participated in the focus group discussion and the quantitative survey. Without their involvement the experiences of maternal health care wouldn’t have been explored in this thesis. My profound respect and gratitude goes to the key informants (health extension workers and nurse midwives) and to the other groups (nurses and HEWs) who participated as data collectors and supervisors for the survey. I would also like to acknowledge to my colleague Fisha Haile (FGD note taker and data collector) and all participants of this research including the district maternal health experts as well as archive officers. I have a deep appreciation for the district health offices, the Regional Health Bureau and Mekelle University. I am also very grateful to an American friend Joseph Brew Russel who committed his time and energy and assisted me the data collection task of the first paper. I am extremely indebted to my supervisors and their greatness will forever remain in my heart. It is due to their proper mentorship and guidance that I built my critical and analytical thinking and reached to the level of exploring the inlet and outlet of the research process. I would like to sincerely express my heartfelt gratitude to my main supervisor Isabel Goicolea, it was not only your technical assistance and proper advice but also your humor, kindness and respect that make me feel always welcomed and endured. You also gave me a freedom and energy to optimally utilize my efforts for effective implementation of the research project. Your careful listening and your deep interest to understand the social and cultural context of the study population have contributed in improving the work of the research. You were an advisor, a mentor as well as a sister. I am very grateful to your family Delfin and Peio for their hosting and welcoming reception all the time. My candid and deepest gratitude also goes to my co- supervisor Miguel San Sebastian who accepted me as a PhD student and gave me an opportunity to work under his guidance and supervision. It was much privilege for me to acquire and learn academic and research skills from you over the years. During the whole process, your valid criticisms, provoking comments, endless guidance and support, and your prompt responses and suggestions helped me to critically think, analyze and increase my capacity accordingly. Apart from the 91 mentoring and advising processes, the way you give credit to community values maximized my curiosity and hard work. I would like to thank your family (Anna Karin, Emil and Sisa) for their welcoming reception all the time. I sincerely admire my co-supervisor Kerstin Edin. I am so grateful for your talents, skills and your rich experiences which guide me to follow the standards and procedures and to keep an eye out to make things in different ways. You taught me how to reconcile contextual entities with the international arena. The family dinner at your house was also an extra commitment that builds a strong friendship between me and all my supervisors. I highly appreciate Anna Karin Hurtig, for her dedication and commitment in organizing and coordinating various capacity building workshops to the PhD students and for encouragment and reassurance during the PhD process. I am so grateful for Lars Lindholm for accepting to be my examiner. It would be impossible to forget the co-authors Yalem Tsegay and Hailemariam Lemma who had immense contribution on the second paper. Furthermore, I am very indebted to my good friend Hailemariam Lemma for being a gate opener to my PhD study by accompanying my co-supervisor (Miguel) to my work place. Dear Birgitta Astrom, I totally agree with the wonderful acknowledgments that have been given by former graduates. I am highly indebted for your everlasting care, support and motherhood since my Masters study. It is not only the technical, administrative and logistical arrangements but also your guidance, kindness and advice in all aspects that I appreciate. I give you my sincere thanks for your endless support throughout my PhD process. I would also like to acknowledge Karin Johnsson, Lena Mustonen, Veronica Lodwicka, Ulrika, Susanne Walther, Per Gustaffsson, Andreas and other administrative staffs for their continuous support and reassurance. I offer my appreciation to Anneli Ivarsson for her open interest and willingness to discuss my little daughter’s issue. I am very grateful to Kjerstin Dhalbolm who helped me to prepare my poster presentation for attending an international conference in Addis Ababa and for her reassurance all the time. I am highly indebted to Professor Yemane Berhan, Professor Misganaw Fantahun, Dr. Yirgu Gebrehiwet, and Dr. Nigusse Deyesa for their influencing messages and guidance to choose the PhD 92 journey over other competitive priorities. I also thank Dr. Kidist Lulu for reviewing and forwarding her comments to my last manuscript. I also acknowledge the various funding bodies that sponsored my trips back and forth from Ethiopia to Umeå to undertake my research studies and course work in Umeå. I am grateful to the support of the Swedish Centre Party donation for global health research. I would like to thank the Umeå Centre for Global Health Research, for granting me the support fund from FAS, the Swedish Council for Working Life and Social Research, as well as the Swedish Research School for Global Health. I sincerely wish to express my appreciation to the organization Etiopia-Utopia for covering the expenses of data collection for the qualitative and quantitative research. I sincerely express my gratitude to Professor Carmen, Nawi and Karin for their valuable and constructive comments, suggestions and critics during the mid-term seminar which helped me improve my work. I acknowledge the commitment of my opponents (Katrina and Faustine) on the pre-dissertation, their comments and suggestions were an immense contribution to my work. I am very thankful to Göran who was very supportive in uploading important statistical software and solving my PC problems. I would also like to pass my deepest gratitude on to all the staff at the Unit of Epidemiology and Global health, Peter Byass, John Kinsman, Anna Myleus, Klasse Sahlen, Joakim, Malin Eriksson and others. Special thanks to Barbara Schumann for her continuous reassurance. I am very grateful to Vigendra Ignole for his best friendship and brotherhood all the time while I was in Umeå. I was very lucky to socialize with people from different parts of the world and it has been a pleasure to learn from the diversified experiences by attending various research presentations. I am very grateful to Mariano, Claudia, Steven Maluka, Lorena, Cynthia Anticona, Paola, Juan Antonio Cordoba, Raman Preet, Joseph, Ailina Santosa, Bumi, Tej, Rakal, Dickson, Gladys, Thaddeus, Sewe, Kanyiva, Moses, Paul, Mellissa, Nathnael Sierili, Kateryna Kharina, Nyttin, Prasad, Trang, Henduruw, Barnabas, Kim, Osama, Setaria, Ryan, Fredinah, Ying Lei, Jing Hammersson, Linda Sundberg, Alirezakhatami, Chaya Utamie, Anna Lundgren, Daniel, Jonas Hansson and Johanna Sundquvist. My special thanks to Alison Hernandez for her open interest in helping and sharing her experiences. I acknowledge my good friends Masoud and Julia for their reassurance and encouraging discussion all the time. My sincere 93 thanks to Mikkel Quam for your respect, friendliness and companionship and your open willingness to support. I am very much indebted to Dr. Yohannes Kiros, Dr. Bisrat and Yodit Yohannes from Stockholm, for their continuous reassurance from the beginning to the end of this PhD journey and also for their welcoming reception all the time. My wonderful acknowledgment to Abrham Tsegay for his inspirational encouragement and brotherhood support. My colleagues Hagos, Ataklti and Tsige, I thank you so much for accompanying me halfway through the journey. As well as to Eduardo Subero for his encouragement and reassurance during his short stay in Mekelle. I am also very grateful to Ethiopian community in Umeå: Adane, Tekle, Fitsum, Haile, Measho, Mohamed, Abbe and Addis for your nice reception all the time. Thanks to my friends from Mekelle, Kahsu Bokretsion, Tadesse Alemseged, Mulu Atsbha, Tesfay Gebru, Yohannes Tewolde, Mebrahtom Belay, Awaela Equar, Yemane Ashebir, Solomon G/Mariam, Birhane Hadera, Habtom Gebrehiwet, Solomon Bezabih, Asmelash and Daniel for your encouragement and reassurance. I am so much indebted to Yemane, Tadelech, Mulu, Amaru, Axumawi and Hani Tewolde for the care, protection and their endless reassurance to my family while I have been away for my PhD study. I would like to pass my heartfelt gratitude to my Mom for her pronounced efforts throughout her life to make me a better man. Thank you “Destaye” for cultivating the values of solidarity and social integrity in me. I appreciate the blessing, prayers and the good wish of my Dad. My special thanks to my siblings, my sister “Tiebe” and my brother “Mihure” for your endless support to my family when I am away from home. Your understanding about my academic career gave me a courage, confidence and endurance. I am grateful to my beloved wife Luchia, for sharing my decision from the outset. Apart from accomplishing your own business, I appreciate your endless support in bearing the burden of family responsibility throughout the journey. Congratulations Luci, You are the back of this success. My lovely kids, Henos and Abseala who were very curious about my frequent travel to Umeå. My little daughter Arsema, who is very cute, her melodious words through the phone make me stable when I am away from home. 94 References 1. Unger JP, Van DP, Sen K, De PP: International health policy and stagnating maternal mortality: is there a causal link? Reproductive Health Matters 2009; 17(33): 91-104. 2. Brazier E, Andrzejewski C, Perkins ME, Themmen EM, Knight RJ, Bassane B: Improving poor women’s access to maternity care: Findings from a primary care intervention in Burkina Faso. Social Science and Medicine 2009; 69: 682-690. 3. Prost A, Colbourn T, Seward N, Azad K, Coomarasamy A, Copas A, et al.: Women's groups practicing participatory learning and action to improve maternal and newborn health in low-resource settings: a systematic review and meta-analysis. Lancet 2013; 381(9879): 1736-1746. 4. Marston C, Renedo A, McGowan CR, Portela A: Effects of Community Participation on Improving Uptake of Skilled Care for Maternal and Newborn Health: A Systematic Review. PLoS ONE 2013; 8(2): e55012. doi:10.1371/journal.pone.0055012. 5. Portela A, Santarelli C: Empowerment of women, men, families and communities: true partners for improving maternal and newborn health. British Medical Bulletin 2003; 67:59-72. 6. Grant R, Green D: The Health Care Home Model: Primary Health Care Meeting Public Health Goals. American Journal of Public Health. 2012; 102(6): 1096-1103. doi:10.2105/AJPH .2011.300397. 7. World Health Organization. Declaration of Alma Ata: Report of the international conference on primary health care. World Health Organization 1978; Alma Atta, USSR. 8. Rhode J, Cousens S, Chopra M, Tangcharoensathien V, Black R, Bhutta A Z, et al.: Alma-Ata: Rebirth and Revision 4: 30 years after Alma-Ata: WHO report-2008, 30 years after Alma-Ata: has primary health care worked in countries? Lancet 2008; 372: 950–961. 9. Walley J, Lawn JE, Tinker A, de Francisco A, Chopra M, Rudan I, et al.: Alma-Ata: Rebirth and Revision 8: Primary health care: making Alma-Ata a reality, Lancet series 2008; 372:1001-1007. 10. Haile Mariam D, Kitaw Y. Moving towards global health equity: opportunities and threats: an African perspective. Ethiopian Journal of Health Development 2012; 26: 238-250. 11. Federal Ministry of Health in Ethiopia: The health system of the country during the Transitional Government of Ethiopia. Federal Ministry of Health 1993; Addis Ababa, Ethiopia. 95 12. Federal Ministry of Health in Ethiopia: Health Sector Development Programme I 1997/98-2002/03. Federal Ministry of Health 2003; Addis Ababa, Ethiopia. 13. Federal Ministry of Health in Ethiopia: Health Sector Development Programme III. Annual Performance Report 2002 EFY 2009/2010. Federal Ministry of Health 2010; Addis Ababa, Ethiopia. 14. Federal Ministry of Health in Ethiopia: Health extension programme implementation guidelines. Federal Ministry of Health 2004; Addis Ababa, Ethiopia. 15. Federal Ministry of Health in Ethiopia: Health extension programme in Ethiopia. Federal Ministry of Health 2007; Addis Ababa, Ethiopia. 16. Tigray Health Bureau: Tigray Health Profile, 2010. Tigray Health Bureau 2010; Mekele, Ethiopia. 17. United Nations: Report of International Health Conferences. UN Doc. E/772. United Nations 1948; New York, USA. 18. Mahler H.: Address to the Nairobi Safe Motherhood Conference, World Health Organization 1987; Geneva, Switzerland. 19. World Health Organization: Trends in maternal mortality: 1990 to 2013. Estimates by WHO, UNICEF, UNFPA, the World Bank and the United Nations Population Division I. World Health Organization 2014; Geneva, Switzerland. 20. Nicholas JK, Amelia BV, Megan SC, Katya AS, Caitlyn S, Kyle RH et al.: A systematic analysis for the Global Burden of Disease Study 2013. Lancet. 2014 Sep 13; 384(9947): 980–1004. 21. World Health Organization: Maternal Mortality in 2005 Estimates developed by WHO, UNICEF, UNFPA and The World Bank. World Health Organization 2005; Geneva, Switzerland. 22. Thaddeus S, Maine D: Too far to walk: maternal mortality in context. Social Science and Medicine 1994; 38(8): 1091-1110. 23. Friberg IK, Kinney MV, Lawn JE, Kerber KJ, Odubanjo MO, Bergh AM et al.: Sub-Saharan Africa’s Mothers, Newborns, and Children: How Many Lives Could Be Saved with Targeted Health Interventions? PLoS Medicine 2010; 7(6): e1000295. doi:10. 1371/journal.pmed.1000295. 24. WHO, The UN Children’s Fund, The UN Population Fund. Maternal mortality in 1995: estimates developed by WHO, UNICEF, UNFPA. World Health Organization 2001; Geneva, Switzerland. 25. Travis P, Bennett S, Haines A, Pang T, Bhutta Z, Hyder AA, et al.: Overcoming health systems constraints to achieve the 96  Millennium Development Goals. Lancet 2004; 364(9437): 900- 906. 26. Karim AM, Betemariam W, Yalew S, Alemu H, Carnell M, Mekonnen Y: Programmatic correlates of maternal healthcare seeking behaviors in Ethiopia. Ethiopian Journal of Health Development 2010; 24(1): 92–99. 27. Amano A, Gebeyehu A, Birhanu Z: Institutional delivery service utilization in MunisaWoreda. SouthEast Ethiopia: a community based cross-sectional study. BMC Pregnancy Childbirth 2012; 12:105. 28. Warren C: Care seeking for maternal health: challenges remain for poor women. Ethiopian Journal of Health Development 2010; 24(Special Issue 1): 100–104. 29. Mekonen Y, Mekonen A: Factors influencing the use of maternal health care services in Ethiopia, Addis Ababa. Health Population Nutrition 2003; 21(4): 374–382. 30. Nigussie M, Hailemariam D, Mitkie G: Assessment of Safe Delivery care utilization among women child bearing age in North Gondar Zone, North West Ethiopia. Ethiopian Journal Health Development 2004; 18(3):145–152. 31. Fikre AA, Demissie M: Prevalence of institutional delivery and associated factors in DodotaWoreda (district), Oromia regional state. Ethiopia Reproductive Health 2012; 9: 33. 32. D’ Ambruoso L, Byass P, Qomariyah SN: ‘Maybe it was her fate and maybe she ran out of blood’: final caregivers’ perspectives on access to care in obstetric emergencies in rural Indonesia: Journal Biosocial Science 2010; 42(2): 213–241. 33. Freedman LP, Graham WJ, Brazier E, Smith JM, Ensor T, Fauveau V, et al.: Practical lessons from global safe motherhood initiatives: time for a new focus on implementation. Lancet 2007; 370(9595): 1383–1391. 34. D'Ambruoso L, Byass P, Qomariyah SN, Ouédraogo M: A lost cause? Extending verbal autopsy to investigate biomedical and socio-cultural causes of maternal death in Burkina Faso and Indonesia. Social Science and Medicine 2010; 71(10): 1728- 1738. 35. De Brouwere V, Van Lerberghe W: Safe motherhood strategies, a review of the evidence. Studies Health Service Organization Policy 2001; 17. 36. Gabrysch, S, Lema C, Bedriñana, E, Bautista AM, Malca R, Campbell MR O et al.: Cultural adaptation of birthing services in rural Ayacucho, Peru. Bulletin of the World Health Organization 2009; 87(9): 724–729. 97 37. Otis KE, Brett JA: Barriers to hospital births, why do many Bolivian women give birth at home. Panamerican Journal of Public Health 2008; 24(1): 46-53. 38. Travis P, Bennett S, Haines A, Pang, T, Bhutta Z, Hyder AA et al.: Overcoming health-systems constraints to achieve the Millennium Development Goals. Lancet 2004; 364(9437): 900- 906. 39. Starrs AM: Safe motherhood initiative: 20 years and counting, Family Care International. Lancet 2006; 368(9542): 1130-1132. 40. World Health Organisation. Report of the International Safe Motherhood Conference. World Health Organization 1987; Nairobi, Kenya. 41. Rosenfield A, Maine D: Maternal mortality—a neglected tragedy. Where is the M in MCH? Lancet 1985; 326(8446): 83–85. 42. Freedman LP, Graham WJ, Brazier E, Smith JM, Ensor T, Fauveau V, et al.: Practical lessons from global safe motherhood initiatives: time for a new focus on implementation. Lancet 2007; 370(9595): 1383–1391. 43. Lawn JE, Tinker A, Munjanja SP, Cousens S: Where is maternal and child health now? Lancet 2006; 368(9546): 1474–1477. 44. Darmstadt GL, Lee AC, Cousens S, Sibley L, Bhutta ZA, Donnay F, et al.: 60 million non-facility births: who can deliver in community settings to reduce intra-partum related deaths? International Journal of Gynaecology and Obstetrics 2009; 107: S89–S112. 45. Greenwood AM, Bradley AK, Byass P, Greenwood BM, Snow RW, Bennett S, et al.: Evaluation of a primary health care programme in the Gambia. I. The impact of trained traditional birth attendants on the outcome of pregnancy. Journal of Tropical Medicine and Hygiene 1990; 93(1): 58–66. 46. Koblinsky M, Tain F, Gaym A: Responding to the maternal health care challenge: The Ethiopian Health Extension Program. Ethiopian Journal of Health Development 2010; 24(1): 105– 109. 47. Nyamtema AS, Urassa DP, van Roosmalen J: Maternal health interventions in resource limited countries: a systematic review of packages, impacts and factors for change. BMC Pregnancy Childbirth 2011, 11: 30. 48. Starrs A: Preventing the tragedy of maternal deaths: a report on the international safe motherhood conference. Family Care International 1987; New York, USA. 98 49. Joseph L: WHO, UNICEF: Monitoring emergency obstetric care: a handbook. World Health Organization 2009; Geneva, Switzerland. 50. United Nations. Report of the International Conference on Population and Development 1994. United Nations 1995; Cairo, Egypt. 51. United Nations: Report of the International Women’s Conference, 1995. United Nations 1996; Beijing, China. 52. World Health Organisation: Report of the 2nd Inter-Agency Meeting of Reproductive Health Indicators for Global Monitoring. World Health Organization 1999; Available at http://www.who.int.reproductive _health/publications/. [Accessed April 2014]. 53. World Health Organisation. Reducing Maternal Mortality. A Joint WHO/UNFPA/World Bank Statement. World Health Organization 1999; Geneva, Switzerland. 54. McIntosh C, Alison F, Jason L: The Cairo conference on population and development: A new paradigm? Population and Development Review 1995; 223-260. 55. World Health Organization. Health and the millennium development goals. World Health Organization 2005; Geneva, Switzerland. 56. World Health Organization. The partnership for maternal, newborn and child health. World Health Organization 2011; Geneva, Switzerland. Available at: http://www.who.int /pmnch/en/, [accessed on 2011 May 20]. 57. Kerber KJ, de Graft-Johnson JE, Bhutta ZA, Okong P, Starrs A, Lawn JE: Continuum of care for maternal, newborn, and child health: from slogan to service delivery. Lancet 2007, 370(9595): 1358–1369. 58. Rasanathan K, Montesinos VE, Matheson D, Etienne C, Evans T: Primary health care and the social determinants of health: essential and complementary approaches for reducing inequities in health. Journal Epidemiology Community Health 2011; 65: 656-660. 59. World Health Organization. The World Health Report 2008: primary health care now more than ever. World Health Organization 2008; Geneva, Switzerland. 60. Parkhurst OJ, Loveday KP, Blaauw D, Balabanova D, Danishevski K, Rahman AS, et al.: Health systems factors influencing maternal health services a four-country comparison. Health Policy 2005, 73: 127–138. 99 61. Yamin, AE. From ideals to tools: applying human rights to maternal health. PLoS Medicine 2013, 10.11: e1001546. 62. United Nations Population Fund Agency. Accelarated campaign to reduce maternal mortality in Africa: (CARMMA), Maternal Health Thematic Fund. United Nations Population Fund Agency 2012; New York, USA. 63. David P: Fossils From Ethiopia May Be Earliest Human Ancestor National Geographic news, reporting your world daily, ASU, Institute of Human Origins. Available at: http://news.nationalgeographic.com/news/2001/07/0712_ethi opianbones.html [accessed April 2014]. 64. Munro-Hay S: Ethiopia, The Unknown Land: a Cultural and Historical Guide. I.B. Tauris and Co. Ltd., 2002, London and New York. 65. Grierson R, Munro-Hay S: The Ark of the Covenant, Phoenix 2000, London, UK. 66. Adejumobi AS: The History of Ethiopia. Greenwood Press Westport, 2007, London, UK. 67. Population Census Commission. Summary and statistical report of the 2007 population and housing census: population size by age and sex. Federal Democratic Republic of Ethiopia 2008, Addis Ababa, Ethiopia. 68. Federal Government of Ethiopia. Situational Analysis on the Libreville Declaration on Health and Environment Inter linkage Country Report Ethiopia. Federal Government of Ethiopia 2010; Addis Ababa, Ethiopia. 69. Central Statistical Agency: Report of Ethiopian Demographic and Health survey 2011. Central Statistical Agency and ICF International 2012; 270-272. Addis Ababa, Ethiopia. 70. Federal Democratic Republic of Ethiopia: The constitution of the transitional government of Ethiopia 1993. Federal Democratic Republic of Ethiopia 1994; Addis Ababa, Ethiopia. 71. Federal Ministry of Health. Health Sector Development Programme IV 2010/11-2014/15. Federal Ministry of Health 2011; Addis Ababa, Ethiopia. 72. Leta A: World trade organization and its implication for low income countries Addis Ababa University 2011; Addis Ababa, Ethiopia. Available at http//www.etd.aau.edu.et. 73. Teshome A: Agriculture, Growth and Poverty Reduction in Ethiopia: Policy Processes Around the New PRSP (PASDEP). Research Paper 004 2006; Addis Ababa, Ethiopia. Available at www.future.agricultures.org. 100 74. Federal Ministry of Health. Growth and Transformation plan 2010/11-2014/15. Ministry of Finance and Economic Development 2010; Addis Ababa, Ethiopia. 75. Gieger M. Ethiopian economic update, overcoming inflation: raising competitiveness. World Bank 2012; Washington DC, USA. Available at: http://documents.worldbank.org/curated /en/2012/11/17073947/ [accessed April 2014]. 76. Geda A, Shimeles A: The Pattern of growth, poverty and inequality in Ethiopia: which way for a pro-poor growth? Ethiopian Economic Association 2013; Addis Ababa, Ethiopia. Available at http//www.eeaecon.org/node/4880. 77. United Nations Development Program in Ethiopia. Annual Report, 2012; Transformation, Excellence and Action. United Nations Development Program 2012; Addis Ababa, Ethiopia. 78. Malik K: Human development report 1990-2013. The rise of the South: Human progress in a diverse world. United Nations Development Program 2013; New York, USA. 79. Erulkar A: Early marriage, marital relations and intimate partner violence in Ethiopia. International Perspectives on Sexual and Reproductive Health 2013; 39(1): 6–13, doi: 10.1363/3900613. 80. Getahun H, Berhane Y: Abortion among rural women in north Ethiopia. International Journal Gynaecology and Obstetrics 2000; 71(3): 265-266. 81. Demowez T, Molakign A, Brandon R, Duncan A, Hayden M: Ethiopian Cultural Profile: Information about Ethiopian history, culture and community with emphasis on health related issues: University of Washington and Ethiopian community Association, 2008; Available at [https://ethnomed.org/ culture/ethiopian/copy_of_ethiopian-cultural-profile/@view- documentation, Accessed October 10/2013]. 82. World Economic Forum: An Insight Report: The Global Gender Gap Report 2014. World Economic Forum 2014; Geneva, Switzerland. 83. Weldearegawi B, Ashebir Y, Gebeye E, Gebregziabiher T, Yohannes M, Mussa S, et al.: Emerging chronic non- communicable diseases in rural communities of Northern Ethiopia: evidence using population-based verbal autopsy method in Kilite Awlaelo surveillance site. Health Policy and Planning 2013; 28(8):891-8 101 84. Federal Ministry of Health: Health and Health indicators, 2011- 2012. Federal Ministry of Health Policy and Planning Directorate Ethiopia 2012; Addis Ababa, Ethiopia. 85. World Health Organization. Millennium Development Goals report, Ethiopia. World Health Organization 2012; Addis Ababa, Ethiopia. 86. Sutcliffe JP: Economic assessment of land degradation in the Ethiopian highlands: A case study. Transitional Government of Ethiopia, Ministry of Planning and Economic Development 1993; Addis Ababa, Ethiopia. 87. Paulos C: Clientelism and Ethiopia's post-1991 decentralization, The Journal of Modern African Studies 2007; 45: 355-384. doi: 10.1017/Soo22278x07002662. 88. Derick Brinkerhoff D, Bossert T: Health Governance: Concepts, Experience, and Programming Options. Health Finance and Governance 2008. 89. World Health Organization. Country cooperation strategy 2002-2005. World Health Organization 2006; Addis Ababa, Ethiopia. 90. Bilal KN, Herbst HC, Zhoo F, Soucat A, Lemiere C: Yes, Africa can: Success stories from a dynamic continent, HEWs in Ethiopia: Improved Access and coverage for the rural poor. International Bank for Reconstruction and Development 2011; Washington DC; USA. 91. Banteyerga, H: Ethiopia's health extension program: improving health through community involvement. MEDICC Review 2011; 13(3): 46-49. 92. Roman T, Accorsi S, Akalu T, Vella V, Mamo D: Assessing the performance of the health sector: achievements and challenges in EFY 2001 [G.C.2008/9]. Policy and Practice: information for action. Quarterly Health Bulletin, Federal Ministry of Health 2010; Addis Ababa, Ethiopia. 93. Ministry of Health. HSDP IV Annual Performance Report, version I, FDRE 2012/2013. Ministry of Health 2013; Addis Ababa, Ethiopia. 94. Tigray Health Bureau. Tigray Health Profile, 2012. Tigray Health Bureau 2012; Mekelle, Ethiopia. 95. Central Statistical Agency. Ethiopian demographic and health survey 2005. Central Statistical Agency and ICF International, 2006; Addis Ababa, Ethiopia. 96. Abera M, Belachew T: Predictors of safe delivery service utilization in Arsi zone South-East Ethiopia. Ethiopian Journal of Health Science 2012; 21(3): 101–113. 102 97. Teferra AS, Alemu FM, Woldeyohannes SM: Institutional delivery service utilization and associated factors among mothers who gave birth in the last 12 months in Sekela District, North West of Ethiopia: a community-based cross sectional study. BMC Pregnancy Childbirth 2012; 12: 74. 98. Medhanyie A, Spigt M, Kifle Y, Schaay N, Sanders D, Blanco R, et al: The role of health extension workers in improving utilization of maternal health services in rural areas in Ethiopia: a cross sectional study. BMC Health Services Research 2012; 12: 352. 99. Medhanyie A, Spigt M, Dinant G, Blanco R: Knowledge and performance of the Ethiopian health extension workers on antenatal and delivery care: a cross-sectional study. Human Resources for Health 2012; 10(1): 44. 100. Abebe F, Berhane Y, Girma B: Factors associated with home delivery in Bahirdar, Ethiopia: a case control study. BMC Research Notes 2012; 5: 653. 101. Gebrehiwot T, Goicolea I, Edin K, San Sebastian M: Making pragmatic choices: a grounded theory analysis of women´s experiences of delivery care in Northern Ethiopia. BMC Pregnancy Childbirth 2012; 12: 113. doi: 10.1186/1471-2393- 12-113. 102. Gebrehiwot T, San Sebastian M, Edin K, Goicolea I: Health workers’ perceptions of facilitators and barriers for institutional delivery in Tigray, Northern Ethiopia. BMC Pregnancy Childbirth 2014; 14:137. doi: 10.1186/1471-2393-14-137. 103. Federal Democratic Republic of Ethiopia. Report on the proceedings and results of the 12th Annual Review Meeting of Health Sector Development Program. Federal Ministry of Health 2010; Addis Ababa, Ethiopia. 104. Tsegay Y, Gebrehiwot T, Goicolea I, Edin K, Lemma H, Sebastian MS: Determinants of antenatal and delivery care utilization in Tigray region, Ethiopia: a cross-sectional study. International Journal Equity in Health 2013; 12: 30. doi:10.1186/1475-9276-12-30. 105. Admassie A, Abebaw D, Woldemichael AD: Impact evaluation of the Ethiopian Health Services Extension Programme. Journal of Development Effectiveness 2009; 1(4): 430-449. doi: 10.1080/19439340903375724. 106. Federal Minsitry of Health. Training of Health Extension Workers (HEW) On Family Folder and HMIS Procedures, Federal Ministry of Health 2011; Addis Ababa, Ethiopia. 103 107. Negusse H, McAuliffe E, MacLachlan M: Initial community perspectives on the health service extension programme in welkait, Ethiopia. Human Resources for Health 2007; 5: 21. 108. Teklehaimanot A, Kitaw Y, Girma S, Seyoum S, Desta S, Ye- Ebiyo Y: Study of the working conditions of health extension workers in Ethiopia. Ethiopian Journal of Health Development 2007; 21(3): 246-259. 109. Obrist B, Iteba N, Lengeler C, Makemba A, Mshana C, Nathan R et al.: Access to health care in contexts of livelihood insecurity: A framework for analysis and action. PLoS Medicine 2007; 4(10): e308. doi:10.1371/journal.pmed.0040308. 110. Reda H: Improving efficiency, access to and quality of the rural health Extension Programme in Tigray, Ethiopia: the case of malaria diagnosis and treatment. 2012; Umeå University Medical Dissertations New Series No. 1947. 111. Dhalgren L, Emmelin M, Winkvist A: Qualitative Methodology for International Public Health. Umeå University 2004; Umeå, Sweden. 112. Wagner AK, Soumerai SB, Zhang F, Ross-Degnan D: Segmented regression analysis of interrupted time series studies in medication use research. Journal of Clinical Pharmacology Therapy 2002; 27: 299-309. 113. Stekelenburg J, Kyanamina S, Mukelabai M, Wolffers I, van Roosmalen J: Waiting too long: low use of maternal health services in Kalabo, Zambia. Tropical Medicine International Health 2004; 9: 390-398. 114. Wilkinson D, Gouws E, Sach M, Karim SS: Effect of removing user fees on attendance for curative and preventive primary health care services in rural South Africa. Bulletin of the World Health Organization 2001; 79: 665-671. 115. Robe S, Everitt BS. Statistical analysis using Stata, Fourth edition. Taylor and Francis 2007; New York, USA. 116. Braun V, Clarke V: Using thematic analysis in psychology, Qualitative Research in Psychology. 2006; 3: 77-101. 117. Mamdani M, Bangser M: Poor people’s experiences of health services in Tanzania. A literature review. Reproductive Health Matters 2004; 12: 138–153. 118. Huntington D, Banzonb E, Recidoroc DZ: A systems approach to improving maternal health in the Philippines. Bulletin of the World Health Organization 2012; 90: 104–110. 119. Mkoka DA, Goicolea I, Kiwara A, Mwangu M, Karin AH: Availability of drugs and medical supplies for emergency obstetric care: experience of health facility managers in a rural 104  District of Tanzania. BMC Pregnancy and Childbirth 2014; 14: 108. 120. Feyissa TR, Genemo GA: Determinants of institutional delivery among childbearing age women in Western Ethiopia, 2013: Unmatched case control study. PLoS ONE 2014; 9(5): e97194. doi:10.1371/journal.pone.0097194. 121. Koblinsky M, Matthews Z, Hussein J, Mavalankar D, Mridha MK, Anwar I, et al.: Going to scale with professional skilled care. Lancet 2006; 368(9544): 1377–1386. 122. Bradley S, McAuliffe E: Mid-level providers in emergency obstetric and new-born health care: factors affecting their performance and retention within the Malawian health system. Human Resources for Health 2009; 7:14. 123. Teijlingen ER, Porter M, Simkhada P: Factors affecting the utilization of antenatal care in developing countries: systematic review of the literature. Journal of Advanced Nursing 2008; 61(3): 244-260. 124. Yesuf AE, Margalit CR: Disparities in the use of antenatal care service in Ethiopia over a period of fifteen years. BMC Pregnancy Childbirth 2013; 13:131. 125. Sibley LM, Tesfaye S, Desta FB, Frew HA, Kebede A, Mohammed H, et al.: Improving Maternal and New-born Health Care Delivery in Rural Amhara and Oromiya Regions of Ethiopia through the Maternal and Newborn Health in Ethiopia Partnership. Journal of Midwifery and Women’s Health 2014; 59; S6-S20. 126. Gabrysch S, Campbell OM: Still too far to walk: Literature review of the determinants of delivery service use. BMC Pregnancy Child Birth 2009; 9: 34. 127. Bedford J, Gandhi M, Admassu M, Girma A: ‘A Normal Delivery Takes Place at Home’: A Qualitative Study of the Location of Childbirth in Rural Ethiopia. Maternal Child Health Journal 2013; 17(2): 230-239. 128. Lubbock LA, Stephenson RB: Utilization of maternal health care services in the department of Matagalpa, Nicaragua. Pan American Journal of Public Health 2008; 24(2): 75-84. 129. Renfrew J M, McFadden A, Bastos HM, Campbell J, Channon AA, Cheung FN et al.: Midwifery and quality care: findings from a new evidence informed Framework for maternal and newborn care Lancet Series 2014; Available on http://dx.doi.org.10.1076/S0140-6736(14)60789-3. 105 130. Bender HP, Bernis L, Campbell J, Downe S, Fauveau V, Fogstad H et al.: Midwifery 4: Improvement of maternal and newborn health through midwifery, Lancet series 2014; Available on http://dx.doi.org.10.1076/S0140-6736(14)60930-2. 131. Mrisho M, Obrist B, Schellenberg JA, Haws RA, Mushi AK, Mshinda H, et al.: The use of antenatal and postnatal care: perspectives and experiences of women and health care providers in rural southern Tanzania. BMC Pregnancy Childbirth 2009; 9: 10. 132. Mrisho M, Schellenberg JA, Mushi AK, Obrist B, Mshinda H, Tanner M, et al.: Factors affecting home delivery in rural Tanzania. Tropical Medicine and International Health 2007; 12: 862–872. 133. Magoma M, Requejo J, Campbell OM, Cousens S, Filippi V: High ANC coverage and low skilled attendance in a rural Tanzanian district: a case for implementing a birth plan intervention. BMC Pregnancy Childbirth 2010; 10: 13. 134. Shiferaw S, Spigt M, Godefrooij M, Melkamu Y, Tekie M: Why do women prefer home births in Ethiopia? BMC Pregnancy Childbirth 2013; 13: 5. 135. Ochako R, Fotso JC, Ikamari L, Khasakhala A: Utilization of maternal health services among young women in Kenya: Insights from the KenyaDemographic and Health Survey, 2003. BMC Pregnancy Childbirth 2011: 11(1): 1. 136. Mwakipa H: Use pattern of maternal health services and determinants of skilled care during delivery in Southern Tanzania: implications for achievement of MDG-5 targets. BMC Pregnancy Childbirth 2007; 6: 7–29. 137. Nikiema B, Beninguisse G, Haggerty JL: Providing information on pregnancy complications during antenatal visits: unmet educational needs in sub-Saharan Africa. Health Policy Planning 2009; 24(5): 367–376. 138. Lincoln Y, Guba E. Naturalistic Inquiry. SAGE 1985; New York, USA. 139. Hennekens CH, Buring JE. Epidemiology in Medicine. Lippincott-Raven 1987; Philadelphia, USA. 140. Bonita R, Beaglehole R, Kjellström T: Basic Epidemiology, 2nd edition. World Health Organization 2006; Geneva, Switzerland. 141. Husserl E: Experience and judgment. Northwestern University Press 1975; Chicago, USA. 142. Little A, Medhanyie A, Yebyo H, Spigt M, Dinant G-J, Blanco R: Meeting Community Health Worker Needs for Maternal Health 106 Care Service Delivery Using Appropriate Mobile Technologies in Ethiopia . PLoS ONE. 2013; 8(10): e77563. doi:10.1371/journal.pone.0077563. 107