Pakistan. Maternal Mortality Survey 2019

Publication date: 2020

Pakistan Maternal Mortality Survey 2019 P akistan 2019 M aternal M ortality S urvey PAKISTAN MATERNAL MORTALITY SURVEY 2019 National Institute of Population Studies Islamabad, Pakistan The DHS Program ICF Rockville, Maryland, USA December 2020 The 2019 Pakistan Maternal Mortality Survey (2019 PMMS) was implemented by the National Institute of Population Studies (NIPS) under the aegis of the Ministry of National Health Services, Regulations and Coordination, Islamabad, Pakistan. ICF provided technical assistance through The DHS Program, a project funded by the United States Agency for International Development (USAID) that provides support and technical assistance in the implementation of population and health surveys in countries worldwide. Support for the survey was also provided by the Department for International Development (DFID), the United Nations Population Fund (UNFPA), and Bill and Melinda Gates Foundation (BMGF). Additional information about the 2019 PMMS may be obtained from the National Institute of Population Studies, Ministry of National Health Services, Regulations and Coordination, National Institute of Health (NIH), Park Road, Chak Shahzad, Islamabad, Pakistan; telephone: +92-51-9255937; fax: +92-51-9255932; internet: www.nips.org.pk. Information about The DHS Program may be obtained from ICF, 530 Gaither Road, Suite 500, Rockville, MD 20850, USA; telephone: +1-301-407-6500; fax: +1-301-407-6501; email: info@DHSprogram.com; internet: www.DHSprogram.com. ISBN: 978-969-9732-07-2 Cover design: Mohsin Khan. Cover photos (from left to right): ‘Quaid-e-Azam tomb’ ©2020 by Muhammad Masood Khan; ‘Quaid e Azam Residency’ ©2020 by Zainullah Kakar; ‘Minar-e-Pakistan’ © 2020 by Muhammad Masood Khan; Bab-e-Khyber © 2020 by Muhammad Saqib; Faisal Masjid © 2020 by Zainullah Kakar. Suggested citation: National Institute of Population Studies (NIPS) [Pakistan] and ICF. 2020. Pakistan Maternal Mortality Survey 2019. Islamabad, Pakistan, and Rockville, Maryland, USA: NIPS and ICF. The contents of this report are the sole responsibility of NIPS and ICF and do not necessarily reflect the views of USAID, the United States Government, or other donor agencies. http://www.dhsprogram.com/ Contents • iii CONTENTS TABLES AND FIGURES . v FOREWORD . xi ACKNOWLEDGEMENTS . xiii 2019 PAKISTAN MATERNAL MORTALITY SURVEY TECHNICAL ADVISORY COMMITTEE . xv CONTRIBUTORS TO THE REPORT . xvii ACRONYMS AND ABBREVIATIONS . xix READING AND UNDERSTANDING TABLES FROM THE 2019 PMMS . xxi 1 INTRODUCTION AND SURVEY METHODOLOGY . 1 1.1 Survey Objectives . 1 1.2 Sample Design . 2 1.3 Questionnaires . 3 1.3.1 Development of Section 5 (Maternal Morbidity) of the Woman’s Questionnaire . 4 1.4 Pretest . 4 1.5 Training of Field Staff . 5 1.6 Fieldwork . 5 1.7 Quality Control . 6 1.7.1 Verification of Verbal Autopsy Questionnaires . 6 1.8 Data Processing . 6 1.9 Response Rates . 7 2 HOUSEHOLD POPULATION, HOUSEHOLD AND RESPONDENTS’ CHARACTERISTICS . 9 2.1 Household Population and Composition . 9 2.2 Drinking Water Sources and Treatment . 11 2.3 Sanitation . 12 2.4 Household Characteristics . 12 2.5 Household Wealth . 13 2.6 Services in Rural Areas . 14 2.7 Basic Characteristics of Survey Respondents . 14 2.8 Educational Attainment . 15 3 ADULT AND MATERNAL MORTALITY . 27 3.1 Mortality Rates . 28 3.2 Reproductive Age Mortality . 30 3.3 Trends in Adult Mortality . 32 3.4 Life Expectancy . 34 3.5 Estimates of Pregnancy-related and Maternal Mortality . 35 4 CAUSES OF DEATH . 51 4.1 Verbal Autopsy Questionnaire . 51 4.2 Verbal Autopsy Fieldwork . 51 4.3 Cause of Death Certification and ICD-10 Coding . 52 4.4 Characteristics of Deceased Women . 54 4.5 Respondents to the Verbal Autopsy Questionnaires . 55 4.6 Cause-specific Mortality . 55 iv • Contents 4.7 Maternal Causes of Death . 56 4.8 Deceased Women and Health Care . 57 5 MATERNAL HEALTH CARE . 63 5.1 Antenatal Care Coverage and Content . 63 5.1.1 Skilled Providers . 63 5.1.2 Timing and Number of ANC Visits . 64 5.1.3 Components of ANC . 65 5.1.4 Protection against Neonatal Tetanus . 65 5.2 Delivery Services . 66 5.2.1 Institutional Deliveries . 66 5.2.2 Skilled Assistance during Delivery . 67 5.2.3 Delivery by Caesarean . 68 5.3 Postnatal Care . 68 5.3.1 Postnatal Health Check for Mothers . 68 5.3.2 Complete Maternity Care . 69 6 MATERNAL MORBIDITY . 87 6.1 Maternal Morbidities during Pregnancy . 88 6.2 Maternal Morbidities during Delivery . 88 6.3 Maternal Morbidities during the Postpartum Period . 89 6.4 Maternal Morbidities about which Women were Informed by a Health Care Provider . 90 6.5 Morbidities Reported before the Last Pregnancy . 90 6.6 Maternal Complications or Morbidities and Treatment . 91 7 HEALTH CARE SEEKING BEHAVIOUR . 97 7.1 Treatment for Maternal Health Complications . 98 7.2 Place where ANC Was Received . 98 7.3 Pregnancy Complications and Receiving ANC . 99 7.4 Health Care Providers for Pregnancy Complications during ANC visits . 100 7.5 Health Care for Complications during Delivery . 100 7.6 Health Care for Complications during the Postpartum Period . 101 REFERENCES . 109 APPENDIX A SAMPLE DESIGN . 111 A.1 Introduction . 111 A.2 Sample Frame . 111 A.3 Sample Design and Implementation . 111 A.4 Sample Probabilities and Sampling Weights . 116 APPENDIX B ESTIMATES OF SAMPLING ERRORS . 119 APPENDIX C DATA QUALITY TABLES . 129 APPENDIX D INTERNATIONAL CLASSIFICATION OF DISEASES (ICD-10) CODES . 145 APPENDIX E PERSONS INVOLVED IN THE 2019 PAKISTAN MATERNAL MORTALITY SURVEY . 147 APPENDIX F QUESTIONNAIRES . 153 Tables and Figures • v TABLES AND FIGURES 1 INTRODUCTION AND SURVEY METHODOLOGY . 1 Table 1.1 Results of the household, women’s, and verbal autopsy interviews . 8 Figure 1.1 Pakistan Maternal Mortality Survey sample design . 3 2 HOUSEHOLD POPULATION, HOUSEHOLD AND RESPONDENTS’ CHARACTERISTICS . 9 Table 2.1 Household population by age, sex, and residence . 17 Table 2.2 Household population by age, sex, and region . 18 Table 2.3 Household composition . 19 Table 2.4.1 Household drinking water . 19 Table 2.4.2 Treatment of household drinking water . 20 Table 2.5 Household sanitation facilities household members usually use . 20 Table 2.6 Household characteristics . 21 Table 2.7 Household possessions . 22 Table 2.8 Wealth quintiles . 22 Table 2.9 Availability of services in rural areas . 23 Table 2.10 Background characteristics of respondents . 24 Table 2.11 Educational attainment . 25 Figure 2.1 Population pyramid . 10 Figure 2.2 Household drinking water by residence . 11 Figure 2.3 Household toilet facilities by residence . 12 Figure 2.4 Household wealth by residence. 13 Figure 2.5 Education of survey respondents . 15 Figure 2.6 Secondary education by household wealth . 15 3 ADULT AND MATERNAL MORTALITY . 27 Table 3.1 Mortality rates by sex . 40 Table 3.2 Age-specific mortality rates by residence and region . 41 Table 3.3.1 Mortality rates by residence and region: Females . 42 Table 3.3.2 Mortality rates by residence and region: Males . 43 Table 3.4 Adult mortality rates (15-49 years) . 44 Table 3.5 Complete life table for Pakistan . 45 Table 3.6 Complete life table for the total population of Pakistan by sex . 46 Table 3.7 Pregnancy-related mortality . 47 Table 3.8 Pregnancy-related mortality ratio (PRMR) using live births as the denominator (pregnancy-related deaths divided by live births reported in the household survey) . 48 Table 3.9 Maternal mortality . 48 Table 3.10 Maternal mortality ratio . 49 Table 3.11 Maternal mortality ratio using direct method . 49 Figure 3.1 Age-specific mortality rates in the 3 years preceding the survey by sex (log scale) . 28 Figure 3.2 Age-specific mortality rates in the 3 years preceding the survey by residence (log scale) . 29 vi • Tables and Figures Figure 3.3 Age-specific mortality rates in the 3 years preceding the survey by region (log scale) . 29 Figure 3.4 Age-specific mortality rates in the 3 years preceding the survey in Azad Jammu and Kashmir and Gilgit Baltistan (log scale) . 30 Figure 3.5 Age-specific female mortality rates in the 3 years preceding the survey by residence (log scale) . 31 Figure 3.6 Age-specific male mortality rates in the 3 years preceding the survey by residence (log scale) . 31 Figure 3.7 Crude mortality rates in the 3 years preceding the survey by sex and region . 32 Figure 3.8 All-cause adult mortality rates in the 3 years preceding the survey by sex and age . 33 Figure 3.9 All-cause adult mortality rates (15-49 years) in the 3 years preceding the survey by sex and residence . 33 Figure 3.10 All-cause adult mortality rates (15-49 years) in the 3 years preceding the survey by sex and region . 34 Figure 3.11 Life expectancy according to sex . 34 Figure 3.12 Number of females and males living at the beginning of each age interval across the life span . 35 Figure 3.13 Age-specific pregnancy-related mortality ratio trends, 2006-07 PDHS and 2019 PMMS . 37 Figure 3.14 Pregnancy-related mortality ratios by region . 37 Figure 3.15 Age-specific maternal mortality ratio trends, 2006-07 PDHS and 2019 PMMS . 38 Figure 3.16 Maternal mortality ratios by region . 39 4 CAUSES OF DEATH . 51 Table 4.1 Background characteristics of deceased women . 58 Table 4.2 Respondents to Verbal Autopsy Questionnaires . 59 Table 4.3 All cause-specific mortality . 59 Table 4.4 Causes of maternal deaths . 60 Table 4.5 Treatment received for deceased women . 60 Table 4.6 Place of death . 61 Figure 4.1 Physician review process for verbal autopsies . 54 Figure 4.2 All-cause mortality. 55 Figure 4.3 Maternal causes of death . 56 Figure 4.4 Trends in obstetric-coded deaths . 56 Figure 4.5 Treatment received for deceased women . 57 5 MATERNAL HEALTH CARE . 63 Table 5.1 Antenatal care . 71 Table 5.2.1 Antenatal care: Miscarriage and abortion . 72 Table 5.2.2 Antenatal care: Stillbirths . 73 Table 5.2.3 Providers among those receiving antenatal care . 74 Table 5.3 Number of antenatal care visits and timing of first visit . 75 Table 5.4 Components of antenatal care . 76 Table 5.5 Tetanus toxoid injections . 77 Table 5.6 Place of delivery . 78 Table 5.7 Assistance during delivery . 79 Table 5.8.1 Assistance during delivery of stillbirths . 80 Table 5.8.2 Assistance during abortions and miscarriages . 81 Table 5.9 Caesarean section . 82 Table 5.10 Pregnancy outcomes . 83 Tables and Figures • vii Table 5.11 Timing of first postnatal check for the mother . 84 Table 5.12 Type of provider of first postnatal check for the mother . 85 Table 5.13 Combinations of antenatal care, assistance at delivery, and postnatal checks . 86 Figure 5.1 Trends in antenatal care coverage . 64 Figure 5.2 Components of antenatal care . 65 Figure 5.3 Trends in place of birth . 66 Figure 5.4 Health facility births by education . 67 Figure 5.5 Assistance during delivery . 67 Figure 5.6 Skilled assistance at delivery by mother’s education . 68 Figure 5.7 Skilled assistance during ANC, delivery, and PNC by birth order . 69 6 MATERNAL MORBIDITY . 87 Table 6.1 Maternal complications or morbidities reported by women during the last pregnancy . 93 Table 6.2 Maternal complications or morbidities reported by women during the last delivery . 94 Table 6.3 Maternal complications or morbidities reported by women during the postpartum period . 94 Table 6.4 Maternal health complications about which women were informed by a health care provider . 95 Table 6.5 Morbidities reported by women before the last pregnancy . 95 Table 6.6 Maternal complications or morbidities . 96 Figure 6.1 Major maternal health complications that women were told by a health care provider that they had . 90 Figure 6.2 Complications during delivery by birth order . 91 Figure 6.3 Complications during the postpartum period by household wealth . 92 7 HEALTH CARE SEEKING BEHAVIOUR . 97 Table 7.1 Treatment received for maternal complications about which women were informed by a health care provider . 103 Table 7.2 Place where ANC received . 104 Table 7.3 Pregnancy complications and receiving ANC . 105 Table 7.4 Visits to health care providers for complications . 106 Table 7.5 Health care for delivery complications . 107 Table 7.6 Health care for complications during the postpartum period . 108 Figure 7.1 Treatment for maternal health complications . 98 Figure 7.2 Women with pregnancy complications who did not receive ANC by region . 99 Figure 7.3 Skilled assistance during delivery complications by education . 101 Figure 7.4 Skilled assistance during delivery complications by household wealth . 101 APPENDIX A SAMPLE DESIGN . 111 Table A.1 Number of enumeration blocks by region and by type of residence . 111 Table A.2 Sample allocation of enumeration blocks and households by domain and by type of residence . 112 Table A.3 Sample allocation of phase two households and expected number of interviews with women by domain and by type of residence . 112 Table A.4 Sample implementation: Women . 113 Table A.5 Sample implementation (Short Household Questionnaire) . 114 Table A.6 Sample implementation (Long Household Questionnaire): Women . 115 viii • Tables and Figures APPENDIX B ESTIMATES OF SAMPLING ERRORS . 119 Table B.1 List of selected variables for sampling errors, Pakistan MMS 2019 . 120 Table B.2 Sampling errors: Total sample, Pakistan MMS 2019 . 121 Table B.3 Sampling errors: Urban sample, Pakistan MMS 2019 . 121 Table B.4 Sampling errors: Rural sample, Pakistan MMS 2019 . 121 Table B.5 Sampling errors: Punjab sample, Pakistan MMS 2019 . 122 Table B.6 Sampling errors: Punjab urban sample, Pakistan MMS 2019 . 122 Table B.7 Sampling errors: Punjab rural sample, Pakistan MMS 2019 . 122 Table B.8 Sampling errors: Sindh sample, Pakistan MMS 2019 . 123 Table B.9 Sampling errors: Sindh urban sample, Pakistan MMS 2019 . 123 Table B.10 Sampling errors: Sindh rural sample, Pakistan MMS 2019 . 123 Table B.11 Sampling errors: Khyber Pakhtunkhwa sample, Pakistan MMS 2019 . 124 Table B.12 Sampling errors: Khyber Pakhtunkhwa urban sample, Pakistan MMS 2019 . 124 Table B.13 Sampling errors: Khyber Pakhtunkhwa rural sample, Pakistan MMS 2019 . 124 Table B.14 Sampling errors: Balochistan sample, Pakistan MMS 2019 . 125 Table B.15 Sampling errors: Balochistan urban sample, Pakistan MMS 2019 . 125 Table B.16 Sampling errors: Balochistan rural sample, Pakistan MMS 2019 . 125 Table B.17 Sampling errors: Gilgit Baltistan sample, Pakistan MMS 2019 . 126 Table B.18 Sampling errors: Azad Jammu and Kashmir sample, Pakistan MMS 2019 . 126 Table B.19 Sampling errors: Azad Jammu and Kashmir urban sample, Pakistan MMS 2019 . 126 Table B.20 Sampling errors: Azad Jammu and Kashmir rural sample, Pakistan MMS 2019 . 127 Table B.21 Sampling errors for pregnancy-related mortality rates/ratios (PRMR), Pakistan MMS 2019 . 127 Table B.22 Sampling errors for maternal mortality rates/ratios (MMR), Pakistan MMS 2019 . 128 APPENDIX C DATA QUALITY TABLES . 129 Table C.1.1 Household age distribution . 129 Table C.1.2 Household age distribution: Short Household Questionnaire . 130 Table C.1.3 Household age distribution: Long Household Questionnaire . 131 Table C.1.4 Household age distribution: Azad Jammu and Kashmir . 132 Table C.1.5 Household age distribution (short questionnaire): Azad Jammu and Kashmir . 133 Table C.1.6 Household age distribution (long questionnaire): Azad Jammu and Kashmir . 134 Table C.1.7 Household age distribution: Gilgit Baltistan . 135 Table C.1.8 Household age distribution (short questionnaire): Gilgit Baltistan . 136 Table C.1.9 Household age distribution (long questionnaire): Gilgit Baltistan . 137 Table C.2.1 Age distribution of eligible and interviewed women . 138 Table C.2.2 Age distribution of eligible and interviewed women: Azad Jammu and Kashmir . 138 Table C.2.3 Age distribution of eligible and interviewed women: Gilgit Baltistan . 138 Table C.3.1 Completeness of reporting . 139 Table C.3.2 Completeness of reporting: Azad Jammu and Kashmir . 139 Table C.3.3 Completeness of reporting: Gilgit Baltistan . 139 Table C.4.1 Births by calendar years . 140 Table C.4.2 Births by calendar years: Azad Jammu and Kashmir . 140 Table C.4.3 Births by calendar years: Gilgit Baltistan . 141 Table C.5.1 Reporting of age at death in days . 142 Table C.5.2 Reporting of age at death in days: Azad Jammu and Kashmir . 142 Tables and Figures • ix Table C.5.3 Reporting of age at death in days: Gilgit Baltistan . 143 Table C.6.1 Reporting of age at death in months . 143 Table C.6.2 Reporting of age at death in months: Azad Jammu and Kashmir . 144 Table C.6.3 Reporting of age at death in months: Gilgit Baltistan . 144 Foreword • xi FOREWORD aternal mortality information remains as one of the persistent gaps in health indicators worldwide. This constitutes a serious challenge in developing countries, as nearly 90% of all maternal deaths occur in low-income countries. Considering its importance, the United Nations adopted the maternal mortality ratio (MMR) as an indicator of maternal health and set targets of substantially reducing the MMR in the Millennium Development Goals (MDGs) as well as the Sustainable Development Goals (SDGs). This emphasis has resulted in a drop of around 40% in the MMR at the international level since 2000. To assess progress in relation to the SDG targets and the targets of specific country programmes, it is necessary to have access to accurate MMR data. Since its inception 35 years ago, the National Institute of Population Studies (NIPS) has sought to provide evidence-based data for planning and implementation. The 2006-07 Pakistan Demographic and Health Survey (PDHS) included a direct estimate of maternal mortality for the first time in Pakistan. However, this vital information could not be obtained in either the 2012-13 PDHS or the 2017-18 PDHS, mainly as a result of challenges related to resources, sample sizes, and methodologies. NIPS was finally able to meet these challenges, launching the Pakistan Maternal Mortality Survey (PMMS) in collaboration with a Technical Advisory Committee (TAC) consisting of national and international experts. The PMMS is the first exclusive survey in Pakistan with a nationally representative household sample carried out to collect comprehensive information on maternal health issues, maternal mortality, and specific causes of death among women in the country in accordance with international survey design, listing, fieldwork, and data processing and analysis standards. The survey also gathered information on care women received before, during, and following pregnancy and their utilisation of maternal health services. The information collected is intended to help policymakers and programme managers evaluate and design health sector policies, programmes, and strategies for improving maternal health in Pakistan. The key indicators report released in August 2019 shows that Pakistan has made progress in reducing the MMR, which decreased from 276 per 100,000 live births in 2006-07 to 186 in 2019. The availability of the PMMS data opens avenues for in-depth studies to understand the dynamics of MMR reductions in areas of high prevalence. NIPS is indebted to ICF, the Pakistan Bureau of Statistics, the National Committee for Maternal and Neonatal Health, and Dr. Tauseef Ahmed, the principal investigator, for making this survey possible. The PMMS core team, including officers from NIPS and the project staff, implemented the survey. Special appreciation is extended for the support provided by the Ministry of National Health Services, Regulations and Coordination, DFID, UNFPA, USAID, and the Bill and Melinda Gates Foundation. Pervaiz Ahmed Junejo Executive Director, National Institute of Population Studies, Ministry of National Health Services, Regulations and Coordination, Islamabad M Acknowledgements • xiii ACKNOWLEDGEMENTS he 2019 Pakistan Maternal Mortality Survey (PMMS) is the result of the dedicated efforts of several individuals and organisations. The survey was conducted under the aegis of the Ministry of National Health Services, Regulations and Coordination and implemented by the National Institute of Population Studies (NIPS). The Department for International Development (DFID), the United States Agency for International Development (USAID), the United Nations Population Fund (UNFPA), and the Bill and Melinda Gates Foundation provided financial support for the survey. The Pakistan Bureau of Statistics (PBS) assisted in the selection of the sample and the household listing for the sampled primary units. The technical support provided by ICF was invaluable during all stages of the survey. The technical assistance and contributions of the National Committee for Maternal and Neonatal Health in reviewing verbal autopsies and coding causes of death using the ICD-10 to identify maternal deaths were commendable. NIPS is indebted to these organisations. NIPS appreciates the overall supervision, guidance, and dedicated support of Mr. Khizar Hayat Khan, the former Executive Director of NIPS, and acknowledges Dr. Farid Midhet, who highlighted the necessity of the survey with the stakeholders; voluntarily helped in the study design, sampling strategy, and questionnaire development; and remained associated with NIPS during different stages of the project. Mr. Ali Anwar (Fellow) contributed to training, fieldwork, and monitoring. Ms. Rabia Zafar (Fellow) served as a master trainer of the survey teams in addition to managing responsibilities related to fieldwork and monitoring. Mr. Zafar Iqbal Niazi (Administrative Officer), Mr. Muhammad Arif (Accounts Officer), and Mr. Asif Amin Khan (Personal Secretary to the Executive Director) provided logistical support for the project. Ms. Rizwana Timsal (Research Associate), Mr. Qamar Rasool (Data Entry Operator), and Mr. Farman Ali (Data Entry Operator) facilitated research and operational activities. We commend the efforts and dedication of Mr. Zafar Zahir (Operations Advisor), Mr. Muhammad Ali Raza (Data Processing Manager/Data Analyst), and Ms. Gulnaz Mushtaq (Network and Data Supervisor) in project implementation and technical support. Mr. Asif Mehmood (Office Coordinator), Ms. Maida Umer (Office Coordinator), Mr. Junaid Khan (Research Associate), and Mr. Muhammad Ishtiaque, Zeeshan Ali, Farukh Bilal, and Muhammad Saad Alam (Provincial Coordinators) supported fieldwork by coordinating the movement of teams, dispatching questionnaires, and collecting completed questionnaires. Mian Muhammad Arif, Mr. Wajahat Saeed, and Mr. Asghar Ali managed administrative and accounts aspects of the project. We appreciate their contributions. NIPS fully acknowledges the hard work put in by the survey field teams, who collected data under tough and at times hazardous circumstances, and the quality control interviewers for their efficient follow-up and monitoring of the overall fieldwork. We extend special appreciation to the Technical Advisory Committee, which included experts from different fields of population and health. The committee advised and guided different aspects of the survey from the conceptual stage to implementation. The guidance provided by the experts ensured smooth implementation of the survey. We acknowledge with deep gratitude the relentless efforts of Ms. Anjushree Pradhan, Senior Survey Coordinator, ICF, for providing immense technical support during all stages of the project. We extend our thanks to Dr. Ruilin Ren (Sampling Statistician) for his valuable advice on sample design, to Mr. Ruben Hume (Data Processing Specialist) for his contributions in data processing and tabulation, and to all other T xiv • Acknowledgements technical experts from ICF who contributed to the survey. The support provided by ICF experts made it possible to undertake the survey in accordance with international requirements. Ms. Azra Aziz Director (Research and Survey) Team Leader, PMMS Aysha Sheraz, PhD Senior Fellow Deputy Project Director, PMMS Tauseef Ahmed, PhD Principal Investigator, PMMS 2019 Pakistan Maternal Mortality Survey Technical Advisory Committee • xv 2019 PAKISTAN MATERNAL MORTALITY SURVEY TECHNICAL ADVISORY COMMITTEE Mr. Pervaiz Ahmed Junejo, Executive Director, National Institute of Population Studies Chairman Mr. Muhammad Ali Shahzada, Additional Secretary, MNHSR&C Member Dr. Assad Hafeez, Director General, MNHSR&C Member Dr. Hassan Mohtashami, former Country Representative, UNFPA Member Dr. Zeba A. Sathar, Country Director, Population Council Member Dr. Sadiqa Jaffery, President, NCMNCH Member Dr. Farid Midhet, Team Leader, DAFPAK, The Palladium Group Member Mr. Tauseef Ahmed, PhD, Principal Investigator, PMMS Member Dr. Durr-e-Nayab, Joint Director, PIDE Member Dr. Nasir Sarfraz, Health Advisor, DFID Member Mr. Andy Murray, Statistics Adviser and Team Leader, DFID Member Dr. Muhammad Ahmed Isa, Senior Technical Advisor, USAID Member Dr. Yasmeen Sabeeh Qazi, Sr. Country Advisor, The David and Lucile Packard Foundation Member Dr. Nabeela Ali, Chief of Party, JSI Research & Training Institute, Inc. Member Ms. Rabia Awan, Director (Sample & Design), Pakistan Bureau of Statistics Member Dr. Azra Ahsan, Vice President and Technical Consultant, (NCMNH) Member Dr. Zulfiqar Bhutta, Prof./Director, Women and Child Health Centre of Excellence, AKU Member Dr. Nasira Tasneem, Professor/Consultant, MCH Center, PIMS Member Mr. Muqaddar Shah, Program Analyst (Population and Development), UNFPA Member Dr. Saba Shuja, Nutrition Officer, UNICEF Member Dr. Lamia Mahmood, Medical Officer, WHO Member Syed Mubashir Ali, Freelance Consultant Member Dr. G.M. Arif, Principal Investigator, PDHS, National Institute of Population Studies Member Dr. Najma Javed, Senior Medical Officer, PHRC Member Mrs. Azra Aziz, Director (Research and Survey), National Institute of Population Studies Member Dr. Aysha Sheraz, Senior Fellow, National Institute of Population Studies Member Contributors to the Report • xvii CONTRIBUTORS TO THE REPORT National Institute of Population Studies Mr. Pervaiz Ahmed Junejo, Executive Director Ms. Azra Aziz, Director (Research and Survey) Dr. Aysha Sheraz, Senior Fellow/DPD Ms. Saima Mukhtar, Senior Fellow Mr. Ali Anwar Buriro, Fellow Ms. Rabia Zafar, Fellow Ms. Rizwana Timsal, Research Associate Mr. Muhammad Arif, Accounts Officer/DDO External Contributors Dr. Farid Midhet, Team Leader, DAFPAK, The Palladium Group Mr. Tauseef Ahmed, PhD, Principal Investigator, PMMS Mr. Muhammad Ali Raza, DPM, NIPS Dr. Azra Ahsan, Vice President and Technical Consultant, NCMNH Dr. Sarah Saleem, Professor, Head Population and Reproductive Health Section, Department of Community Health Sciences, Aga Khan University Dr. Ejaz Ahmed Khan, Associate Professor, HSA Dr. Shahzad Ali Khan, Professor, Head of Department of Public Health, Health Services Academy Dr. Jasim Anwar, Federal Consultant, CRVS/UN-ESCAP Acronyms and Abbreviations • xix ACRONYMS AND ABBREVIATIONS AIDS acquired immunodeficiency syndrome AJK Azad Jammu and Kashmir ANC antenatal care ASMR age-specific mortality rate BHU basic health unit CAFE computer-assisted field editing CBR crude birth rate CI confidence interval CSPro Census and Survey Processing System DFID Department for International Development DHS Demographic and Health Survey EB enumeration block EmONC emergency obstetric and newborn care FATA Federally Administered Tribal Areas GB Gilgit Baltistan GFR general fertility rate GoP Government of Pakistan ICD International Classification of Diseases ICT Islamabad Capital Territory IFSS internet file streaming system IT information technology LHV lady health visitor LHW lady health worker LPG liquid petroleum gas MCH maternal and child health MDGs Millennium Development Goals MMR maternal mortality ratio MNCH maternal, neonatal, and child health MNH maternal and newborn health MoNHSRC Ministry of National Health Services, Regulations and Coordination NADRA National Database and Registration Authority NCMNH National Committee for Maternal and Neonatal Health NGO nongovernmental organisation NIH National Institutes of Health NIPS National Institute of Population Studies xx • Acronyms and Abbreviations PBS Pakistan Bureau of Statistics PDHS Pakistan Demographic and Health Survey PGPHC Pakistan General Population and Housing Census PHC primary health care PMMS Pakistan Maternal Mortality Survey PNC postnatal care PPS probability proportional to size PRMR pregnancy-related mortality ratio PSU primary sampling unit SDGs Sustainable Development Goals TB tuberculosis TFR total fertility rate TT tetanus toxoid UN United Nations UNFPA United Nations Population Fund USAID United States Agency for International Development VA verbal autopsy VAQ Verbal Autopsy Questionnaire VIP ventilated improved pit WHO World Health Organization Reading and Understanding Tables from the 2019 PMMS • xxi READING AND UNDERSTANDING TABLES FROM THE 2019 PAKISTAN MATERNAL MORTALITY SURVEY (PMMS) he 2019 Pakistan Maternal Mortality Survey final report is based on approximately 60 tables of data. For quick reference, they are located at the end of each chapter and can be accessed through links in the pertinent text (electronic version). This reader-friendly version features about 40 figures that clearly highlight subnational patterns and background characteristics. The text highlights key points in bullets and to clearly identify indicator definitions in boxes. While the text and figures featured in each chapter highlight some of the most important findings from the tables, not every finding can be discussed or displayed graphically. For this reason, PMMS data users should be comfortable reading and interpreting tables. The following pages provide an introduction to the organisation of PMMS tables, the presentation of background characteristics, and a brief summary of sampling and understanding denominators. In addition, this section provides some exercises for users as they practice their new skills in interpreting PMMS tables. T xxii • Reading and Understanding Tables from the 2019 PMMS Example 1: Antenatal Care A Question Asked of Survey Respondents Table 5.1 Antenatal care Percent distribution of ever-married women age 15-49 who had a live birth in the 3 years preceding the survey by antenatal care (ANC) provider during pregnancy for the most recent birth and percentage receiving antenatal care from a skilled provider for the most recent birth, according to background characteristics, Pakistan MMS 2019 Antenatal care provider No ANC Total Percentage receiving antenatal care from a skilled provider1 Number of women Background characteristic Obste- trician/ specialist Doctor Nurse/mid- wife/lady health visitor Community midwives Dai/TBA Other Age at birth <20 35.3 41.9 13.5 0.4 1.2 0.7 7.1 100.0 91.0 447 20-34 47.9 32.5 11.2 0.2 0.8 0.4 7.1 100.0 91.7 3,690 35-49 41.0 31.6 15.0 0.5 0.5 0.7 10.7 100.0 88.1 620 Birth order 1 49.4 36.2 11.0 0.0 0.0 0.1 3.3 100.0 96.7 1,014 2-3 49.3 31.0 10.3 0.3 0.7 0.3 8.1 100.0 90.9 1,845 4-5 43.9 33.4 12.4 0.2 1.6 1.0 7.4 100.0 90.0 1,183 6+ 34.6 34.4 16.5 0.5 0.6 0.8 12.6 100.0 85.9 714 Residence Urban 57.0 31.5 6.8 0.1 0.6 0.5 3.5 100.0 95.4 1,559 Rural 40.3 34.1 14.4 0.3 0.9 0.5 9.6 100.0 89.1 3,197 Education No education 35.8 34.7 14.6 0.4 1.3 0.8 12.4 100.0 85.5 2,439 Primary 45.5 34.1 15.0 0.3 0.0 0.2 4.9 100.0 94.9 833 Middle 55.4 33.5 8.3 0.0 0.3 0.3 2.2 100.0 97.2 408 Secondary 59.7 32.5 6.1 0.0 0.7 0.0 1.0 100.0 98.3 499 Higher 69.5 26.0 3.8 0.0 0.0 0.0 0.7 100.0 99.3 577 Wealth quintile Lowest 33.2 29.1 16.7 0.0 2.1 1.0 17.9 100.0 79.0 1,023 Second 31.2 36.9 18.1 0.9 0.8 0.4 11.7 100.0 87.1 954 Middle 43.3 39.3 11.2 0.3 0.3 0.7 4.8 100.0 94.1 965 Fourth 56.6 33.0 8.4 0.0 0.0 0.2 1.8 100.0 98.0 975 Highest 68.0 27.2 4.0 0.0 0.6 0.0 0.2 100.0 99.2 839 Region Punjab2 54.2 27.1 14.7 0.3 0.8 0.4 2.5 100.0 96.4 2,426 Sindh 41.7 40.5 6.8 0.0 0.9 0.7 9.3 100.0 89.0 1,067 Khyber Pakhtunkhwa3 31.2 41.0 12.4 0.5 0.1 0.6 14.3 100.0 85.0 988 Balochistan 39.8 30.7 5.1 0.2 2.3 0.0 21.9 100.0 75.8 275 Total4 45.8 33.2 11.9 0.2 0.8 0.5 7.6 100.0 91.2 4,756 Azad Jammu and Kashmir 68.3 24.7 4.0 0.1 0.2 0.4 2.2 100.0 97.1 647 Urban 69.9 25.6 1.0 0.5 0.7 0.8 1.5 100.0 97.0 90 Rural 68.0 24.6 4.5 0.0 0.2 0.4 2.4 100.0 97.1 557 Gilgit Baltistan 38.3 44.7 2.5 0.0 0.6 0.1 13.8 100.0 85.5 572 Note: If more than one source of ANC was mentioned, only the provider with the highest qualifications is considered in this tabulation. 1 Skilled provider includes obstetrician/specialist, doctor, nurse, midwife, lady health visitor, and community midwife 2 Punjab includes ICT 3 Khyber Pakhtunkhwa includes merged districts of former FATA 4 Total excludes Azad Jammu and Kashmir and Gilgit Baltistan Step 1: Read the title and subtitle, highlighted in orange in the table above. They tell you the topic and the specific population group being described. In this case, the table is about ever-married women age 15-49 who had a live birth in 3 years before the survey by type of antenatal care provider. All eligible ever- married female respondents age 15-49 were asked these questions. Step 2: Scan the column headings—highlighted in green in Example 1. They describe how the information is categorised. In this table, the first six columns of data show the percent distribution of ever-married women who received ANC by different types of providers – obstetrician/specialist, doctor, nurse/midwife/lady health visitor, community midwife, Dai/traditional birth attendant (TBA), and other. The seventh column shows ever-married women who received no ANC, while the eight column totals 100%, indicating that the first 7 columns are a percent distribution, and that all ever-married women with a live birth in the 3 years before the survey are captured in one of these 7 columns. The ninth column shows 1 2 3 6 5 4 4 Reading and Understanding Tables from the 2019 PMMS • xxiii the percentage of ever-married women who received ANC from a skilled provider (the sum of the obstetrician/specialist, doctor, nurse/midwife/lady health visitor, and community midwife columns). The last column lists the number of ever-married women age 15-49 interviewed in the survey. Step 3: Scan the row headings—the first vertical column highlighted in blue in Example 1. These show the different ways the data are divided into categories based on population characteristics. In this case, the table presents the percent distribution of ever-married women who received ANC by age at birth, birth order, urban-rural residence, level of education, wealth quintile, and region. Most of the tables in the PMMS report will be divided into these same categories. Step 4: Look at the row at the bottom of the table highlighted in red. These percentages represent the totals (excluding Azad Jammu and Kashmir and Gilgit Baltistan) of all ever-married women age 15-49 who received ANC by type of provider. In this case, 45.8% * of ever-married women age 15-49 with a live birth in the 3 years before the survey received ANC from an obstetrician/specialist, 33.2% from a doctor, 11.9% from a nurse/midwife/lady health visitor, and 0.2% from a community midwife. Overall, 91.2% of ever-married women received ANC from a skilled provider – the sum of the totals in the first four columns. Step 5: Scan the last four rows highlighted in grey in Example 1. While the 2019 PMMS collected data in Azad Jammu and Kashmir (AJK) and Gilgit Baltistan (GB), those data are not included in the national total or the background characteristics. The data for these regions are presented separately in the last four rows. For more information on sampling, see Example 3. Step 6: To find out what percentage of ever-married women with higher education received ANC from a nurse/midwife/lady health visitor, draw two imaginary lines, as shown on the table. This shows that 3.8% of ever-married women age 15-49 with higher education received ANC from a nurse/midwife/lady health visitor. By looking at patterns by background characteristics, we can see how ANC coverage varies across Pakistan. *For the purpose of this document data are presented exactly as they appear in the table including decimal places. However, the text in the remainder of this report rounds data to the nearest whole percentage point. Practice: Use the table in Example 1 to answer the following questions: a) What percentage of ever-married women in Pakistan did not receive ANC for their most recent birth? b) Compare ever-married women in urban areas to ever-married women in rural areas—which group is more likely to receive ANC from a nurse/midwife/lady health visitor? c) What are the lowest and highest percentages (range) of ever-married women who received no ANC by region (excluding AJK and GB))? d) Is there a clear relationship in ANC from an obstetrician/specialist by education level? e) Is there a clear relationship in ANC from a skilled provider by wealth quintile? Answers: a) 7.6% b) Ever-married women in rural areas; 14.4% of ever-married women in rural areas received ANC from a nurse/midwife/lady health visitor, compared to 6.8% of urban women. c) Ever-married women who did not receive ANC ranges from a low of 2.5% in Punjab to a high of 21.9% in Balochistan. d) Yes. ANC from an obstetrician specialist increases as a woman’s level of education increases; 35.8% of ever-married women with no education received ANC from an obstetrician/specialist, compared to 69.5% of women with higher education. e) Yes. ANC from a skilled provider increases with household wealth, from 79.0% of ever-married women in the poorest households to 99.2% of ever-married women in the wealthiest households. xxiv • Reading and Understanding Tables from the 2019 PMMS Example 2: Maternal Mortality Direct Estimates of Maternal Mortality Rates and Ratios Table 3.9 Maternal mortality Direct estimates of maternal mortality rates and ratios for the 3 years preceding the survey, by 5-year age groups, residence, and region, Pakistan MMS 2019 Background characteristic Percentage of female deaths that are maternal Number of maternal deaths1 Weighted number of woman years2 Maternal mortality rate3 Maternal mortality ratio4 Age 15-19 13.0 12 117,365 0.10 194 20-24 17.4 19 100,449 0.19 99 25-29 23.4 24 90,591 0.26 115 30-34 29.1 30 68,283 0.44 263 35-39 20.5 31 61,286 0.50 481 40-44 2.7 4 44,828 0.08 286 45-49 0.6 1 41,395 0.03 331 Residence Urban 11.4 32 199,897 0.16 158 Rural 14.5 88 324,300 0.27 199 Region Punjab5 10.5 52 278,770 0.19 157 Sindh 15.7 33 117,149 0.28 224 Khyber Pakhtunkhwa6 15.8 23 99,292 0.23 165 Balochistan 29.2 13 28,987 0.45 298 Total 15-497 13.5 120 524,197 0.23a 186a Azad Jammu and Kashmir 6.4 9 81,048 0.11 104 Gilgit Baltistan 15.8 12 56,225 0.22 157 1 A maternal death is defined as the death of a woman while pregnant or during childbirth or within 42 days after delivery, for which there was a verbal autopsy that classified deaths as being either a direct or indirect maternal death 2 Woman-years lived in that age group during the 36 months before the survey. For example, for the age group 15-19, it is calculated by taking ½ of the number of women age 15, plus 1½ times the number age 16, plus 2½ times the number age 17, plus 3 times the number age 18, plus 3 times the number age 19, plus 2½ times the number age 20, plus 1½ times the number age 21, plus ½ times the number age 22, plus 1½ times the number of deaths to women 15-49 in the previous 36 months. 3 Expressed per 1,000 woman-years of exposure 4 Expressed per 100,000 live births; calculated as the age-adjusted maternal mortality rate times 100 divided by the age-adjusted general fertility rate 5 Punjab includes ICT 6 Khyber Pakhtunkhwa includes the merged districts of former FATA 7 Total excludes Azad Jammu and Kashmir and Gilgit Baltistan a Age-adjusted rate Step 1: Read the title and subtitle. In this case, the table is about direct estimates of maternal mortality rates and ratios for the three-year period before the survey. Step 2: Scan the column headings—highlighted in green. The first column presents the percentage of female deaths that are maternal. The second column shows the number of maternal death cases. For the definition of a maternal death, see footnote 1 at the bottom of the table. A maternal death is defined as the death of a woman while pregnant or during childbirth or within 42 days after delivery, for which there was a verbal autopsy that classified deaths as either a direct or indirect maternal death. The third column shows the weighted number of woman-years or woman-years lived in that age group during the 36 months before the survey (footnote 2). The fourth column shows the maternal mortality rate, and the final column includes the maternal mortality ratio (MMR). The maternal mortality rate is expressed per 1,000 woman- years of exposure (footnote 3), while the maternal mortality ratio is expressed per 100,000 live births (footnote 4). What is the difference between the maternal mortality rate and maternal mortality ratio (MMR)? While the numerator is the same in both indicators (maternal deaths shown in column 2), the MMR uses live births as the denominator and the maternal mortality rate uses the person-years lived by women age 15-49 during the 3 year period before the survey (weighted number of woman-years shown in column 3). To learn more 2 3 4 1 Reading and Understanding Tables from the 2019 PMMS • xxv about how estimates of maternal mortality are calculated, see Chapter 3, section 5, Estimates of Pregnancy-related and Maternal Mortality. To learn about how The DHS Program calculates pregnancy- related and maternal mortality ratios, watch a tutorial video series on The DHS Program’s YouTube channel. Step 3: Scan the row headings—the first vertical column highlighted in blue. The table presents maternal mortality indicators by five-year age groups, urban-rural residence, and region. Step 4: The row near the bottom of the table highlighted in red displays the totals for Pakistan (excluding AJK and GB).  What percentage of female deaths are maternal deaths? It’s 13.5%.  How many maternal deaths were identified by the 2019 PMMS? 120 maternal deaths were included in the survey.  What is the maternal mortality rate? The rate of mortality associated with pregnancy and childbearing is 0.23 maternal deaths per 1,000 woman-years.  What is the MMR in Pakistan? The MMR is 186 maternal deaths per 100,000 live births. Step 5: By looking at patterns by background characteristics, we can see how maternal mortality varies across Pakistan.  Which age group has the highest maternal mortality ratio (MMR)? The MMR is highest in among women age 35-39 at 331 deaths per 100,000 live births.  Is MMR higher in urban areas or rural areas? MMR is higher in rural areas at 199 deaths per 100,000 live births, compared to 158 deaths per 100,000 live births in urban areas.  What is the range in MMR by region (excluding AJK and GB)? MMR is lowest in Punjab at 157 deaths per 100,000 live births and highest in Balochistan at 298 deaths per 100,000 live births. https://www.youtube.com/playlist?list=PLagqLv-gqpTO6YQasXjBnRvPGJpB6jePf xxvi • Reading and Understanding Tables from the 2019 PMMS Example 3: Understanding Sampling Weights in PMMS Tables A sample is a group of people who have been selected for a survey. In the PMMS, the sample is designed to represent the national population age 15-49. In addition to national data, most countries want to collect and report data on smaller geographical or administrative areas. However, doing so requires a large enough sample size in each area. For the 2019 PMMS, the survey sample is representative at the national level; for urban and rural areas separately; for four provinces including Punjab (combined with Islamabad Capital Territory), Sindh, Khyber Pakhtunkhwa (combined with FATA), and Balochistan; and for two regions including Azad Jammu and Kashmir (AJK) and Gilgit Baltistan (GB). To generate statistics that are representative of Pakistan (excluding AJK and GB) and the 4 regions, the number of ever-married women surveyed in each region should contribute to the size of the total (excluding AJK and GB) sample in proportion to the size of the region. However, if some regions have small populations, then a sample allocated in proportion to each region’s population may not include sufficient women from each region for analysis. To solve this problem, regions with small populations are oversampled. For example, let’s say that you have enough money to interview 11,859 women and want to produce results that are representative of Pakistan (excluding AJK and GB) and its 4 regions (as in Table 2.10). However, the total population of Pakistan (excluding AJK and GB) is not evenly distributed among the regions: some regions, such as Punjab, are heavily populated while others, such as Balochistan are not. Thus, Balochistan must be oversampled. A sampling statistician determines how many ever-married women should be interviewed in each region in order to get reliable statistics. The blue column (1) in the table at the right shows the actual number of ever-married women interviewed in each region. Within the regions, the number of women interviewed ranges from 1,779 in Balochistan to 4,387 in Punjab. The number of interviews is sufficient to get reliable results in each region. With this distribution of interviews, some regions are over-represented, and some regions are underrepresented. For example, the population in Punjab is about 53% of the population in Pakistan (excluding AJK and GB), while Balochistan’s population contributes only 5% of the population in Pakistan (excluding AJK and GB). But as the blue column shows, the number of ever-married women interviewed in Punjab accounts for only about 37% of the total sample of ever-married women interviewed (4,387 / 11,859) and the number of ever-married women interviewed in Balochistan accounts for 15% of women interviewed (1,779 / 11,859). This unweighted distribution of ever-married women does not accurately represent the population. In order to get statistics that are representative of Pakistan (excluding AJK and GB), the distribution of the ever-married women in the sample needs to be weighted (or mathematically adjusted) such that it resembles the true distribution in Pakistan (excluding AJK and GB). Ever-married women from a small region, like Balochistan, should only contribute a small amount to the national total. Ever-married women from a large region, like Punjab, should contribute much more. Therefore, DHS statisticians mathematically calculate a “weight” which is used to adjust the number of ever-married women from each region so that each region’s contribution to the total is proportional to the actual population of the region. The numbers in the purple column (2) represent the “weighted” values. The weighted values can be smaller or larger than the unweighted values at the regional level. The total national sample size (excluding AJK and GB) of 11,859 ever-married women has not changed after weighting, but the distribution of the Table 2.10 Background characteristics of respondents Percent distribution of ever-married women age 15-49 by selected background characteristics, Pakistan MMS 2019 Pakistan Background characteristic Weighted percent Weighted number Unweighted number Region Punjab1 53.2 6,308 4,387 Sindh 22.7 2,697 2,857 Khyber Pakhtunkhwa2 19.2 2,271 2,836 Balochistan 4.9 582 1,779 Total3 100.0 11,859 11,859 1 Punjab includes ICT. 2 Khyber Pakhtunkhwa includes the merged districts of FATA. 3 Total excludes Azad Jammu and Kashmir and Gilgit Baltistan. 1 2 3 Reading and Understanding Tables from the 2019 PMMS • xxvii ever-married women in the regions has been changed to represent their contribution to the total population size. How do statisticians weight each category? They take into account the probability that a woman was selected in the sample. If you were to compare the green column (3) to the actual population distribution of Pakistan (excluding AJK and GB), you would see that ever-married women in each region are contributing to the total sample with the same weight that they contribute to the population of Pakistan (excluding AJK and GB). The weighted number of ever-married women in the survey now accurately represents the proportion of ever-married women who live in Punjab and the proportion of ever-married women who live in Balochistan. With sampling and weighting, it is possible to interview enough ever-married women to provide reliable statistics at national (excluding AJK and GB) and regional levels. In general, only the weighted numbers are shown in each of the PMMS tables, so don’t be surprised if these numbers seem low: they may actually represent a larger number of ever-married women interviewed. Introduction and Survey Methodology • 1 INTRODUCTION AND SURVEY METHODOLOGY 1 he 2019 Pakistan Maternal Mortality Survey (PMMS) is the first exclusive nationwide survey on maternal mortality implemented by the National Institute of Population Studies (NIPS) of the Ministry of National Health Services, Regulations and Coordination. Data collection took place from 20 January 2019 to 30 September 2019. ICF provided technical assistance through The DHS Program, which is funded by the United States Agency for International Development (USAID) and offers financial support and technical assistance for population and health surveys in countries worldwide. Other agencies and organisations that facilitated the successful implementation of the survey through technical or financial support were USAID, the United Nations Population Fund (UNFPA), the Department for International Development (DFID), and the Bill and Melinda Gates Foundation. Pakistan was a signatory to the United Nations Millennium Development Goals (MDGs) in 2000 and has focused on Goal 5 (improving maternal health), which set a target for significantly reducing the maternal mortality ratio (MMR) to 140 deaths per 100,000 live births by 2015 by increasing the proportion of births attended by skilled birth attendants and achieving universal access to reproductive health care. The MDG progress assessment showed that Pakistan was on track for Goal 5 but was not close to achieving the set target in 2015 (Government of Pakistan 2013). Pakistan has recently endorsed the UN’s Sustainable Development Goals (SDGs), making a commitment to reducing the MMR to less than 70 per 100,000 live births by 2030 (SDG 3.1) through increased skilled birth attendance, access to modern contraception, and expanded coverage of community health workers as an essential component of universal health coverage. Pursuing the MDG and SDG targets, Pakistan launched a series of initiatives during the past decade and made good progress in safe motherhood (as shown in the results of the 2012-13 PDHS [Pakistan Demographic and Health Survey] and the 2017-18 PDHS). Indirect estimates of the MMR through modelling have shown a substantial decline from 276 per 100,000 live births in the 2006-07 PDHS to 170 (Government of Pakistan 2019). The programmatic initiatives undertaken by the provincial departments of health include the following1: enhanced scope of basic health units to provide around-the-clock basic emergency obstetric and newborn care (EmONC) services, provision of ambulance services for transferring patients at the time of labour (household to basic EmONC and further to a comprehensive EmONC hospital if required), improved knowledge among lady health workers (LHWs) with respect to appropriate referrals (through multiple training opportunities), and a general increase in the number of secondary care hospitals, leading to growth in their utilisation. The information collected through the Pakistan Maternal Mortality Survey is intended to assist policymakers and programme managers in evaluating and designing programmes and strategies for further improving the health of the country’s population. The hope is that this information will enhance the pace towards achieving the goal set for reducing maternal mortality to 70 per 100,000 live births by 2030. 1.1 SURVEY OBJECTIVES The primary objective of the 2019 PMMS is to provide up-to-date estimates of basic demographic and health indicators. Specifically, the survey was designed and carried out with the purpose of assessing where Pakistan stands on maternal health indicators and how well the country is moving toward these targets. Overall aims of the 2019 PMMS were as follows: 1 Community midwives probably have a minor effect on the MMR as their share of normal deliveries remained at less than 5% of public sector deliveries in 2008-2015 and has been less than 3% in 2016 and subsequent years. T 2 • Introduction and Survey Methodology  To estimate national and regional levels of maternal mortality for the 3 years preceding the survey and determine whether the MMR has declined substantially since 2006-07  To identify medical causes of maternal deaths and the biological and sociodemographic risk factors associated with maternal mortality  To assess the impact of maternal and newborn health services, including antenatal and postnatal care and skilled birth attendance, on prevention of maternal mortality and morbidity  To estimate the prevalence and determinants of common obstetric complications and morbidities among women of reproductive age during the 3 years preceding the survey 1.2 SAMPLE DESIGN The 2019 PMMS used a multistage and multiphase cluster sampling methodology based on updated sampling frames derived from the 6th Population and Housing Census, which was conducted in 2017 by the Pakistan Bureau of Statistics (PBS). The sampling universe consisted of urban and rural areas of the four provinces of Pakistan (Punjab, Sindh, Khyber Pakhtunkhwa, and Balochistan), Azad Jammu and Kashmir (AJK), Gilgit Baltistan (GB), Federally Administered Tribal Areas (FATA), and the Islamabad Capital Territory (ICT). A total of 153,560 households (81,400 rural and 72,160 urban) were selected using a two-stage and two-phase stratified systematic sampling approach. The survey was designed to provide representative results for most of the survey indicators in 11 domains: four provinces (by urban and rural areas with Islamabad combined with Punjab and FATA combined with Khyber Pakhtunkhwa), Azad Jammu and Kashmir (urban and rural), and Gilgit Baltistan (see Figure 1.1). Restricted military and protected areas were excluded from the sample. The sampled households were randomly selected from 1,396 primary sampling units (PSUs) (740 rural and 656 urban) after a complete household listing. In each PSU, 110 randomly selected households were administered the various questionnaires included in the survey. All 110 households in each PSU were asked about births and deaths during the previous 3 years, including deaths among women of reproductive age (15-49 years). Households that reported at least one death of a woman of reproductive age were then visited, and detailed verbal autopsies were conducted to determine the causes and circumstances of these deaths to help identify maternal deaths. In the second phase, 10 households in each PSU were randomly selected from the 110 households selected in the first phase to gather detailed information on women of reproductive age. All eligible ever-married women age 15-49 residing in these 10 households were interviewed to gather detailed information, including a complete pregnancy history. A detailed description of the sample design is provided in Appendix A. Introduction and Survey Methodology • 3 Figure 1.1 Pakistan Maternal Mortality Survey sample design 1.3 QUESTIONNAIRES Six questionnaires were used in the 2019 PMMS: the Short Household Questionnaire, the Long Household Questionnaire, the Woman’s Questionnaire, the Verbal Autopsy Questionnaire, the Community Questionnaire, and the Fieldworker Questionnaire. A Technical Advisory Committee was established to solicit comments on the questionnaires from various stakeholders, including representatives of government ministries and agencies, nongovernmental organisations, and international donors. The survey protocol was reviewed and approved by the National Bioethics Committee, the Pakistan Health Research Council, and the ICF Institutional Review Board. After being finalised in English, the questionnaires were translated into Urdu and Sindhi. The 2019 PMMS used paper-based questionnaires for data collection, while computer-assisted field editing (CAFE) was used to edit questionnaires in the field. The Short and Long Household Questionnaires listed all usual household members and visitors who stayed in the selected households the night preceding the interview. Basic demographic information was collected on each person listed, including age, sex, marital status, education, and relationship to head of household. These demographic data were used to identify women who were eligible for an individual interview. The Household Questionnaires also collected information on births and deaths in the household in the 3 years preceding the survey date so that female deaths in the household could be identified and verbal autopsies conducted. In addition, the Long Household Questionnaire collected information on the environmental circumstances of the household, characteristics of the household’s dwelling unit (source of drinking water, type of toilet facilities, and materials used for flooring, external walls, and roofing), and household ownership of assets and various durable goods. The Woman’s Questionnaire was used to collect information from all eligible ever-married women age 15-49. These women were asked questions on the following topics:  Background characteristics (including age and education)  Pregnancy history  Use of family planning methods  Antenatal, delivery, and postnatal care 100 Households selected for Short Questionnaire From each PSU Identify Deceased Women 1,396 (PSUs) Rural 81,400 Urban 72,160 Total Households 153,560 Identify Eligible & Deceased Women 10 Households for Long Questionnaire 4 • Introduction and Survey Methodology  Maternal morbidity  Health service utilisation The Verbal Autopsy Questionnaire was based on the 2016 World Health Organization (WHO) standardised instrument and was primarily adapted from the 2006-07 PDHS for consistency. The questionnaire was finalised after incorporating key inputs from ICF’s health experts in accordance with WHO International Classification of Diseases (ICD-10) guidance, and information was recorded on the circumstances surrounding the event that led to the death, the cause of death, and the health services sought. The questionnaire included both structured (precoded) and unstructured (open-ended) questions that were answered by the member of the household who knew the most about the woman’s last illness and the circumstances of her death. The Community Questionnaire was administered during the fieldwork to collect information on basic infrastructure in the survey clusters and access to health facilities and services. The Community Questionnaire was implemented in rural clusters only. Community representatives who provided information for the questionnaire included, among others, village leaders, counsellors, religious leaders, local teachers, lady health visitors, and lady health workers. The Community Questionnaire was based on the instrument that has been used in PDHS surveys. The Fieldworker Questionnaire recorded background information from the interviewers that will serve as a tool in conducting analyses of data quality. Each interviewer completed a self-administered Fieldworker Questionnaire after the final selection of interviewers and before the fieldworkers entered the field. No personal identifiers are attached to the 2019 PMMS fieldworkers’ data files. 1.3.1 Development of Section 5 (Maternal Morbidity) of the Woman’s Questionnaire A separate section on maternal morbidity was included in the Woman’s Questionnaire in which different types of questions were added to collect information from eligible women, including information on their last pregnancy. Most of the questions related to maternal morbidity were included to obtain data on any problems or illnesses and complications women faced or may have suffered during their last pregnancy. Specific questions were added about abortions or miscarriages, labour pains, delivery complications, postnatal care, treatment received, and use of medicines. Also, questions were added about use of tobacco during pregnancy. To determine maternal health conditions among women, separate information was gathered from them on their experience of having a fever or fever-related symptoms during their last pregnancy or postpartum period (42 days after delivery). Additional questions were asked about fits or seizures, excessive bleeding during pregnancy (before and after delivery or abortion/miscarriage), jaundice, tetanus toxoid injections, use of misoprostol tablets, and so forth. Women were also asked about their utilisation of health service outlets. 1.4 PRETEST Thirty enumerators, eight members of the core project team, and two data processing personnel from NIPS participated in the training to pretest the PMMS protocol; this training was held from 19 November to 6 December 2018. Most participants had previous experience carrying out PDHS surveys and other household surveys. The data processing staff participated in the pretest so that they could familiarise themselves with the survey instruments. ICF provided technical support for the training. Along with discussions on the technical aspects of the survey, the pretest training was designed to train the trainers for the main survey training. The training focused on key components such as age probing, interviewing techniques, and procedures for completing the survey questionnaires. The participants worked in groups using various training techniques, for example interactive question-and-answer sessions, case studies, and role plays. Along with the enumerators, the trainers administered the questionnaires in the Introduction and Survey Methodology • 5 field, provided feedback on the content and language of the questionnaires, and learned the various training techniques. As noted, the questionnaires were translated into Urdu and Sindhi, and the fieldwork for the pretest was carried out in those two languages. Questionnaires were pretested in nonsampled clusters in Islamabad and Rawalpindi. The Verbal Autopsy Questionnaire was pretested using real cases from Lahore, Sukkur, Islamabad, and Peshawar after necessary information from the registers of district hospitals had been collected and family members had been interviewed. Following the fieldwork, a debriefing session was held with the pretest field staff, and modifications to the questionnaires were made based on lessons drawn from the exercise. 1.5 TRAINING OF FIELD STAFF Training of household listers and mappers (67 two-member teams) was organised in the first week of December 2018 to prepare them for identifying the exact locations of PSUs and to show them how to prepare household listings and maps. Training included 15 field supervisors who monitored the household listing operation and undertook a validation of 5% of the listings. The main training for the field staff was held from 17 December 2018 to 6 January 2019 in Islamabad. Separate training was arranged for interviewers selected to conduct verbal autopsies. The participants in the main training included 158 enumerators, selected through a strict process (including a 10% surplus of candidates to account for attrition and quality control staff). Prior to the training, NIPS staff visited the provincial headquarters to screen, interview, and select participants. Applicants came from different parts of Pakistan and represented major language groups within the country. Most of the candidates had previous fieldwork experience, and some had experience gained through PDHS surveys. The training sessions included discussions of concepts, procedures, and methodologies for conducting the survey. Participants were guided through the questionnaires. In-class exercises were carried out in recognition that involving participants in the training process allows them to have a better understanding of the training content. Various techniques were used to facilitate the training, including role-playing on completing the household schedule, age probing in pairs, consistency checking for age and date of birth, correcting errors in pregnancy history tables, and training field editors on using the CAFE system. Special training was organised for interviewers who were selected to administer Verbal Autopsy Questionnaires. Participants in the training were evaluated through classwork, in-class exercises, quizzes, observation during training, and a final test given to all trainees. 1.6 FIELDWORK A comprehensive household listing of all sampled PSUs was conducted in December 2018 to January 2019 to identify sampled households to be visited. Data collection took place from 20 January to 30 September 2019 in all provinces and regions other than Balochistan and Gilgit Baltistan, where fieldwork was completed in October 2019. Forty-one teams consisting of a supervisor, a field editor, and four interviewers were deployed for data collection. All data entry was conducted by the field editors at the end of each day’s fieldwork. Fieldwork monitoring was an integral part of the 2019 PMMS, and several rounds of monitoring were carried out by the core team members and the provincial coordinators. The monitors were provided with guidelines for overseeing the fieldwork. Quality Control Teams and Field Editors focused on various quality of data matters in all regions. The quality and progress of data collection were also monitored through weekly field check tables that were generated from completed interviews received at the NIPS central office, and regular feedback was sent out to the teams and monitors. 6 • Introduction and Survey Methodology 1.7 Quality Control Data quality was a priority of the survey and was ensured through the engagement of 10 quality control teams (each comprising one female and one male evaluator), a proactive information technology (IT) team, and senior management personnel who provided oversight. Quality control teams validated 5% of the Household and Woman’s Questionnaires. The NIPS core team members monitored the field teams during the data collection phase to support them in conducting successful interviews by using structured questionnaires with set procedures of skipping, probing, and understanding the selection of households; using field control sheets; assigning households to female interviewers; checking the field editing of the questionnaires; and observing the supervisors’ efforts to contact PBS regional offices for help in identifying sample areas or with any issues regarding listing and mapping. The core team remained focused on such areas as observing field team coordination, ensuring efficient use of field time, maintaining the condition of vehicles, and checking logbooks. 1.7.1 Verification of Verbal Autopsy Questionnaires The 2019 PMMS data collection process involved a heightened effort to ensure the quality of responses, especially in PSUs with high non-response rates in general or the unavailability of a close relative to allow completion of the verbal autopsy. Furthermore, special attention was given to households in which inconsistencies in reported data were detected during secondary editing, particularly inconsistencies related to female deaths. An exercise involving verification of household responses was conducted using the following methodology. First, PSUs with high non-response rates or inconsistencies in the death section were selected. Second, a randomly selected 20% of households per PSU were visited (and the Short or Long Household Questionnaire re-administered) to cross-check the household roster and the completeness of data on births and deaths for the past 3 years. Third, if a female death was found in the household, a team member undertook a verbal autopsy. Finally, a team member completed any incomplete Verbal Autopsy Questionnaires. Teams also visited households to verify and reinterview or ask verbatim related questions to ensure the completeness of events surrounding female deaths. Verification teams were given blank household questionnaires to be completed during the verification process to cover the selected clusters. A total of 181 clusters were covered during the verification process. In all of the verification revisits, different teams and quality control teams were assigned to conduct the interviews. 1.8 DATA PROCESSING The processing of the 2019 PMMS data began simultaneously with the fieldwork. As soon as data collection was completed in each cluster, all electronic data files were transferred via the Internet File Streaming System (IFSS) to the NIPS central office in Islamabad. These data files were registered and checked for inconsistencies, incompleteness, and outliers. A double entry procedure was adopted by NIPS to ensure data accuracy. The field teams were alerted about any inconsistencies and errors. Secondary editing of completed questionnaires, which involved resolving inconsistencies and coding open-ended questions, was carried out in the central office. The survey core team members assisted with secondary editing, and the NIPS data processing manager coordinated the work at the central office. Data entry and editing were carried out using the CSPro software package. The concurrent processing of the data offered a distinct advantage because it maximised the likelihood of the data being error-free and accurate. Assessment of Verbal Autopsy Questionnaires The Verbal Autopsy Questionnaires were reviewed and coded based on ICD-10 categories to determine causes of death. NIPS organised a workshop on ICD-10 coding from 29 July to 2 August 2019 to provide an orientation for experts and reviewers from Pakistan’s National Committee for Maternal and Neonatal Health (NCMNH). The workshop was supported by ICF. Introduction and Survey Methodology • 7 NCMNH provided technical assistance to identify causes and determinants of maternal deaths based on a review of Verbal Autopsy Questionnaires and to assign causes of death. The Verbal Autopsy Questionnaires were reviewed by six panels of experts; each panel included two obstetricians and one general physician based at NCMNH. For each female death, reviewers used prescribed forms to identify the category of the death. Categories were as follows: direct obstetric death, indirect obstetric death, probable obstetric death, coincidental obstetric death, late maternal death, non-obstetric death, and undecided category of death. As part of quality assurance, each Verbal Autopsy Questionnaire was independently reviewed and assigned a cause (or causes) of death by two experts from the panel and validated by the third expert. In cases in which the findings of the two experts were discordant, the questionnaire was re-reviewed by the lead expert for final assignment of the category and cause of death. Cause of death coding was carried out by separate trained personnel. To help the reviewers summarise and comprehend the complex data, a checklist was prepared to list the main signs and symptoms of the fatal illness and thus facilitate assignment of a cause of death. A condition was regarded as the main cause of death if it was entered in each of the three cause of death forms as an immediate, underlying, possible, or associated cause of death. The timing of death in relation to pregnancy, delivery, or the postpartum period was categorised as follows: (1) death during pregnancy (before any signs of labour or abortion), (2) death during or less than 24 hours after delivery, (3) death during or less than 24 hours after abortion or miscarriage, (4) death more than 24 hours but less than 42 days after delivery, (5) death more than 24 hours but less than 42 days after abortion/miscarriage, (6) death more than 42 days but less than 1 year after delivery or abortion/miscarriage, (7) cannot be determined (insufficient information), (8) not applicable (not a pregnancy-related death), and (9) delay in seeking treatment (all three types of delay). The main cause of death (immediate and underlying), possible cause(s) of death (immediate and underlying), associated cause(s) of death (if any), delays in receiving treatment during a fatal illness, and the reviewer’s assessment of the quality of the data were determined. 1.9 RESPONSE RATES Table 1.1 shows the response rates for the 2019 Pakistan Maternal Mortality Survey. In the four provinces, the sample contained a total of 116,169 households. All households were visited by the field teams, and 110,483 households were found to be occupied. Of these households, 108,766 were successfully interviewed, yielding a household response rate of 98%. The subsample selected for the Long Household Questionnaire comprised 11,080 households, and interviews were carried out in 10,479 of these households. A total of 12,217 ever-married women age 15-49 were eligible to be interviewed based on the Long Household Questionnaire, and 11,859 of these women were successfully interviewed (a response rate of 97%). In Azad Jammu and Kashmir, 16,755 households were occupied, and interviews were successfully carried out in 16,588 of these households (99%). A total of 1,707 ever-married women were eligible for individual interviews, of whom 1,666 were successfully interviewed (98%). In Gilgit Baltistan, 11,005 households were occupied, and interviews were conducted in 10,872 households (99%). A total of 1,219 ever-married women were eligible for interviews, of whom 1,178 were successfully interviewed (97%). A total of 944 verbal autopsy interviews were conducted in Pakistan overall, 150 in Azad Jammu and Kashmir, and 88 in Gilgit Baltistan. The Verbal Autopsy Questionnaire was used in almost all of the interviews, and response rates were nearly 100%. 8 • Introduction and Survey Methodology Table 1.1 Results of the household, women’s, and verbal autopsy interviews Number of households, number of interviews, and response rates, according to residence (unweighted), Pakistan MMS 2019 Pakistan Azad Jammu and Kashmir Gilgit Baltistan Result Urban Rural Total Urban Rural Total Urban Rural Total Household interviews (total) Households selected 57,510 58,659 116,169 8,558 8,952 17,510 3,293 8,460 11,753 Households occupied 54,649 55,834 110,483 8,159 8,596 16,755 3,071 7,934 11,005 Households interviewed 53,510 55,256 108,766 8,064 8,524 16,588 3,061 7,811 10,872 Household response rate1 97.9 99.0 98.4 98.8 99.2 99.0 99.7 98.4 98.8 Household interviews (short questionnaire) Households selected 52,120 52,969 105,089 7,758 8,102 15,860 2,993 7,620 10,613 Households occupied 49,495 50,344 99,839 7,396 7,781 15,177 2,789 7,135 9,924 Households interviewed 48,474 49,813 98,287 7,307 7,721 15,028 2,779 7,025 9,804 Household response rate1 97.9 98.9 98.4 98.8 99.2 99.0 99.6 98.5 98.8 Household interviews (long questionnaire) Households selected 5,390 5,690 11,080 800 850 1,650 300 840 1,140 Households occupied 5,154 5,490 10,644 763 815 1,578 282 799 1,081 Households interviewed 5,036 5,443 10,479 757 803 1,560 282 786 1,068 Household response rate1 97.7 99.1 98.4 99.2 98.5 98.9 100.0 98.4 98.8 Interviews with ever-married women Number of eligible women 5,747 6,470 12,217 803 904 1,707 317 902 1,219 Number of eligible women interviewed 5,540 6,319 11,859 777 889 1,666 309 869 1,178 Eligible women response rate2 96.4 97.7 97.1 96.8 98.3 97.6 97.5 96.3 96.6 Verbal autopsy interviews Number of verbal autopsies/ deceased women selected 416 528 944 67 83 150 18 70 88 Number of verbal autopsy interviews 412 528 940 67 82 149 18 70 88 Eligible verbal autopsy response rate3 99.0 100.0 99.6 100.0 98.8 99.3 100.0 100.0 100.0 1 Households interviewed/households occupied 2 Women interviewed/eligible women 3 Verbal autopsies selected/verbal autopsies conducted Household Population, Household and Respondents’ Characteristics • 9 HOUSEHOLD POPULATION, HOUSEHOLD AND RESPONDENTS’ CHARACTERISTICS 2 Key Findings  Population age distribution: 40% of the household population is less than age 15 and 4% is age 65 or above.  Household composition: The average household size is 6.7 persons; 74% of households have more than 4 persons.  Drinking water: 97% of households have an improved drinking water source.  Sanitation: 79% of households have an improved sanitation facility, and 67% use a basic sanitation service (an improved facility that is not shared with other households).  Education: 52% of women overall and 91% of women in the lowest wealth quintile have no education. nformation on the socioeconomic characteristics of the household population in the 2019 PMMS provides a context to interpret demographic and health indicators and can furnish an approximate indication of the representativeness of the survey. In addition, this information sheds light on the living conditions of the population. This chapter presents information on sources of household drinking water, treatment of drinking water, sanitation, exposure to smoke inside the home, wealth status, household population and composition, educational attainment, availability of services in rural areas, and background characteristics of respondents. 2.1 HOUSEHOLD POPULATION AND COMPOSITION Household A person or group of related or unrelated persons who live together in the same dwelling unit(s), who acknowledge one adult male or female as the head of the household, who share the same housekeeping arrangements, and who are considered a single unit. De facto population All persons who stayed in the selected households the night before the interview (whether usual residents or visitors). I 10 • Household Population, Household and Respondents’ Characteristics De jure population All persons who are usual residents of the selected households, whether or not they stayed in the household the night before the interview. How data are calculated All tables are based on the de facto population unless otherwise specified. The distribution of the population by age, sex, and residence is a primary demographic classification. It also provides information on trends in population growth based on the size of age brackets, including dependency groups. The de facto survey population (those who stayed in the surveyed households the night before the survey) is 728,135. The sex ratio (number of males per 100 females) is 100 (Table 2.1). Forty percent of the population is under age 15 and 4% is age 65 and above, with the bulk of the population in the 15-64 age brackets (Figure 2.1). The pattern essentially depicts a high fertility rate with slight recent declines. The age dependency ratio (the sum of the population under age 15 and the population age 65 and over relative to the working-age population age 15-64) is 0.77. When categorised by different age groups, it is interesting to note that 46% of the total population falls in the 0-17 group, whereas the adult population (age 18 or above) is 54% of the total. The adolescent population age 10-19 accounts for 23% of the total population, and the young adult population age 15-24 accounts for 20%. Table 2.2 shows the distribution of the population by region. In Khyber Pakhtunkhwa and Balochistan, 15% of the population is under age 5, whereas in Punjab and Sindh 13% is under age 5. The percentages are similar in Azad Jammu and Kashmir (13%) and Gilgit Baltistan (15%). The high percentage of children under age 5 in all regions indicates a high fertility rate. The percentage of children age 5-9 is highest in Balochistan (17%) and lowest in Azad Jammu and Kashmir (12%). This may be an indicator of family size preference in these regions. Thirteen percent of children in Balochistan and Khyber Pakhtunkhwa are age 10-14, which is slightly higher than the percentage is Sindh (12%) and Punjab (11%). The percentage of the population age 65 or older is highest in Azad Jammu and Kashmir (6%) and lowest in Balochistan and Sindh (3% each). Balochistan has the highest percentage of the population age 0-14 (46%), while the highest percentage of the population age 15-64 (58%) is in Punjab and Sindh. In general, Pakistani households are large, with an average household size of 6.7 persons (Table 2.3). Mean household sizes are 6.1 and 7.6 persons, respectively, in Azad Jammu and Kashmir and Gilgit Baltistan. Pakistani households are predominantly headed by men (91%). Ten percent of households in rural areas are headed by women, compared with 8% in urban areas. By far the highest percentage of households that are headed by women is in Azad Jammu and Kashmir (23%) Figure 2.1 Population pyramid 10 6 2 2 6 10 <5 5-9 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79 80+ Age Percent distribution of the household population Male Female 2610 Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan Household Population, Household and Respondents’ Characteristics • 11 Trends: The population distribution has varied slightly over the past 14 years. The percentage of the population less than age 15 was 41% in the 2006-07 PDHS (a survey of comparable design) and 40% in 2019. The size of the adolescent population has decreased slightly since 2006-07 (from 25% to 23%), while there has been a slight increase in the size of the population age 15-64 (from 55% to 57%). The sex ratio in 2019 was 100, compared with 102 in 2006-07. Patterns by background characteristics  The proportion of the population age 0-17 is higher in rural areas (48%) than in urban areas (42%).  The average household size is larger in rural (6.9) than urban (6.3) areas. 2.2 DRINKING WATER SOURCES AND TREATMENT Improved sources of drinking water Include piped water, public taps, standpipes, tube wells, boreholes, protected dug wells and springs, rainwater, filtration plants, water delivered via a tanker truck or a cart with a small tank, and bottled water. Sample: Households Improved sources of drinking water are essential for the health of the population and provide safeguards against contamination. In Pakistan, 97% of households use an improved source of drinking water (Table 2.4.1 and Figure 2.2). A tube well or borehole (51%) is the predominant source of drinking water, followed by piped water (23%). Among other improved sources, filtered drinking water is used by 9% of households. A tube well or borehole is the most common source in rural areas (63%), followed by piped water (17%). Thirty-three percent of urban households use piped water, and 31% use a tube well or borehole. Seventy-four percent of households have drinking water on their premises, while 5% of households spend more than 30 minutes to obtain drinking water. Only 6% of households follow appropriate water treatment practices prior to drinking water. Appropriate treatment practices are followed more often in urban areas (12%) than in rural areas (2%) (Table 2.4.2). Figure 2.2 Household drinking water by residence 23 33 17 5 5 5 51 31 63 2 1 4 5 9 22 5 09 16 5 3 1 4 Total Urban Rural Percent distribution of households by source of drinking water Unimproved source Filtration plant Bottled water Tanker truck/cart with small tank Protected well or spring Tube well or borehole Public tap/ standpipe Piped water into dwelling/yard/plot/ neighbour’s yard Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan 12 • Household Population, Household and Respondents’ Characteristics 2.3 SANITATION Improved toilet facilities Flush/pour flush toilets that flush water and waste to a piped sewer system, septic tank, pit latrine, or unknown destination; ventilated improved pit (VIP) latrines; pit latrines with slabs; or composting toilets. Sample: Households Target 6.2 of the Sustainable Development Goals is to provide adequate and equitable sanitation for all. This target is tracked with the indicator of safely managed sanitation services or use of an improved type of sanitation facility that is not shared with other households (United Nations 2018). Improved sanitation facilities help to prevent communicable diseases such as cholera and typhoid. Overall, 79% of households (69% in rural areas and 95% in urban areas) use improved sanitation facilities (Figure 2.3). Five percent of households have an unimproved sanitation facility, and 16% have no toilet facility (25% in rural areas and 1% in urban areas) (Table 2.5). Among households having an improved toilet facility, the facility is almost always in the household dwelling (96% in urban areas and 91% in rural areas). In only 1% of households is the facility elsewhere. Basic sanitation service Use of improved facilities that are not shared with other households. Sample: De jure population Limited sanitation service Use of improved facilities shared by two or more households. Sample: De jure population Overall, 69% of de jure household members have basic sanitation service, while 10% have limited service (Table 2.5). 2.4 HOUSEHOLD CHARACTERISTICS The survey collected data on access to electricity, flooring materials, the number of rooms used for sleeping, and other household characteristics. A vast majority (94%) of the households in Pakistan (99% in urban areas and 91% in rural areas) have access to electricity (Table 2.6). In Pakistan, most households use cement (40%) or earth/sand (35%) as materials for flooring. Earth and sand are more commonly used in rural households (51%), and cement is most common in urban households (56%). Nine percent of all households have marble flooring. Forty percent of households have only one room for sleeping, 39% have two rooms for sleeping, and 21% have three or more rooms for sleeping. Ninety-three percent of households have a place for cooking within the dwelling. Liquefied petroleum gas (LPG) or natural gas is the most common cooking fuel (49% of households). Use of this fuel varies Figure 2.3 Household toilet facilities by residence 79 95 69 5 4 6 16 1 25 Total Urban Rural Percent distribution of households by type of toilet facilities No facility/ bush/field Unimproved facility Improved facility Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan Household Population, Household and Respondents’ Characteristics • 13 dramatically from urban households (86%) to rural households (26%). Overall, 53% of residents use solid fuel for cooking (e.g., wood, agricultural crops, or animal dung), while 47% rely on clean fuel (mostly LPG or natural gas). 2.5 HOUSEHOLD WEALTH Household Durable Goods The survey collected information on household possessions, means of transportation, and farm animals (Table 2.7). Mobile phones and televisions are the most common information and communication devices used in Pakistan. Almost all households (95%) have at least one mobile phone, and 4% have land-line phones (6% in urban areas and 2% in rural areas). More than 6 in 10 households (62%) in Pakistan own a television (84% in urban areas and 48% in rural areas). Three percent of urban households and 6% of rural households own a radio. As expected, ownership of farm animals is more common among rural (59%) than urban (14%) households. Overall, 64% of urban households and 51% of rural households have a motorbike or scooter as a means of transportation. Only 11% of urban households and 5% of rural households own a car or truck. Wealth Index Wealth index Households are given scores based on the number and kinds of consumer goods they own, ranging from a television to a bicycle or car, and housing characteristics such as source of drinking water, toilet facilities, and flooring materials. These scores are derived using principal component analysis. National wealth quintiles are compiled by assigning the household score to each usual (de jure) household member, ranking each person in the household population by her or his score, and then dividing the distribution into five equal categories, each comprising 20% of the population. Sample: Households Table 2.8 presents data on wealth quintiles and the Gini coefficient according to residence, region, and province. The Gini coefficient indicates the level of concentration of wealth, with 0 representing an equal wealth distribution and 1 representing a totally unequal distribution. Pakistan’s Gini coefficient is 0.28, indicating a somewhat uneven distribution of wealth in the population. In urban areas, 41% of the population is in the highest wealth quintile and only 3% is in the lowest quintile. Conversely, in rural areas only 8% of the population is in the highest wealth quintile, while 30% is in the lowest quintile (Figure 2.4). Among the provinces, 26% of the population in Punjab is in the highest wealth quintile and 11% is in the lowest wealth quintile. Sindh has a higher level of poverty than Punjab, with 33% of the population falling in the lowest wealth quintile. In Khyber Pakhtunkhwa, 18% of the population is in the lowest wealth quintile. In that region, the highest percentage of the population falls in the second wealth quintile (33%). The percentage of the population in the lowest wealth quintile in Balochistan (45%) is alarming. Only 5% of the population in Balochistan falls in the Figure 2.4 Household wealth by residence 3 30 8 27 19 21 29 15 41 8 Urban Rural Percent distribution of de jure population by wealth quintiles Highest Fourth Middle Second Lowest Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan 14 • Household Population, Household and Respondents’ Characteristics highest wealth quintile. Poverty is higher in rural areas than urban areas in all of the provinces, especially in rural Sindh, where 70% of the population is in the lowest wealth quintile. In Azad Jammu and Kashmir, 32% of the population falls in the middle wealth quintile and 28% in the fourth quintile. Only 7% of the population in this region is in the lowest wealth quintile. In Gilgit Baltistan, 25% of the population is in the lowest wealth quintile and just 2% in the highest quintile. 2.6 SERVICES IN RURAL AREAS The 2019 PMMS Community Questionnaire was administered in each of the 610 selected rural sample points. It included questions about the availability of various public services, such as district headquarters, post offices, banks, educational institutions, shops, transportation services, and health facilities. As the data were provided by community informants and distances to services were not verified, the results should be viewed with some caution. Table 2.9 shows the percent distribution of rural households by distance to various services. Ninety-four percent of households in rural areas have electricity coverage in their community, and 93% have mobile phone coverage. Three-quarters of rural households have access to a general store or shop, and 65% have access to motorised public transport. Although television signals are available to 86% of rural households, only 36% have access to cable television connections. A post office is available to only 20% of rural households, while 2-12% of households have access to National Database and Registration Authority (NADRA) offices, banks, and courier services. A large majority of rural households have primary schools located in their community. However, it is noteworthy that primary schools for boys are more prevalent than primary schools for girls (85% versus 80%). The majority of rural households (64%) are 10 or more kilometres from degree-granting colleges. With respect to health services, only 17% of rural households have a functioning basic health unit in their community, while 32% have one within 1-4 kilometres. Only 9% of rural households have a female doctor in their community; 50% have to travel 10 or more kilometres to see a female doctor when needed. Also, the majority of households are 10 or more kilometres away from ultrasound or ambulance services, a functioning maternal and child health centre, rural health centre, or family welfare centre. The health- related personnel most likely to be available in the community are traditional birth attendants (dais) (51%) and dispensers/compounders (39%). 2.7 BASIC CHARACTERISTICS OF SURVEY RESPONDENTS The survey results indicate that 42% of ever-married women age 15-49 are under age 30 (Table 2.10). Ninety-five percent of women are currently married. Women are more likely to reside in rural areas (63%) than in urban areas (37%). More than half of women have no education. The distribution of women across wealth quintiles is generally equal, with 18% being in the lowest wealth quintile and 22% in the highest quintile. Women are most likely to reside in Punjab (53%) and least likely to reside in Balochistan (5%). In Azad Jammu and Kashmir, 34% of women are under age 30, 95% are currently married, and 84% live in rural areas. More than one-quarter (28%) have no education. One-third of respondents in Azad Jammu and Kashmir are in the middle wealth quintile, while only 6% fall in the lowest quintile. Forty-two percent of women in Gilgit Baltistan are under age 30, 97% are married, and 83% reside in rural areas. Half have no education. Twenty-three percent of women fall in the lowest wealth quintile, while only 9% fall in the upper two quintiles. Household Population, Household and Respondents’ Characteristics • 15 2.8 EDUCATIONAL ATTAINMENT Median educational attainment Half of the population has completed less than the median number of years of schooling, and half of the population has completed more than the median number of years of schooling. Sample: Ever-married women age 15-49 Education and access to information are important in determining behaviours. They are also crucial in developing approaches toward reproductive and maternal health. The 2019 PMMS results showed that one in two ever-married women have no education (52%), while 18% have a primary education. Overall, 23% of ever-married women have a secondary or higher level of education (Table 2.11). In Azad Jammu and Kashmir, about half of urban women (51%) have a secondary or higher education. The median number of years of education in Azad Jammu and Kashmir is 6.2, which is higher than in other provinces and Gilgit Baltistan. Trends: The percentages of educated women in different age brackets depict a mixed trend. In 2019, the highest percentage of uneducated women (67%) were age 45-49, and 14% of women in this age group had a secondary or higher level of education. In 2006-07, 79% women in this age group were uneducated, and only 8% had a secondary or higher education. Patterns by background characteristics  Among ever-married women age 15-49, the highest percentages with no education are in the older age brackets (Table 2.11). Two-thirds of women age 45-49 have no education, compared with 49% of women age 15-24.  Sixty-two percent of rural women have no education, compared with 34% of urban women (Figure 2.5). Urban women (39%) are more likely to have a secondary or higher education than their rural counterparts (14%). The median years of education is 4.9 in urban areas and 0.0 in rural areas.  Women in the highest wealth quintile (61%) are much more likely than those in the lowest wealth quintile (1%) to have a secondary or higher education (Table 2.11 and Figure 2.6).  Nine percent of women in Balochistan have a secondary or higher level of education, compared with 26% of women in Punjab. Thirty- four percent of women in Azad Jammu and Kashmir have a secondary or higher education. Figure 2.5 Education of survey respondents Figure 2.6 Secondary education by household wealth 34 62 17 18 10 6 17 722 7 Urban Rural Percent distribution of ever-married women age 15-49 by highest level of schooling attended or completed Higher Secondary Middle Primary No education Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan 1 4 10 32 61 Lowest Second Middle Fourth Highest Percentage of ever-married women age 15-49 with a secondary education or higher WealthiestPoorest Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan 16 • Household Population, Household and Respondents’ Characteristics  Four in five rural women in Sindh (82%) and Balochistan (80%) have no education. Two in five urban women in Punjab (41%) and Sindh (40%) have a secondary or higher education. LIST OF TABLES For more information on household population and housing characteristics, see the following tables:  Table 2.1 Household population by age, sex, and residence  Table 2.2 Household population by age, sex, and region  Table 2.3 Household composition  Table 2.4.1 Household drinking water  Table 2.4.2 Treatment of household drinking water  Table 2.5 Household sanitation facilities household members usually use  Table 2.6 Household characteristics  Table 2.7 Household possessions  Table 2.8 Wealth quintiles  Table 2.9 Availability of services in rural areas  Table 2.10 Background characteristics of respondents  Table 2.11 Educational attainment Household Population, Household and Respondents’ Characteristics • 17 Table 2.1 Household population by age, sex, and residence Percent distribution of the de facto household population by various age groups, dependency age groups, and child and adult populations, and percentage of adolescents and young people, according to sex and residence, Pakistan MMS 2019 Urban Rural Total Age Male Female Total Male Female Total Male Female Total <5 12.4 12.0 12.2 15.3 14.2 14.7 14.2 13.4 13.8 5-9 12.3 12.0 12.1 15.4 14.3 14.8 14.2 13.4 13.8 10-14 11.5 11.3 11.4 12.8 11.7 12.2 12.3 11.5 11.9 15-19 11.0 11.0 11.0 11.0 11.1 11.0 11.0 11.0 11.0 20-24 9.4 10.2 9.8 7.9 9.2 8.6 8.5 9.6 9.0 25-29 8.4 9.2 8.8 6.9 8.5 7.7 7.5 8.8 8.1 30-34 7.0 7.3 7.1 5.7 6.4 6.1 6.2 6.7 6.5 35-39 6.2 6.4 6.3 5.2 5.9 5.6 5.6 6.1 5.8 40-44 5.0 4.9 4.9 4.0 4.0 4.0 4.4 4.3 4.4 45-49 4.4 4.5 4.5 3.7 4.0 3.8 4.0 4.1 4.0 50-54 3.6 3.3 3.4 3.0 2.9 3.0 3.2 3.1 3.1 55-59 2.6 2.5 2.5 2.4 2.4 2.4 2.5 2.4 2.4 60-64 2.5 2.1 2.3 2.3 2.0 2.2 2.4 2.0 2.2 65-69 1.6 1.3 1.4 1.6 1.4 1.5 1.6 1.4 1.5 70-74 1.2 0.9 1.0 1.2 1.0 1.1 1.2 0.9 1.1 75-79 0.5 0.4 0.5 0.6 0.5 0.5 0.6 0.4 0.5 80+ 0.7 0.6 0.7 0.8 0.7 0.8 0.8 0.7 0.7 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Dependency age groups 0-14 36.1 35.2 35.7 43.5 40.1 41.8 40.8 38.4 39.6 15-64 60.0 61.5 60.7 52.2 56.3 54.3 55.1 58.2 56.6 65+ 3.9 3.2 3.6 4.3 3.5 3.9 4.2 3.4 3.8 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Child and adult populations 0-17 42.5 41.6 42.1 50.0 46.7 48.3 47.2 44.9 46.1 18+ 57.5 58.4 57.9 50.0 53.3 51.7 52.8 55.1 53.9 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Percentage of adolescents age 10-19 22.5 22.3 22.4 23.8 22.7 23.3 23.3 22.6 22.9 Percentage of young people age 15-24 20.4 21.2 20.8 18.9 20.3 19.6 19.5 20.6 20.0 Number of persons 134,382 130,773 265,154 229,956 233,025 462,981 364,337 363,798 728,135 Note: Table excludes Azad Jammu and Kashmir and Gilgit Baltistan. 18 • Household Population, Household and Respondents’ Characteristics Ta bl e 2. 2 H ou se ho ld p op ul at io n by a ge , s ex , a nd r eg io n P er ce nt d is tri bu tio n of th e de fa ct o ho us eh ol d po pu la tio n by v ar io us a ge g ro up s, d ep en de nc y ag e gr ou ps , a nd c hi ld a nd a du lt po pu la tio ns , a nd p er ce nt ag e of a do le sc en ts a nd y ou ng p eo pl e, a cc or di ng to s ex a nd re gi on , P ak is ta n M M S 2 01 9 R eg io n P un ja b S in dh K hy be r P ak ht un kh w a B al oc hi st an A za d Ja m m u an d K as hm ir G ilg it B al tis ta n A ge M al e Fe m al e To ta l M al e Fe m al e To ta l M al e Fe m al e To ta l M al e Fe m al e To ta l M al e Fe m al e To ta l M al e Fe m al e To ta l <5 13 .7 12 .7 13 .2 13 .6 13 .0 13 .3 15 .8 15 .0 15 .4 15 .4 15 .3 15 .4 13 .9 11 .5 12 .6 15 .7 13 .9 14 .8 5- 9 13 .5 12 .5 13 .0 13 .8 13 .8 13 .8 15 .7 14 .6 15 .1 17 .6 16 .5 17 .1 13 .6 11 .4 12 .4 15 .5 13 .8 14 .7 10 -1 4 11 .8 10 .9 11 .4 12 .1 11 .9 12 .0 13 .4 12 .2 12 .8 14 .0 12 .9 13 .4 13 .0 11 .3 12 .1 13 .2 13 .0 13 .1 15 -1 9 10 .6 10 .9 10 .8 11 .4 11 .1 11 .3 11 .5 11 .2 11 .4 11 .0 11 .4 11 .2 11 .0 10 .8 10 .9 11 .6 12 .0 11 .8 20 -2 4 8. 5 9. 9 9. 2 9. 0 9. 5 9. 3 8. 0 9. 1 8. 6 7. 7 8. 6 8. 1 8. 1 9. 6 8. 9 7. 2 8. 8 8. 0 25 -2 9 7. 6 9. 0 8. 3 8. 0 8. 8 8. 4 6. 6 8. 4 7. 5 7. 3 8. 2 7. 7 6. 5 8. 6 7. 6 6. 1 8. 0 7. 1 30 -3 4 6. 3 6. 9 6. 6 6. 6 7. 0 6. 7 5. 5 6. 2 5. 9 5. 7 6. 3 6. 0 5. 4 7. 1 6. 3 5. 2 6. 2 5. 7 35 -3 9 5. 8 6. 2 6. 0 5. 9 6. 1 6. 0 4. 7 5. 8 5. 3 5. 1 5. 3 5. 2 4. 9 6. 7 5. 9 5. 0 5. 3 5. 2 40 -4 4 4. 6 4. 5 4. 6 4. 6 4. 5 4. 5 3. 7 3. 9 3. 8 3. 7 3. 5 3. 6 4. 0 4. 8 4. 4 4. 1 4. 2 4. 1 45 -4 9 4. 2 4. 4 4. 3 4. 0 4. 3 4. 1 3. 5 3. 5 3. 5 3. 1 3. 6 3. 3 4. 0 4. 6 4. 3 3. 6 3. 5 3. 6 50 -5 4 3. 5 3. 3 3. 4 3. 2 2. 8 3. 0 2. 7 2. 7 2. 7 2. 6 2. 5 2. 5 3. 3 3. 4 3. 3 2. 6 2. 7 2. 6 55 -5 9 2. 7 2. 5 2. 6 2. 3 2. 5 2. 4 2. 4 2. 3 2. 3 1. 8 1. 9 1. 8 2. 8 2. 7 2. 7 2. 0 2. 2 2. 1 60 -6 4 2. 5 2. 2 2. 3 2. 3 1. 9 2. 1 2. 3 1. 9 2. 1 1. 9 1. 6 1. 8 2. 9 2. 5 2. 7 2. 4 2. 0 2. 2 65 -6 9 1. 7 1. 5 1. 6 1. 4 1. 2 1. 3 1. 7 1. 4 1. 5 1. 2 1. 0 1. 1 2. 1 1. 6 1. 8 1. 9 1. 7 1. 8 70 -7 4 1. 4 1. 1 1. 3 1. 0 0. 7 0. 9 1. 1 0. 8 0. 9 1. 0 0. 7 0. 8 1. 8 1. 3 1. 5 1. 7 1. 1 1. 4 75 -7 9 0. 6 0. 5 0. 6 0. 4 0. 3 0. 4 0. 6 0. 5 0. 5 0. 3 0. 3 0. 3 1. 1 0. 7 0. 9 0. 8 0. 6 0. 7 80 + 1. 0 0. 8 0. 9 0. 5 0. 5 0. 5 0. 6 0. 6 0. 6 0. 6 0. 4 0. 5 1. 6 1. 3 1. 5 1. 3 0. 9 1. 1 To ta l 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 D ep en de nc y ag e gr ou ps 0- 14 39 .0 36 .2 37 .6 39 .5 38 .8 39 .1 45 .0 41 .7 43 .3 46 .9 44 .7 45 .9 40 .5 34 .2 37 .1 44 .4 40 .8 42 .6 15 -6 4 56 .3 59 .9 58 .1 57 .3 58 .5 57 .9 51 .0 55 .1 53 .1 50 .0 52 .9 51 .4 52 .9 60 .9 57 .1 49 .9 54 .9 52 .5 65 + 4. 8 3. 9 4. 3 3. 2 2. 7 3. 0 4. 0 3. 2 3. 6 3. 0 2. 4 2. 7 6. 6 5. 0 5. 7 5. 7 4. 3 5. 0 To ta l 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 C hi ld a nd ad ul t po pu la tio ns 0- 17 45 .1 42 .6 43 .8 46 .1 45 .3 45 .7 52 .1 48 .4 50 .2 53 .5 51 .5 52 .6 47 .1 40 .6 43 .7 51 .8 48 .4 50 .1 18 + 54 .9 57 .4 56 .2 53 .9 54 .7 54 .3 47 .9 51 .6 49 .8 46 .4 48 .5 47 .4 52 .9 59 .4 56 .3 48 .2 51 .6 49 .9 To ta l 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 10 0. 0 P er ce nt ag e of ad ol es ce nt s 10 -1 9 22 .4 21 .8 22 .1 23 .5 23 .0 23 .3 25 .0 23 .4 24 .2 25 .0 24 .3 24 .7 24 .0 22 .1 23 .0 24 .8 25 .0 24 .9 P er ce nt ag e of y ou ng pe op le 15 -2 4 19 .1 20 .8 19 .9 20 .4 20 .6 20 .5 19 .5 20 .3 19 .9 18 .7 20 .0 19 .4 19 .2 20 .4 19 .8 18 .8 20 .9 19 .8 N um be r o f pe rs on s 18 7, 32 5 18 9, 22 1 37 6, 54 6 83 ,7 88 79 ,9 34 16 3, 72 2 70 ,8 31 73 ,1 30 14 3, 96 2 22 ,3 93 21 ,5 12 43 ,9 05 48 ,4 71 54 ,1 76 10 2, 64 8 40 ,9 42 42 ,0 29 82 ,9 71 Household Population, Household and Respondents’ Characteristics • 19 Table 2.3 Household composition Percent distribution of households by sex of head of household and household size, and mean size of households, according to residence, Pakistan MMS 2019 Pakistan Azad Jammu and Kashmir Gilgit Baltistan Characteristic Urban Rural Total Urban Rural Total Urban Rural Total Household headship Male 92.1 89.6 90.6 84.0 76.2 77.5 94.6 91.8 92.3 Female 7.9 10.4 9.4 16.0 23.8 22.5 5.4 8.2 7.7 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Number of usual members 1 1.5 1.1 1.2 1.6 1.6 1.6 0.5 0.7 0.7 2 5.0 4.7 4.8 5.5 4.6 4.8 1.7 2.8 2.6 3 8.2 7.6 7.8 8.6 8.6 8.6 4.4 4.1 4.1 4 13.4 11.3 12.1 14.5 12.8 13.1 8.4 8.5 8.5 5 17.1 14.5 15.5 18.7 17.3 17.6 13.0 13.3 13.2 6 17.1 15.3 16.0 17.7 16.6 16.8 16.8 14.3 14.7 7 11.8 12.8 12.4 11.2 13.1 12.7 15.2 13.3 13.7 8 8.4 9.5 9.1 8.0 9.0 8.8 11.2 11.0 11.0 9+ 17.5 23.2 21.0 14.1 16.4 16.0 28.9 31.9 31.4 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Mean size of households 6.3 6.9 6.7 5.9 6.2 6.1 7.5 7.6 7.6 Number of households 41,607 67,159 108,766 2,827 13,761 16,588 1,924 8,948 10,872 Note: Table is based on de jure household members, i.e., usual residents. Table 2.4.1 Household drinking water Percent distribution of households and de jure population by source of drinking water and by time to obtain drinking water, percentage of households and de jure population with basic drinking water service, and percentage with limited drinking water service, according to residence, Pakistan MMS 2019 Households Population Characteristic Urban Rural Total Urban Rural Total Source of drinking water Improved source 98.8 96.1 97.1 98.9 95.9 97.0 Piped into dwelling/yard/plot 31.5 15.2 21.4 31.7 15.7 21.5 Piped to neighbour 1.4 1.7 1.6 1.6 1.7 1.7 Public tap/standpipe 4.6 4.7 4.7 4.7 4.8 4.8 Tube well or borehole 31.0 63.4 51.0 31.8 62.1 51.0 Protected dug well 0.6 2.4 1.7 0.6 2.7 2.0 Protected spring 0.1 1.2 0.7 0.1 1.3 0.9 Rainwater 0.0 0.1 0.0 0.0 0.1 0.0 Tanker truck/cart with small tank 8.5 2.1 4.5 8.7 2.8 4.9 Bottled water 4.7 0.4 2.0 3.8 0.3 1.6 Filtration plant 16.3 5.1 9.4 15.9 4.3 8.6 Unimproved source 1.2 3.9 2.9 1.1 4.1 3.0 Unprotected dug well 0.1 1.4 0.9 0.1 1.5 1.0 Unprotected spring 0.1 0.8 0.5 0.1 0.8 0.6 Surface water 0.2 1.5 1.0 0.2 1.6 1.1 Other 0.9 0.1 0.4 0.7 0.1 0.3 Total 100.0 100.0 100.0 100.0 100.0 100.0 Time to obtain drinking water (round trip) Water on premises1 74.1 74.4 74.2 74.8 75.0 74.9 30 minutes or less 22.0 18.7 20.0 21.1 17.9 19.1 More than 30 minutes 2.5 6.5 4.9 2.6 6.5 5.0 Don’t know/missing 1.4 0.4 0.8 1.6 0.7 1.0 Total 100.0 100.0 100.0 100.0 100.0 100.0 Percentage with basic drinking water service2 95.0 91.1 92.6 94.9 90.7 92.3 Percentage with limited drinking water service3 3.7 5.0 4.5 4.0 5.2 4.8 Number of households/population 4,009 6,470 10,479 25,574 44,346 69,920 Note: Table excludes Azad Jammu and Kashmir and Gilgit Baltistan. 1 Includes water piped to a neighbour and those reporting a round-trip collection time of zero minutes 2 Defined as drinking water from an improved source, provided either water is on the premises or round-trip collection time is 30 minutes or less. Includes safely managed drinking water, which is not shown separately. 3 Drinking water from an improved source, and round-trip collection time is more than 30 minutes or is unknown 20 • Household Population, Household and Respondents’ Characteristics Table 2.4.2 Treatment of household drinking water Percentage of households and de jure population using various methods to treat drinking water, and percentage using an appropriate treatment method, according to residence, Pakistan MMS 2019 Households Population Water treatment method Urban Rural Total Urban Rural Total Boil 10.5 1.6 5.0 10.2 1.5 4.7 Bleach/chlorine added 0.6 0.1 0.3 0.5 0.1 0.2 Strain through cloth 4.1 1.8 2.7 4.5 1.9 2.8 Ceramic, sand, or other filter 1.2 0.2 0.6 1.2 0.2 0.6 Let stand and settle 0.2 0.4 0.3 0.2 0.5 0.4 Other 0.6 0.0 0.2 0.7 0.0 0.3 No treatment 84.2 96.0 91.5 84.1 96.0 91.6 Percentage using an appropriate treatment method1 12.1 1.9 5.8 11.7 1.7 5.4 Number of households/population 4,009 6,470 10,479 25,574 44,346 69,920 Note: Table excludes Azad Jammu and Kashmir and Gilgit Baltistan. Respondents may report multiple treatment methods, so the sum of treatment may exceed 100%. 1 Appropriate water treatment methods include boiling, bleaching, filtering, and solar disinfecting. Table 2.5 Household sanitation facilities household members usually use Percent distribution of households and de jure population by type of toilet/latrine facilities household members usually use, percent distribution of households and de jure population with a toilet/latrine facility by location of the facility, percentage of households and de jure population with basic sanitation service, and percentage with limited sanitation service, according to residence, Pakistan MMS 2019 Households Population Type and location of toilet/latrine facility Urban Rural Total Urban Rural Total Improved sanitation facility 94.5 69.2 78.8 94.3 69.8 78.7 Flush/pour flush to piped sewer system 58.2 5.9 25.9 57.0 6.4 24.9 Flush/pour flush to septic tank 28.4 43.8 37.9 28.4 43.0 37.7 Flush/pour flush to pit latrine 6.6 16.8 12.9 7.5 17.4 13.8 Flush/pour flush, don’t know where 0.8 0.7 0.7 0.8 0.7 0.7 Ventilated improved pit (VIP) latrine 0.2 0.2 0.2 0.2 0.1 0.2 Pit latrine with slab 0.4 1.6 1.1 0.4 2.0 1.4 Composting toilet 0.0 0.1 0.1 0.1 0.2 0.1 Unimproved sanitation facility 4.1 6.1 5.3 4.3 6.5 5.7 Flush/pour flush not to sewer/septic tank/pit latrine 3.6 4.1 3.9 3.6 4.0 3.9 Pit latrine without slab/open pit 0.3 1.4 1.0 0.4 1.8 1.3 Bucket 0.0 0.3 0.2 0.0 0.4 0.3 Hanging toilet/hanging latrine 0.0 0.0 0.0 0.0 0.1 0.0 Other 0.2 0.3 0.3 0.2 0.3 0.2 Open defecation (no facility/bush/field) 1.4 24.8 15.8 1.4 23.7 15.6 Total 100.0 100.0 100.0 100.0 100.0 100.0 Number of households/population 4,009 6,470 10,479 25,574 44,346 69,920 Location of toilet facility In own dwelling 96.2 91.2 93.4 96.7 91.4 93.6 In own yard/plot 3.3 7.2 5.5 3.0 7.3 5.4 Elsewhere 0.4 1.5 1.0 0.2 1.2 0.8 Total 100.0 100.0 100.0 100.0 100.0 100.0 Number of households/population with a toilet/latrine facility 3,952 4,868 8,821 25,208 33,830 59,039 Percentage with basic sanitation service1 86.1 55.9 67.4 86.7 58.0 68.5 Percentage with limited sanitation service2 8.4 13.3 11.4 7.6 11.7 10.2 Number of households/population 4,009 6,470 10,479 25,574 44,346 69,920 Note: Table excludes Azad Jammu and Kashmir and Gilgit Baltistan. 1 Defined as use of improved facilities that are not shared with other households. Includes safely managed sanitation service, which is not shown separately. 2 Defined as use of improved facilities shared by 2 or more households Household Population, Household and Respondents’ Characteristics • 21 Table 2.6 Household characteristics Percent distribution of households and de jure population by housing characteristics, percentage using solid fuel for cooking, and percentage using clean fuel for cooking, according to residence, Pakistan MMS 2019 Households Population Housing characteristic Urban Rural Total Urban Rural Total Electricity Yes 99.3 90.5 93.9 99.3 90.6 93.8 No 0.7 9.5 6.1 0.7 9.4 6.2 Total 100.0 100.0 100.0 100.0 100.0 100.0 Flooring material Earth/sand 8.2 51.4 34.9 8.8 52.3 36.4 Dung 0.5 4.3 2.9 0.5 4.0 2.7 Wood/planks 0.0 0.0 0.0 0.0 0.1 0.0 Ceramic tiles 3.8 1.1 2.1 3.5 1.1 2.0 Cement 55.8 29.5 39.5 55.9 28.9 38.8 Carpet 1.2 0.6 0.8 1.2 1.0 1.1 Chips/terrazzo 8.8 2.0 4.6 8.6 1.8 4.3 Bricks 4.9 5.5 5.3 4.9 5.4 5.2 Mats 0.7 0.7 0.7 0.8 0.9 0.9 Marble 16.1 4.7 9.1 15.6 4.5 8.6 Other 0.1 0.1 0.1 0.0 0.1 0.1 Total 100.0 100.0 100.0 100.0 100.0 100.0 Rooms used for sleeping One 34.8 43.0 39.8 26.2 33.2 30.7 Two 41.9 37.1 38.9 40.7 36.8 38.2 Three or more 23.2 19.8 21.1 33.0 29.9 31.0 Total 100.0 100.0 100.0 100.0 100.0 100.0 Place for cooking In the house 93.2 92.4 92.7 93.6 92.6 93.0 In a separate building 6.2 6.0 6.1 6.0 6.3 6.2 Outdoors 0.2 1.1 0.8 0.2 1.0 0.7 No food cooked in household 0.4 0.5 0.5 0.2 0.1 0.1 Total 100.0 100.0 100.0 100.0 100.0 100.0 Cooking fuel Electricity 0.3 0.2 0.2 0.2 0.1 0.2 LPG/natural gas/biogas 86.0 25.9 48.9 85.0 24.7 46.8 Coal/lignite 0.1 0.1 0.1 0.1 0.1 0.1 Charcoal 0.6 1.6 1.2 0.6 1.6 1.2 Wood 10.4 50.1 34.9 11.4 53.0 37.8 Straw/shrubs/grass 0.2 2.6 1.7 0.3 2.7 1.9 Agricultural crop 1.0 9.7 6.4 1.0 8.9 6.0 Animal dung 1.1 9.5 6.3 1.2 8.7 6.0 No food cooked in household 0.4 0.5 0.5 0.2 0.1 0.1 Total 100.0 100.0 100.0 100.0 100.0 100.0 Percentage using solid fuel for cooking1 13.4 73.5 50.5 14.6 75.0 52.9 Percentage using clean fuel for cooking2 86.2 26.0 49.1 85.2 24.8 46.9 Number of households/population 4,009 6,470 10,479 25,574 44,346 69,920 Note: Table excludes Azad Jammu and Kashmir and Gilgit Baltistan. LPG = Liquefied petroleum gas 1 Includes coal/lignite, charcoal, wood, straw/shrubs/grass, agricultural crops, and animal dung 2 Includes electricity and LPG/natural gas/biogas 22 • Household Population, Household and Respondents’ Characteristics Table 2.7 Household possessions Percentage of households possessing various household effects, means of transportation, and livestock/farm animals, by residence, Pakistan MMS 2019 Residence Total Possession Urban Rural Household effects Radio 2.6 6.3 4.9 Television 83.9 47.9 61.7 Mobile phone 97.8 92.4 94.5 Watch 60.7 51.4 54.9 Non-mobile telephone 6.2 1.8 3.5 Computer 19.5 5.7 11.0 Refrigerator 75.2 45.2 56.6 Almirah/cabinet 76.7 44.6 56.9 Chair 64.5 45.8 53.0 Room cooler 23.2 13.3 17.1 Air conditioner 17.0 3.6 8.8 Washing machine 81.8 44.9 59.1 Water pump 68.4 46.2 54.7 Bed 77.8 56.1 64.4 Clock 81.3 53.1 63.9 Sofa 50.0 27.2 35.9 Camera 5.9 2.4 3.7 Sewing machine 70.2 54.5 60.5 Internet connection 15.9 2.7 7.7 Means of transport Bicycle 15.9 14.6 15.1 Animal-drawn cart 2.2 9.6 6.7 Motorcycle/scooter 63.6 50.5 55.5 Car/truck 10.7 5.3 7.4 Tractor 0.6 3.7 2.5 Boat without a motor 0.0 0.1 0.1 Rickshaw/chingchi 3.0 2.1 2.4 Ownership of farm animals1 13.8 59.4 42.0 Number 4,009 6,470 10,479 Note: Table excludes Azad Jammu and Kashmir and Gilgit Baltistan. 1 Cows, bulls, other cattle, horses, donkeys, mules, goats, sheep, camels, chickens, or other poultry Table 2.8 Wealth quintiles Percent distribution of the de jure population by wealth quintiles, and the Gini coefficient, according to residence and region, Pakistan MMS 2019 Wealth quintile Total Number of persons Gini coefficient Residence/region Lowest Second Middle Fourth Highest Residence Urban 3.3 7.7 18.6 29.2 41.3 100.0 25,423 0.16 Rural 29.5 27.0 20.8 14.8 8.0 100.0 44,950 0.29 Region Punjab1 11.1 18.6 21.5 22.3 26.3 100.0 35,177 0.21 Urban 0.7 5.9 17.4 26.0 49.9 100.0 12,891 0.14 Rural 17.2 26.0 23.9 20.2 12.7 100.0 22,286 0.23 Sindh 33.1 11.8 13.6 21.5 20.0 100.0 16,506 0.35 Urban 4.4 7.1 18.0 35.7 34.8 100.0 9,303 0.18 Rural 70.2 17.8 7.9 3.1 0.9 100.0 7,203 0.32 Khyber Pakhtunkhwa2 18.1 33.2 24.8 14.7 9.1 100.0 13,810 0.30 Urban 5.9 12.4 21.7 25.9 34.2 100.0 2,004 0.21 Rural 20.2 36.7 25.3 12.9 4.9 100.0 11,806 0.29 Balochistan 44.9 20.3 16.9 13.0 5.0 100.0 4,880 0.30 Urban 17.8 23.1 30.2 17.6 11.2 100.0 1,224 0.23 Rural 53.9 19.3 12.5 11.4 2.9 100.0 3,656 0.30 Total3 20.0 20.0 20.0 20.0 20.0 100.0 70,373 0.28 Azad Jammu and Kashmir 6.8 22.2 31.8 27.6 11.7 100.0 9,613 0.20 Urban 1.0 8.9 23.8 38.2 28.2 100.0 1,606 0.15 Rural 7.9 24.9 33.4 25.4 8.4 100.0 8,006 0.20 Gilgit Baltistan 24.7 44.7 22.3 6.0 2.3 100.0 8,312 0.25 1 Punjab includes ICT. 2 Khyber Pakhtunkhwa includes merged districts of former FATA. 3 Total excludes Azad Jammu and Kashmir and Gilgit Baltistan. Household Population, Household and Respondents’ Characteristics • 23 Table 2.9 Availability of services in rural areas Percent distribution of rural households by distance to selected services from their communities, Pakistan MMS 2019 Number of kilometres to service Total Number Service In community1 1-4 km 5-9 km 10+ km Don’t know/ missing Distance not asked Administrative services District headquarters 0.2 2.5 5.3 91.6 0.5 na 100.0 89,527 Post office 20.2 21.6 20.8 35.9 1.5 na 100.0 89,527 Courier services 4.9 12.5 17.6 61.2 3.8 na 100.0 89,527 Bank 11.8 17.9 19.0 48.5 2.8 na 100.0 89,527 NADRA office 1.8 6.6 15.9 71.7 4.1 na 100.0 89,527 Public call office 9.1 na na na na 90.9 100.0 89,527 Health services Hospital 7.4 14.3 20.3 54.3 3.6 na 100.0 89,527 Functioning basic health unit 16.5 31.7 26.5 24.1 1.2 na 100.0 89,527 Rural health centre 6.3 14.2 19.3 53.8 6.3 na 100.0 89,527 Functioning government dispensary 15.5 24.2 22.2 34.8 3.3 na 100.0 89,527 Functioning MCH centre 5.9 15.9 17.0 56.4 4.8 na 100.0 89,527 Female doctor 8.5 17.1 21.7 49.7 3.0 na 100.0 89,527 Private doctor 12.9 18.1 21.8 44.4 2.8 na 100.0 89,527 Dispenser/compounder 39.2 26.1 16.1 16.7 2.0 na 100.0 89,527 Family welfare centre 10.3 13.5 22.0 49.9 4.3 na 100.0 89,527 Hakeem 16.2 12.5 19.6 48.1 3.6 na 100.0 89,527 Dai 51.1 18.4 11.1 17.9 1.4 na 100.0 89,527 Homeopath 12.1 12.7 16.5 55.2 3.6 na 100.0 89,527 Any ambulance service 9.3 12.6 17.9 56.2 3.9 na 100.0 89,527 Ultrasound service 8.3 12.9 19.2 56.1 3.6 na 100.0 89,527 Medical store 35.1 19.1 17.6 27.5 0.7 na 100.0 89,527 Transportation services Motorised public transport 64.6 14.9 8.9 10.4 1.2 na 100.0 89,527 Non-motorised public transport 60.3 8.8 6.0 11.3 13.6 na 100.0 89,527 Educational services Primary school for boys 84.9 11.0 2.9 1.0 0.1 na 100.0 89,527 Primary school for girls 79.7 11.6 4.9 3.1 0.6 na 100.0 89,527 Secondary school for boys 37.5 22.1 18.7 19.4 2.2 na 100.0 89,527 Secondary school for girls 30.6 22.0 19.8 25.2 2.3 na 100.0 89,527 Degree college for boys or girls 3.5 10.9 17.4 63.6 4.5 na 100.0 89,527 Basic infrastructure services Wastewater drainage scheme 14.5 na na na na 85.5 100.0 89,527 Sewerage system 11.0 na na na na 89.0 100.0 89,527 Drinking water scheme 26.4 na na na na 73.6 100.0 89,527 Television signal/service 86.0 na na na na 14.0 100.0 89,527 Cable television connections 35.9 na na na na 64.1 100.0 89,527 Land-line telephone service 30.7 na na na na 69.3 100.0 89,527 Mobile telephone coverage 93.4 na na na na 6.6 100.0 89,527 Electricity 93.8 na na na na 6.2 100.0 89,527 Gas connection 17.6 na na na na 82.4 100.0 89,527 General store or shop 74.5 9.3 6.8 9.2 0.2 na 100.0 89,527 Note: Table is based on community profile of rural clusters. na = Not applicable NADRA = National Database and Registration Authority MCH = Maternal and Child Health 1 Includes responses of “0” kilometres 24 • Household Population, Household and Respondents’ Characteristics Table 2.10 Background characteristics of respondents Percent distribution of ever-married women age 15-49 by selected background characteristics, Pakistan MMS 2019 Pakistan Azad Jammu and Kashmir Gilgit Baltistan Background characteristic Weighted percent Weighted number Unweighted number Weighted percent Weighted number Unweighted number Weighted percent Weighted number Unweighted number Age 15-19 5.1 604 633 2.5 41 37 4.7 55 47 20-24 15.5 1,839 1,828 12.9 215 196 15.6 183 190 25-29 21.0 2,486 2,459 18.3 305 310 21.8 257 248 30-34 18.0 2,139 2,176 19.0 317 321 17.0 200 197 35-39 16.8 1,987 1,963 18.3 304 319 15.3 181 190 40-44 12.1 1,432 1,411 15.0 249 248 13.5 159 162 45-49 11.6 1,373 1,389 14.1 234 235 12.2 143 144 Marital status Married 95.2 11,290 11,382 95.2 1,586 1,581 97.0 1,143 1,142 Divorced/separated 1.8 214 156 2.3 38 32 0.6 7 7 Widowed 3.0 355 321 2.5 42 53 2.4 28 29 Residence Urban 37.0 4,386 5,540 16.1 269 777 17.2 203 309 Rural 63.0 7,473 6,319 83.9 1,397 889 82.8 975 869 Education No education 51.7 6,131 6,477 28.3 471 405 50.1 590 573 Primary1 17.8 2,108 1,770 19.0 317 296 11.3 133 140 Middle2 7.7 912 823 18.5 308 296 10.9 129 127 Secondary3 10.4 1,239 1,222 16.9 282 298 13.5 159 156 Higher4 12.4 1,469 1,567 17.3 288 371 14.2 167 182 Wealth quintile Lowest 18.0 2,139 2,395 5.9 98 64 22.5 265 260 Second 19.3 2,289 2,286 22.2 370 289 44.9 529 509 Middle 19.7 2,333 2,231 32.5 541 497 23.4 276 293 Fourth 21.1 2,501 2,267 27.6 460 523 6.7 79 89 Highest 21.9 2,597 2,680 11.9 198 293 2.4 28 27 Region Punjab5 53.2 6,308 4,387 na na na na na na Urban 20.1 2,379 2,089 na na na na na na Rural 33.1 3,929 2,298 na na na na na na Sindh 22.7 2,697 2,857 na na na na na na Urban 12.5 1,488 1,356 na na na na na na Rural 10.2 1,209 1,501 na na na na na na Khyber Pakhtunkhwa6 19.2 2,271 2,836 na na na na na na Urban 2.9 342 1,259 na na na na na na Rural 16.3 1,929 1,577 na na na na na na Balochistan 4.9 582 1,779 na na na na na na Urban 1.5 177 836 na na na na na na Rural 3.4 406 943 na na na na na na Total 100.0 11,859 11,859 100.0 1,666 1,666 100.0 1,178 1,178 na = Not applicable 1 Primary refers to classes 1-5. 2 Middle refers to classes 6-8. 3 Secondary refers to classes 9-10. 4 Higher refers to classes 11 and above. 5 Punjab includes ICT. 6 Khyber Pakhtunkhwa includes the merged districts of former FATA. Household Population, Household and Respondents’ Characteristics • 25 Table 2.11 Educational attainment Percent distribution of ever-married women age 15-49 by highest level of schooling completed, and median years completed, according to background characteristics, Pakistan MMS 2019 Highest level of schooling Total Median years completed Number of women Background characteristic No education Primary1 Middle2 Secondary3 Higher4 Age 15-24 48.8 20.4 9.4 11.7 9.7 100.0 1.1 2,443 15-19 51.3 23.7 12.0 9.8 3.1 100.0 0.0 604 20-24 48.0 19.3 8.5 12.3 11.9 100.0 1.4 1,839 25-29 44.3 16.7 9.9 12.0 17.2 100.0 3.6 2,486 30-34 45.2 21.3 7.6 10.3 15.5 100.0 2.7 2,139 35-39 55.6 16.4 7.1 9.6 11.3 100.0 0.0 1,987 40-44 59.1 15.0 4.7 10.9 10.3 100.0 0.0 1,432 45-49 66.9 14.4 4.8 6.5 7.3 100.0 0.0 1,373 Residence Urban 33.6 17.4 10.4 16.9 21.7 100.0 4.9 4,386 Rural 62.3 18.0 6.1 6.6 6.9 100.0 0.0 7,473 Wealth quintile Lowest 91.2 6.9 1.1 0.6 0.2 100.0 0.0 2,139 Second 73.8 18.3 3.9 2.4 1.5 100.0 0.0 2,289 Middle 57.1 25.3 7.5 5.9 4.2 100.0 0.0 2,333 Fourth 32.0 23.6 12.8 18.6 12.9 100.0 4.6 2,501 Highest 13.7 13.9 11.7 21.8 38.8 100.0 9.4 2,597 Region Punjab5 41.7 22.0 9.8 11.4 15.0 100.0 3.8 6,308 Urban 27.2 20.4 11.9 16.7 23.9 100.0 6.5 2,379 Rural 50.5 23.0 8.6 8.3 9.6 100.0 0.0 3,929 Sindh 57.1 13.4 5.4 11.9 12.2 100.0 0.0 2,697 Urban 36.5 14.7 8.4 19.8 20.5 100.0 4.9 1,488 Rural 82.4 11.8 1.7 2.2 1.9 100.0 0.0 1,209 Khyber Pakhtunkhwa6 66.8 13.3 5.0 7.5 7.4 100.0 0.0 2,271 Urban 48.5 11.9 9.5 12.3 17.8 100.0 1.8 342 Rural 70.0 13.6 4.3 6.6 5.6 100.0 0.0 1,929 Balochistan 76.2 9.6 5.2 4.5 4.5 100.0 0.0 582 Urban 66.5 10.0 7.9 5.4 10.2 100.0 0.0 177 Rural 80.4 9.5 4.1 4.0 2.0 100.0 0.0 406 Total7 51.7 17.8 7.7 10.4 12.4 100.0 0.0 11,859 Azad Jammu and Kashmir 28.3 19.0 18.5 16.9 17.3 100.0 6.2 1,666 Urban 16.3 16.0 17.2 18.6 32.1 100.0 8.2 269 Rural 30.6 19.6 18.8 16.6 14.4 100.0 5.0 1,397 Gilgit Baltistan 50.1 11.3 10.9 13.5 14.2 100.0 0.0 1,178 1 Primary refers to classes 1-5. 2 Middle refers to classes 6-8. 3 Secondary refers to classes 9-10. 4 Higher refers to classes 11 and above. 5 Punjab includes ICT. 6 Khyber Pakhtunkhwa includes the merged districts of former FATA. 7 Total excludes Azad Jammu and Kashmir and Gilgit Baltistan. Adult and Maternal Mortality • 27 ADULT AND MATERNAL MORTALITY 3 Key Findings  Adult mortality: During the 3 years before the survey, the mortality rate among women age 15-49 was 1.72 per 1,000 persons, and the mortality rate among men age 15- 49 was 2.48 per 1,000. The mortality rate in the 15-49 age bracket is almost 50% higher among men than women.  Life expectancy: A newborn child in Pakistan in 2019 is expected to reach age 65.4. A newborn girl is expected to live approximately 2 years longer (66.5 years) than a newborn boy (64.3 years).  Maternal mortality ratio: The maternal mortality ratio for the 3 years preceding the survey is estimated at 186 maternal deaths per 100,000 live births.  Pregnancy-related mortality ratio: The pregnancy- related mortality ratio is 255 overall, 188 in Azad Jammu and Kashmir, and 202 in Gilgit Baltistan. dult and maternal mortality indicators can be used to assess the health status of a population. Mortality indicators are also used to estimate the life expectancy of the population and subsequently to assess the country’s level of development. The issue of reproductive health care, particularly health care during pregnancy, childbirth, and the postpartum period, has been of major concern to governments in most developing countries, and Pakistan is no exception. Maternal mortality represents one of the largest and most persistent gaps in health indicators between developed and developing countries. The maternal mortality ratio (MMR), which is the ratio of maternal deaths per 100,000 live births, is several times higher in some developing countries than in the developed countries of Northern Europe (Abou Zahr and Wardlaw 2004). The MMR is believed to be the most sensitive indicator of women’s health status in a society and of the quality and accessibility of maternal health services available to women. A maternal death is not merely a result of treatment failure; rather, it is the final outcome of a complex interplay among a myriad of social, cultural, and economic factors. Therefore, maternal mortality is widely recognised as a key human rights issue (Rosenfield et al. 2006). In the vast majority of cases, a maternal death reflects the failure of society to look after the life and health of its mothers. The Sustainable Development Goals (SDGs) include the MMR as a target of Goal 3 (ensuring healthy lives and promoting well-being for all at all ages), with an aim of reducing the global MMR to less than 70 maternal deaths per 100,000 live births by 2030. Many experts believe that it is possible to achieve this target in a majority of developing countries where the MMR is currently higher than 100 by increasing access to high-quality skilled birth attendance and emergency obstetric care (Campbell and Graham 2006). A 28 • Adult and Maternal Mortality 3.1 MORTALITY RATES The crude death rate (number of deaths per 1,000 population in a given year) is not considered an appropriate indicator of a country’s health status, as this rate is usually higher in developed and higher income countries due to their having a larger proportion of the elderly in the population. Age-specific mortality rates (ASMRs), on the other hand, provide a more precise picture of the health and mortality indicators of a country. ASMRs are also used to generate life tables, which estimate life expectancy at birth and in subsequent age groups. Table 3.1 presents overall age-specific mortality rates by sex for Pakistan (excluding Azad Jammu and Kashmir and Gilgit Baltistan). The first row reflects the mortality rate among infants (less than age 1), which is considerably higher among boys than girls. However, the mortality rate among children age 1-4 is slightly higher among females. Mortality rates are higher among males in the young adult and middle-age groups with the exception of the 35-39 and 60-64 groups, where female mortality is slightly higher than male mortality. In the older age groups (75-79 and 80 or above), female mortality substantially exceeds male mortality. After taking into consideration possible errors in age reporting, the distribution of age- specific mortality rates reflects lower mortality overall among females than males, resulting in higher female life expectancy at birth. The mortality rate in the 15-49 age bracket is almost 50% higher among males than females; the overall mortality rate for all ages (roughly corresponding to the crude death rate) is also higher among males. Correspondingly, the probability of death (q) in the 15-49 and 15-60 age groups is also higher among males than females. Figure 3.1 also depicts age-specific mortality rates among females and males. As expected, a high risk of death is observed in early childhood, dropping to a minimum at age 10-14 and then rising steadily into older ages. As a general rule, mortality rates start to increase rapidly beyond approximately age 40. In Pakistan, mortality rates increase rapidly after age 65. Male mortality rates are slightly higher than female mortality rates, and the most prominent differences are between the 15-19 and 55-54 age groups. The overall mortality rate is 22% higher among males than females (8.11 versus 6.63 per 1,000 population). Figure 3.1 Age-specific mortality rates in the 3 years preceding the survey by sex (log scale) 0.1 1.0 10.0 100.0 1,000.0 Age group (years) Mortality rates per 1,000 Female Male Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan Adult and Maternal Mortality • 29 Patterns by background characteristics  Age-specific mortality rates are generally higher in rural areas than urban areas among women and men less than age 45, but the reverse is true among those above age 45 (Figure 3.2). Figure 3.2 Age-specific mortality rates in the 3 years preceding the survey by residence (log scale)  Mortality rates among infants are higher in Punjab and Sindh than in Khyber Pakhtunkhwa. Mortality rates in Punjab and Sindh are similar across age groups, while rates in Khyber Pakhtunkhwa are generally lower than those in Punjab and Sindh among women and men age 45 and above (Table 3.2).  Mortality rates among infants and in most age groups above 45-49 are somewhat lower in Balochistan than in the other regions (Figure 3.3). Figure 3.3 Age-specific mortality rates in the 3 years preceding the survey by region (log scale)  Mortality rates are typically lower in Azad Jammu and Kashmir and Gilgit Baltistan than in other areas of Pakistan from age 35-39 to age 70-74. 0.1 1.0 10.0 100.0 1,000.0 Age group (years) Mortality rates per 1,000 Urban Rural Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan 0.1 1.0 10.0 100.0 1,000.0 Age group (years) Mortality rates per 1,000 Punjab Sindh Khyber Pakhtunkhwa Balochistan 30 • Adult and Maternal Mortality  The mortality rate among infants is higher in Gilgit Baltistan than in Azad Jammu and Kashmir. In most subsequent age groups, however, rates are lower in Gilgit Baltistan. 3.2 REPRODUCTIVE AGE MORTALITY Reproductive age mortality rate The number of adult deaths in the 15-49 age group per 1,000 population age 15-49 is calculated directly from death and census data collected through the short questionnaire administered to each household. Age-specific mortality rates are higher in rural areas than urban areas for all ages up to age 40-44 (Table 3.2). In all subsequent age groups, however, rural mortality rates are lower than urban rates. Further exploration of the data is needed to determine the reasons for this apparent anomaly. It could be due to reporting errors in the ages of deceased persons in rural areas. This phenomenon is also reflected in the mortality rate for the 15-49 age group and the overall rate for all ages (roughly corresponding to the crude death rate for the population), although the differences as a whole are not large. Both rates are highest in Punjab followed by Sindh, Khyber Pakhtunkhwa, and Balochistan. Two summary measures of adult mortality are also highlighted. The overall mortality rate for Pakistan is 7.37 per 1,000 persons, and the overall rate for the 15-49 age group is 2.07. Patterns by background characteristics  The mortality rate for the 15-49 age group and the overall mortality rate for all ages (roughly corresponding to the crude death rate for the population) are higher in Azad Jammu and Kashmir than Gilgit Baltistan (Figure 3.4). Figure 3.4 Age-specific mortality rates in the 3 years preceding the survey in Azad Jammu and Kashmir and Gilgit Baltistan (log scale)  The rapid increase in the mortality rate beyond age 50 reflects not only the health status of persons in Azad Jammu and Kashmir and Gilgit Baltistan but also the availability of health services in remote areas. Table 3.3.1 and Table 3.3.2 show female and male mortality rates by age, residence, and region and two summary measures of adult mortality. 0.1 1.0 10.0 100.0 1,000.0 Age group (years) Mortality rates per 1,000 Azad Jammu and Kashmir Gilgit Baltistan Adult and Maternal Mortality • 31 In Pakistan, mortality is higher among males than females in almost every age group. The mortality rate is higher among rural females than urban females up to age 49, but this pattern is reversed at age 50 and above (Figure 3.5). Rural males below age 30 have a higher probability of dying than their urban counterparts, but the reverse is true at age 30 and above (Table 3.3.2 and Figure 3.6). Mortality is higher among males than females in all regions, with the largest difference in Azad Jammu and Kashmir (Figure 3.7). Figure 3.5 Age-specific female mortality rates in the 3 years preceding the survey by residence (log scale) Figure 3.6 Age-specific male mortality rates in the 3 years preceding the survey by residence (log scale) 0.1 1.0 10.0 100.0 1,000.0 Age group (years) Mortality rates per 1,000 Urban Rural Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan 0.1 1.0 10.0 100.0 1,000.0 Age group (years) Mortality rates per 1,000 Urban Rural Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan 32 • Adult and Maternal Mortality Figure 3.7 Crude mortality rates in the 3 years preceding the survey by sex and region Patterns by background characteristics  Mortality rates in Punjab and Sindh are similar across almost all age groups, while rates in Khyber Pakhtunkhwa are generally lower than in Punjab and Sindh at age 45 and above.  Balochistan has lower mortality than most regions at age 40 and above. Mortality rates in each age group in Azad Jammu and Kashmir and Gilgit Baltistan are generally lower than in the other regions.  Mortality rates for females and males are consistently lower in Gilgit Baltistan than in Azad Jammu and Kashmir with the exception of the younger age groups. 3.3 TRENDS IN ADULT MORTALITY Table 3.4 shows deaths, person-years of exposure, and mortality rates by 5-year age groups, residence, and region for the 3 years preceding the survey. In Pakistan, mortality in the 15-49 age group is 44% higher among males than females. The overall mortality rate for females age 15-49 is 1.72 per 1,000 persons, and the overall rate for males age 15-49 is 2.48 per 1,000 persons. Male mortality rates are much higher than female mortality rates in every age group other than the 35-39 group, and the most pronounced difference is in the 15-19 age group, in which the male mortality rate is 90% higher than the female rate. Mortality rates among women increase from 0.77 in the 15-19 age group to 4.83 in the 45-49 age group; among men, mortality rates increase from 1.46 in the 15-19 age group to 7.41 in the 45-49 age group (Figure 3.8). The higher mortality among men could be attributed to some extent to men being more involved in activities outside of the house and being exposed to more risks. 7.14 6.46 6.13 4.52 6.63 6.87 4.98 9.13 7.33 7.28 5.11 8.11 9.42 6.48 Punjab Sindh Khyber Pakhtunkhwa Balochistan Pakistan Azad Jammu and Kashmir Gilgit Baltistan Mortality rates per 1,000 Female Male Note: Pakistan total excludes Azad Jammu and Kashmir and Gilgit Baltistan Adult and Maternal Mortality • 33 Figure 3.8 All-cause adult mortality rates in the 3 years preceding the survey by sex and age Patterns by background characteristics  With regard to patterns by residence, mortality rates are higher in rural areas than in urban areas. Rural females have higher mortality rates than urban females (1.89 versus 1.45) (Figure 3.9).  Similarly, rural males have a higher probability of dying than their urban counterparts (2.51 versus 2.43). 0.77 1.10 1.12 1.51 2.45 3.03 4.83 1.46 1.39 1.85 2.09 2.00 4.91 7.41 15-19 20-24 25-29 30-34 35-39 40-44 45-49 Age group (years) Mortality rates per 1,000 Female Male Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan Figure 3.9 All-cause adult mortality rates (15-49 years) in the 3 years preceding the survey by sex and residence 1.45 2.43 1.89 2.51 Female Male Mortality rates per 1,000 Urban Rural Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan 34 • Adult and Maternal Mortality  In Punjab, Sindh, and Khyber Pakhtunkhwa, men age 15-49 have higher mortality rates than women age 15-49, while in Balochistan mortality among women and men is nearly identical (Figure 3.10). Figure 3.10 All-cause adult mortality rates (15-49 years) in the 3 years preceding the survey by sex and region  Mortality rates for males and females are higher in Azad Jammu and Kashmir than in Gilgit Baltistan. 3.4 LIFE EXPECTANCY Table 3.5 presents Pakistan’s life table for both sexes combined (excluding Azad Jammu and Kashmir and Gilgit Baltistan). Life expectancy at birth is 65.4 years (i.e., a newborn in Pakistan in 2019 can expect to reach age 65.4 if current age-specific mortality rates remain constant). A newborn girl is expected to live approximately 2 years longer (66.5 years) than a newborn boy (64.3 years) (Figure 3.11). The female life expectancy advantage in Pakistan is lower than the average female advantage of 5 years worldwide. Life expectancy at birth for both females and males has increased relative to 1998 census figures (63.0 years for females and 62.5 years for males) (NIPS and ICF International 2013). Table 3.6 presents separate life tables for males and females, describing life expectancy at various ages. In all age groups up to age 50-54, life expectancy among women is about 1-2 years higher than life expectancy among men. However, from age 55-59 onward there is almost no difference in average years of remaining life. This might be due to the fact that women in older age groups are no longer protected by female hormones and share the same risk of death as men in terms of factors such as sedentary lifestyles, unhealthy dietary patterns, and co-morbidities such as diabetes and hypertension. 1.81 1.79 1.45 1.52 1.72 1.79 1.41 2.64 2.39 2.43 1.50 2.48 3.06 2.19 Punjab Sindh Khyber Pakhtunkhwa Balochistan Pakistan Azad Jammu and Kashmir Gilgit Baltistan Mortality rates per 1,000 Female Male Note: Pakistan total excludes Azad Jammu and Kashmir and Gilgit Baltistan Figure 3.11 Life expectancy according to sex 65.4 66.5 64.3 Both sexes Female Male Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan Adult and Maternal Mortality • 35 The number of females and males, from the beginning of the birth cohort as 100,000, starts to drop in the first two age intervals but remains fairly steady till age 45. From age 55-59 onward, the number of people alive at the start of each interval starts dropping more rapidly, with sharper decreases for males than females (Figure 3.12). Figure 3.12 Number of females and males living at the beginning of each age interval across the life span 3.5 ESTIMATES OF PREGNANCY-RELATED AND MATERNAL MORTALITY Difference between maternal mortality rate and maternal mortality ratio: While the numerator is the same in both indicators (maternal deaths), the maternal mortality ratio (MMR) uses live births as the denominator and the maternal mortality rate uses the person-years lived by women of reproductive age (15-49 years) during the 3-year recall period for which maternal deaths were recorded. Maternal mortality ratio: The number of maternal deaths per 100,000 live births. The maternal mortality ratio is calculated by dividing the age- standardised maternal mortality rate for women age 15-49 in the 3 years preceding the survey by the general fertility rate (GFR) for the same time period. The MMR is calculated using two different methods: (1) by using the number of live births reported in the entire sample (referred to as the “direct method” in this report) and (2) by using an estimated number of live births from the GFR (referred to as the “indirect method” in this report). Deaths among women of reproductive age (15-49 years) in the preceding 3 years were recorded via the Short Household Questionnaire. All households reporting a female death during that period were revisited to conduct a verbal autopsy to determine the cause of death. Completed Verbal Autopsy Questionnaires were then reviewed by panels of experts (senior obstetricians and general physicians) to determine causes of death as per the International Classification of Diseases (ICD-10) codes and to ascertain whether deaths met the ICD-10 definition of maternal deaths. The number of maternal deaths identified through this process was used to estimate the MMR, where the denominator was derived from the live births reported from the sample households during the 3 years preceding the survey. The MMR thus calculated is referred to as the “direct MMR” in this chapter. The number of live births used in the denominator to estimate the 0 20,000 40,000 60,000 80,000 100,000 Age group (years) Female Male Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan 36 • Adult and Maternal Mortality MMR was also derived in another manner (“indirectly”) by calculating the general fertility rate from women’s pregnancy histories (which were recorded in the 10% subsample households through the Woman’s Questionnaires), which was then used to estimate the number of live births. Maternal deaths are a subset of all female deaths. They are defined as any death that occurred during pregnancy or childbirth or within 42 days after the birth or termination of a pregnancy. Maternal deaths do not include deaths due to accidents or violence (ICD-10 definition of maternal deaths). In the 2019 PMMS, the MMR was estimated directly from maternal deaths (identified through verbal autopsy interviews of the next of kin of deceased women) and live births reported in the household survey. A “pregnancy-related death” is defined as the death of a woman age 15-49 during pregnancy or childbirth/abortion or within 42 days of the termination of a pregnancy, regardless of the cause of death. Pregnancy-related deaths, therefore, may include incidental or accidental deaths. Table 3.7 presents pregnancy-related mortality rates and ratios by age group, residence, and region. The pregnancy-related mortality rate for Pakistan is 0.31 per 1,000 person-years lived by women of reproductive age (15-49 years) during the last 3 years. This rate is lower in Azad Jammu and Kashmir (0.20) and Gilgit Baltistan (0.28). The age-adjusted pregnancy-related mortality ratio (PRMR) is 251 per 100,000 live births, with rates of 179 and 196 in Azad Jammu and Kashmir and Gilgit Baltistan, respectively. There are substantial differences in both indicators by region and residence. Note that in the table, the PRMR is calculated using an estimate of live births derived from the general fertility rate, which in turn is estimated from the pregnancy histories recorded in the Woman’s Questionnaire (administered in a subsample). This method is used here in an effort to ensure that the current rates and ratios are comparable to those reported in the 2006-07 PDHS (NIPS and Macro International 2008). Table 3.8 presents pregnancy-related mortality ratios calculated directly by using live births (reported in the entire sample at the household level) in the denominator. The pregnancy-related mortality ratio is 255 overall, 188 in Azad Jammu and Kashmir, and 202 in Gilgit Baltistan. There are sizable differences in pregnancy-related mortality ratios by urban/rural residence and region. Patterns by background characteristics  Age-specific pregnancy-related mortality ratios show the expected pattern of being low in the younger age groups, increasing in the early reproductive years and reaching a peak in the 40-44 age group, and then decreasing at age 45-49 as pregnancy and childbirth taper off. Adult and Maternal Mortality • 37  The PRMR is notably higher in the 15-19 age group than in 20-24 age group (Figure 3.13). Figure 3.13 Age-specific pregnancy-related mortality ratio trends, 2006-07 PDHS and 2019 PMMS  The probability of pregnancy decreases substantially at older ages. Pregnancies at the older reproductive ages are riskier, resulting in higher mortality rates among women who become pregnant at older ages (Figure 3.13).  By region, the pregnancy-related mortality ratio is lowest in Khyber Pakhtunkhwa (175) and highest in Balochistan (358) (Figure 3.14). Figure 3.14 Pregnancy-related mortality ratios by region Table 3.9 shows direct estimates of maternal mortality rates and ratios for the 3 years preceding the survey by 5-year age groups, residence, and region. The MMR is 186 in Pakistan, 104 in Azad Jammu and Kashmir, and 157 in Gilgit Baltistan. It is almost twice as high in Balochistan (298) as in Punjab (157). The MMR is 26% higher in rural areas than in urban areas. 259 209 262 297 748 967 361 249 131 142 325 644 1,051 331 15-19 20-24 25-29 30-34 35-39 40-44 45-49 Age group (years) Per 100,000 live births 2006-07 PDHS 2019 PMMS Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan 219 345 175 358 251 179 196 Punjab Sindh Khyber Pakhtunkhwa Balochistan Pakistan Azad Jammu and Kashmir Gilgit Baltistan Per 100,000 live births Note: Pakistan total excludes Azad Jammu and Kashmir and Gilgit Baltistan 38 • Adult and Maternal Mortality Figure 3.15 compares the MMR in 2019 with the MMR in 2006-07. The MMR in 2019 is highest at age 35-39 (481) and lowest at age 20-24 (99). The maternal mortality ratio decreased substantially in five of the seven age groups from 2006-07 to 2019. There was a slight increase between the two surveys at age 30-34 and a more substantial increase in the oldest age group (45-49). In general, there was an overall decrease in the MMR between the 2006-07 PDHS and the 2019 PMMS, from 276 maternal deaths per 100,000 live births to 186 (for the 3 years preceding the survey), showing a one-third decline1. Figure 3.15 Age-specific maternal mortality ratio trends, 2006-07 PDHS and 2019 PMMS Table 3.10 shows the total fertility rate, general fertility rate, maternal mortality ratio (with upper and lower confidence interval bounds), and lifetime risk of maternal death for the 3 years preceding the survey by urban-rural residence and region. Maternal mortality ratios (with 95% confidence intervals) are also shown in Figure 3.16. The MMR shown in the table is computed indirectly by using live births derived from the general fertility rate (as described above). The MMR for Pakistan overall is estimated at 186 (95% confidence interval: 138-234), which is higher than in Azad Jammu and Kashmir and Gilgit Baltistan. There are substantial differences in the MMR by region and urban/rural residence. The lifetime risk of maternal mortality is 0.007, which means that 1 in every 143 women in Pakistan will die due to complications during pregnancy, childbirth/abortion, or the postpartum period. 1This evidence of a decline does not take account of statistical uncertainty in the estimates from the two surveys. 242 210 267 246 657 855 234 194 99 115 263 481 286 331 15-19 20-24 25-29 30-34 35-39 40-44 45-49 Age group (years) Per 100,000 live births PDHS 2006-07 PMMS 2019 Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan Adult and Maternal Mortality • 39 Figure 3.16 Maternal mortality ratios by region Patterns by background characteristics  Urban-rural MMR estimates show a difference of 41 deaths per 100,000 live births, with a higher estimate in rural areas (199) than in urban areas (158).  The 95% confidence intervals for all regions indicate that the MMR differences between regions are not statistically significant (Figure 3.16).  The MMR is lowest in Punjab (157 per 100,000 live births), followed by Khyber Pakhtunkhwa (165 per 100,000 live births), Sindh (224 per 100,000 live births), and Balochistan (298 per 100,000 live births). Table 3.11 shows maternal mortality ratios using the direct method (maternal deaths divided by live births from household birth records, as reported in the entire sample). The estimated MMR is 189 maternal deaths per 100,000 live births. The MMR is higher in rural areas (203 per 100,000 live births) than in urban areas (159 per 100,000 live births). Azad Jammu and Kashmir has a lower MMR (108 per 100,000 live births) than Gilgit Baltistan (162 per 100,000 live births). There are substantial regional MMR variations, ranging from 161 per 100,000 live births in Khyber Pakhtunkhwa to 317 per 100,000 live births in Balochistan. LIST OF TABLES For more information on adult and maternal mortality, see the following tables:  Table 3.1 Mortality rates by sex  Table 3.2 Age-specific mortality rates by residence and region  Table 3.3.1 Mortality rates by residence and region: Females  Table 3.3.2 Mortality rates by residence and region: Males  Table 3.4 Adult mortality rates (15-49 years)  Table 3.5 Complete life table for Pakistan  Table 3.6 Complete life table for the total population of Pakistan by sex  Table 3.7 Pregnancy-related mortality 235 299 246 466 234 185 261 79 148 84 130 138 23 53 157 224 165 298 186 104 157 Punjab Sindh Khyber Pakhtunkhwa Balochistan Pakistan Azad Jammu and Kashmir Gilgit Baltistan Maternal deaths per 100,000 live births Note: Pakistan total excludes Azad Jammu and Kashmir and Gilgit Baltistan 40 • Adult and Maternal Mortality  Table 3.8 Pregnancy-related mortality ratio (PRMR) using live births as the denominator (pregnancy-related deaths divided by live births reported in the household survey)  Table 3.9 Maternal mortality  Table 3.10 Maternal mortality ratio  Table 3.11 Maternal mortality ratio using direct method Table 3.1 Mortality rates by sex Direct estimates of mortality rates (per 1,000 persons) from the household listing of usual members who died in the 3 years preceding the survey, according to sex, Pakistan MMS 2019 Females Males Age group Deaths Exposure years Mortality rate Deaths Exposure years Mortality rate <1 1,855 31,442 59.00 2,233 33,248 67.17 1-4 343 116,917 2.94 279 125,440 2.22 5-9 88 139,430 0.63 145 148,800 0.97 10-14 67 123,614 0.54 100 128,370 0.78 15-19 90 117,365 0.77 169 115,300 1.46 20-24 110 100,449 1.10 124 89,522 1.39 25-29 101 90,591 1.12 146 79,032 1.85 30-34 103 68,283 1.51 134 64,327 2.09 35-39 150 61,286 2.45 115 57,463 2.00 40-44 136 44,828 3.03 221 44,956 4.91 45-49 200 41,395 4.83 303 40,846 7.41 50-54 255 32,958 7.74 423 33,427 12.67 55-59 381 26,629 14.31 503 28,756 17.51 60-64 535 19,538 27.40 639 23,469 27.22 65-69 472 14,078 33.52 605 17,415 34.76 70-74 509 8,248 61.66 737 10,793 68.26 75-79 347 5,300 65.44 399 6,341 62.86 80+ 1,215 6,891 176.38 1,279 7,760 164.86 Total age 15-49 890 524,197 1.70 1,212 491,445 2.47 Total all ages 6,959 1,049,243 6.63 8,554 1,055,263 8.11 Probability of dying 35q151 71 100 45q152 168 226 Note: Table excludes Azad Jammu and Kashmir and Gilgit Baltistan. Deaths are from the household listing of usual members who died in the 3 years preceding the survey (excluding the month of the interview); exposure is from usual members of the household and applicable exposure of members who died; deaths with missing age at death have been redistributed proportionately; and cases with missing age in the household schedule (assumed exposure) have been redistributed. 1 The probability of dying between exact ages 15 and 50, expressed per 1,000 persons at age 15 2 The probability of dying between exact ages 15 and 60, expressed per 1,000 persons at age 15 Adult and Maternal Mortality • 41 Table 3.2 Age-specific mortality rates by residence and region Direct estimates of mortality rates (per 1,000 persons) from the household listing of usual members who died in the 3 years preceding the survey, according to residence and region, Pakistan MMS 2019 Residence Region Total3 Azad Jammu and Kashmir Gilgit Baltistan Age group Urban Rural Punjab1 Sindh Khyber Pakhtun- khwa2 Balochistan <1 56.34 66.39 66.62 68.14 55.23 47.59 63.20 47.20 54.97 1-4 1.91 2.87 2.69 2.51 2.10 3.34 2.57 2.11 2.07 5-9 0.56 0.93 0.81 0.86 0.81 0.65 0.81 0.47 1.12 10-14 0.52 0.74 0.57 0.87 0.68 0.62 0.66 0.26 0.32 15-19 1.04 1.16 0.98 1.25 1.34 0.90 1.11 1.15 1.39 20-24 1.08 1.34 1.35 1.19 0.98 1.19 1.23 2.05 1.34 25-29 1.14 1.67 1.37 1.37 1.98 1.06 1.46 2.00 1.58 30-34 1.74 1.82 1.87 1.53 1.91 1.69 1.79 1.84 1.78 35-39 1.66 2.61 2.28 2.41 2.06 1.62 2.23 2.10 1.90 40-44 3.64 4.20 4.18 3.82 3.92 2.55 3.97 3.12 2.07 45-49 6.44 5.89 6.97 6.53 3.53 3.70 6.11 6.72 3.76 50-54 12.37 8.83 11.42 11.42 5.98 5.65 10.22 7.11 4.90 55-59 17.84 14.84 17.11 16.93 12.85 10.99 15.97 16.22 6.00 60-64 32.33 24.36 28.76 29.10 22.13 22.98 27.30 24.60 13.38 65-69 41.15 30.63 36.08 33.77 33.19 18.67 34.21 32.07 17.73 70-74 72.33 61.68 66.19 69.75 59.53 60.72 65.40 57.88 35.51 75-79 68.66 61.66 63.72 66.45 67.80 41.23 64.03 47.71 23.48 80+ 181.92 164.79 162.58 184.36 192.78 141.22 170.28 148.86 118.11 Total age 15-49 1.92 2.17 2.18 2.08 1.91 1.50 2.07 2.33 1.77 Total all ages 7.06 7.55 8.13 6.91 6.70 4.82 7.37 8.07 5.72 Probability of dying 35q154 80 89 91 87 76 62 86 91 67 45q155 210 191 212 208 159 137 198 191 116 Note: Deaths are from the household listing of usual members who died in the 3 years before the survey (excluding the month of the interview); exposure is from usual members of the household and applicable exposure of members who died; deaths with missing age at death have been redistributed proportionately; and cases with missing age in the household schedule (assumed exposure) have been redistributed. 1 Punjab includes ICT. 2 Khyber Pakhtunkhwa includes the merged districts of former FATA. 3 Total excludes Azad Jammu and Kashmir and Gilgit Baltistan. 4 The probability of dying between exact ages 15 and 50, expressed per 1,000 persons at age 15 5 The probability of dying between exact ages 15 and 60, expressed per 1,000 persons at age 15 42 • Adult and Maternal Mortality Table 3.3.1 Mortality rates by residence and region: Females Direct estimates of mortality rates (per 1,000 persons) from the household listing of usual members who died in the 3 years preceding the survey, according to residence and region, Pakistan MMS 2019 Residence Region Total3 Azad Jammu and Kashmir Gilgit Baltistan Age group Urban Rural Punjab1 Sindh Khyber Pakhtun- khwa2 Balochistan <1 47.34 64.39 60.54 66.35 52.65 45.01 59.00 44.61 51.95 1-4 2.21 3.28 3.19 2.99 2.28 3.12 2.94 1.97 2.21 5-9 0.52 0.69 0.72 0.59 0.56 0.39 0.63 0.46 1.05 10-14 0.31 0.67 0.45 0.76 0.65 0.16 0.54 0.24 0.42 15-19 0.72 0.79 0.63 1.11 0.73 0.78 0.77 0.81 0.72 20-24 0.94 1.20 1.18 1.21 0.73 1.15 1.10 1.26 0.94 25-29 0.75 1.34 1.19 1.01 1.30 0.29 1.12 1.48 1.35 30-34 1.21 1.69 1.56 1.26 1.49 2.04 1.51 1.62 1.21 35-39 2.11 2.67 2.32 3.27 1.68 2.83 2.45 2.09 2.02 40-44 2.14 3.63 2.99 3.02 3.38 2.29 3.03 2.58 1.74 45-49 4.43 5.08 5.76 3.89 3.41 3.91 4.83 4.73 3.99 50-54 11.35 5.48 7.72 10.52 5.22 5.40 7.74 4.28 3.36 55-59 15.40 13.68 15.72 13.30 12.53 10.16 14.31 14.31 7.74 60-64 31.82 24.89 27.02 29.89 25.41 29.81 27.40 22.57 12.92 65-69 44.61 28.19 34.55 33.61 34.44 16.91 33.52 29.82 20.37 70-74 69.80 57.30 60.11 72.32 55.40 67.24 61.66 64.41 38.87 75-79 64.40 66.00 64.15 63.51 78.29 26.22 65.44 49.80 25.11 80+ 197.11 166.56 168.53 191.44 193.85 162.77 176.38 144.86 103.26 Total age 15-49 1.42 1.87 1.78 1.78 1.44 1.50 1.70 1.77 1.40 Total all ages 6.25 6.85 7.14 6.46 6.13 4.52 6.63 6.87 4.98 Probability of dying 35q154 60 79 75 71 62 64 71 70 58 45q155 178 163 178 176 141 134 168 153 109 Note: Deaths are from the household listing of usual members who died in the 3 years before the survey (excluding the month of the interview); exposure is from usual members of the household and applicable exposure of members who died; deaths with missing age at death have been redistributed proportionately; and cases with missing age in the household schedule (assumed exposure) have been redistributed. 1 Punjab includes ICT. 12 Khyber Pakhtunkhwa includes the merged districts of former FATA. 3 Total excludes Azad Jammu and Kashmir and Gilgit Baltistan. 4 The probability of dying between exact ages 15 and 50, expressed per 1,000 persons at age 15 5 The probability of dying between exact ages 15 and 60, expressed per 1,000 persons at age 15 Adult and Maternal Mortality • 43 Table 3.3.2 Mortality rates by residence and region: Males Direct estimates of mortality rates (per 1,000 persons) from the household listing of usual members who died in the 3 years preceding the survey, according to residence and region, Pakistan MMS 2019 Residence Region Total3 Azad Jammu and Kashmir Gilgit Baltistan Age group Urban Rural Punjab1 Sindh Khyber Pakhtun- khwa2 Balochistan <1 64.79 68.29 72.35 69.80 57.77 50.00 67.17 49.63 57.71 1-4 1.63 2.49 2.22 2.08 1.92 3.54 2.22 2.24 1.94 5-9 0.59 1.16 0.88 1.12 1.05 0.88 0.97 0.49 1.20 10-14 0.72 0.81 0.68 0.98 0.71 1.04 0.78 0.27 0.22 15-19 1.34 1.54 1.36 1.39 1.97 1.01 1.46 1.53 2.14 20-24 1.22 1.50 1.54 1.17 1.28 1.23 1.39 3.15 1.85 25-29 1.55 2.07 1.58 1.73 2.82 1.87 1.85 2.79 1.87 30-34 2.26 1.97 2.20 1.79 2.40 1.33 2.09 2.17 2.44 35-39 1.21 2.55 2.23 1.55 2.51 0.45 2.00 2.11 1.78 40-44 5.07 4.79 5.39 4.54 4.49 2.81 4.91 3.87 2.42 45-49 8.34 6.74 8.21 9.08 3.67 3.49 7.41 9.24 3.53 50-54 13.35 12.21 15.07 12.26 6.79 5.89 12.67 10.34 6.65 55-59 20.01 15.95 18.39 20.34 13.14 11.71 17.51 18.21 4.31 60-64 32.74 23.91 30.25 28.48 19.40 17.41 27.22 26.37 13.79 65-69 38.57 32.69 37.35 33.90 32.19 19.98 34.76 33.98 15.45 70-74 74.26 65.04 70.92 67.84 62.60 55.92 68.26 52.65 33.22 75-79 72.51 58.17 63.37 68.92 58.52 51.82 62.86 45.96 22.26 80+ 168.35 163.21 157.44 177.22 191.82 124.74 164.86 152.42 128.84 Total age 15-49 2.41 2.51 2.62 2.38 2.44 1.50 2.47 3.08 2.19 Total all ages 7.84 8.26 9.13 7.33 7.28 5.11 8.11 9.42 6.48 Probability of dying 35q154 100 100 106 101 91 59 100 117 77 45q155 238 219 244 236 178 138 226 235 126 Note: Deaths are from the household listing of usual members who died in the 3 years before the survey (excluding the month of the interview); exposure is from usual members of the household and applicable exposure of members who died; deaths with missing age at death have been redistributed proportionately; and cases with missing age in the household schedule (assumed exposure) have been redistributed. 1 Punjab includes ICT. 2 Khyber Pakhtunkhwa includes the merged districts of former FATA. 3 Total excludes Azad Jammu and Kashmir and Gilgit Baltistan. 4 The probability of dying between exact ages 15 and 50, expressed per 1,000 persons at age 15 5 The probability of dying between exact ages 15 and 60, expressed per 1,000 persons at age 15 44 • Adult and Maternal Mortality Table 3.4 Adult mortality rates (15-49 years) Direct estimates of female and male mortality rates for the 3 years preceding the survey, by 5-year age groups, residence, and region, Pakistan MMS 2019 Background characteristic Deaths Exposure years Mortality rate1 FEMALE Age 15-19 90 117,365 0.77 20-24 110 100,449 1.10 25-29 101 90,591 1.12 30-34 103 68,283 1.51 35-39 150 61,286 2.45 40-44 136 44,828 3.03 45-49 200 41,395 4.83 Residence Urban 284 199,897 1.45 Rural 606 324,300 1.89 Region Punjab2 496 278,770 1.81 Sindh 208 117,149 1.79 Khyber Pakhtunkhwa3 143 99,292 1.45 Balochistan 43 28,987 1.52 Total 15-494 890 524,197 1.72a Azad Jammu and Kashmir 143 81,048 1.79 Gilgit Baltistan 79 56,225 1.41 MALE Age 15-19 169 115,300 1.46 20-24 124 89,522 1.39 25-29 146 79,032 1.85 30-34 134 64,327 2.09 35-39 115 57,463 2.00 40-44 221 44,956 4.91 45-49 303 40,846 7.41 Residence Urban 483 200,512 2.43 Rural 729 290,933 2.51 Region Punjab2 670 255,548 2.64 Sindh 286 120,069 2.39 Khyber Pakhtunkhwa3 213 87,431 2.43 Balochistan 43 28,397 1.50 Total 15-494 1,212 491,445 2.48a Azad Jammu and Kashmir 185 60,144 3.06 Gilgit Baltistan 108 49,419 2.19 1 Expressed per 1,000 population 2 Punjab includes ICT. 3 Khyber Pakhtunkhwa includes the merged districts of former FATA. 4 Total excludes Azad Jammu and Kashmir and Gilgit Baltistan. a Age-adjusted rate Adult and Maternal Mortality • 45 Table 3.5 Complete life table for Pakistan Complete life table for the total population, Pakistan MMS 2019 Age group Proportion of persons alive at the beginning of the age interval who died during the interval (qx) Number living at the beginning of the age interval (lx) Number dying during the age interval (dx) Stationary population in the age interval (Lx) Stationary population in this and all subsequent age intervals (Tx) Average number of years of life remaining at the beginning of the age interval (ex) <1 0.0612 100,000 6,125 96,905 6,536,651 65.37 1-4 0.0102 93,875 959 373,580 6,439,746 68.60 5-9 0.0040 92,916 375 463,644 6,066,166 65.29 10-14 0.0033 92,542 307 461,941 5,602,522 60.54 15-19 0.0055 92,235 511 459,896 5,140,581 55.73 20-24 0.0061 91,724 564 457,207 4,680,686 51.03 25-29 0.0073 91,160 662 454,140 4,223,479 46.33 30-34 0.0089 90,497 806 450,469 3,769,339 41.65 35-39 0.0111 89,691 996 445,962 3,318,870 37.00 40-44 0.0197 88,695 1,744 439,101 2,872,909 32.39 45-49 0.0301 86,951 2,616 428,181 2,433,808 27.99 50-54 0.0498 84,335 4,202 411,080 2,005,627 23.78 55-59 0.0767 80,133 6,150 385,085 1,594,547 19.90 60-64 0.1276 73,983 9,439 345,780 1,209,463 16.35 65-69 0.1572 64,544 10,147 296,629 863,682 13.38 70-74 0.2789 54,397 15,173 231,988 567,053 10.42 75-79 0.2740 39,224 10,746 167,822 335,065 8.54 80+ 1.0000 28,477 28,477 167,243 167,243 5.87 Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan. 46 • Adult and Maternal Mortality Table 3.6 Complete life table for the total population of Pakistan by sex Complete life table for the total population by sex, Pakistan MMS 2019 Age group Proportion of persons alive at the beginning of the age interval who died during the interval (qx) Number living at the beginning of the age interval (lx) Number dying during the age interval (dx) Stationary population in the age interval (Lx) Stationary population in this and all subsequent age intervals (Tx) Average number of years of life remaining at the beginning of the age interval (ex) FEMALE <1 0.0573 100,000 5,730 97,107 6,647,938 66.48 1-4 0.0117 94,270 1,101 374,874 6,550,831 69.49 5-9 0.0032 93,169 294 465,110 6,175,957 66.29 10-14 0.0027 92,875 251 463,747 5,710,847 61.49 15-19 0.0038 92,624 354 462,232 5,247,100 56.65 20-24 0.0055 92,269 505 460,084 4,784,868 51.86 25-29 0.0056 91,765 511 457,543 4,324,784 47.13 30-34 0.0075 91,253 685 454,551 3,867,240 42.38 35-39 0.0122 90,568 1,103 450,076 3,412,689 37.68 40-44 0.0150 89,465 1,346 443,951 2,962,613 33.11 45-49 0.0239 88,119 2,103 435,317 2,518,662 28.58 50-54 0.0380 86,016 3,267 421,862 2,083,345 24.22 55-59 0.0691 82,750 5,715 399,291 1,661,483 20.08 60-64 0.1280 77,035 9,861 359,959 1,262,192 16.38 65-69 0.1543 67,174 10,365 309,231 902,233 13.43 70-74 0.2653 56,808 15,072 244,428 593,002 10.44 75-79 0.2791 41,736 11,647 177,980 348,574 8.35 80+ 1.0000 30,090 30,090 170,595 170,595 5.67 MALE <1 0.0650 100,000 6,497 96,715 6,431,015 64.31 1-4 0.0088 93,503 827 372,356 6,334,300 67.74 5-9 0.0049 92,676 450 462,254 5,961,944 64.33 10-14 0.0039 92,226 359 460,232 5,499,690 59.63 15-19 0.0073 91,867 670 457,658 5,039,458 54.86 20-24 0.0069 91,197 630 454,408 4,581,800 50.24 25-29 0.0092 90,567 834 450,747 4,127,392 45.57 30-34 0.0104 89,733 932 446,333 3,676,645 40.97 35-39 0.0100 88,802 885 441,792 3,230,312 36.38 40-44 0.0243 87,917 2,132 434,232 2,788,519 31.72 45-49 0.0364 85,785 3,119 421,079 2,354,287 27.44 50-54 0.0614 82,666 5,073 400,514 1,933,209 23.39 55-59 0.0838 77,593 6,503 371,471 1,532,695 19.75 60-64 0.1272 71,090 9,044 332,327 1,161,223 16.33 65-69 0.1596 62,046 9,900 284,763 828,896 13.36 70-74 0.2892 52,146 15,079 220,893 544,133 10.43 75-79 0.2697 37,067 9,997 159,036 323,240 8.72 80+ 1.0000 27,070 27,070 164,204 164,204 6.07 Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan. Adult and Maternal Mortality • 47 Table 3.7 Pregnancy-related mortality Direct estimates of pregnancy-related mortality for the 3 years preceding the survey, by 5-year age groups, residence, and region, Pakistan MMS 2019 Background characteristic Percentage of female deaths that are pregnancy- related Number of pregnancy- related deaths1 Weighted number of woman-years2 Pregnancy- related mortality rate3 Pregnancy- related mortality ratio4 Age 15-19 16.7 15 117,365 0.13 249 20-24 23.1 25 100,449 0.25 131 25-29 28.7 29 90,591 0.32 142 30-34 36.0 37 68,283 0.54 325 35-39 27.5 41 61,286 0.67 644 40-44 9.8 13 44,828 0.30 1,051 45-49 0.6 1 41,395 0.03 331 Residence Urban 15.7 45 199,897 0.22 218 Rural 19.4 118 324,300 0.37 267 Region Punjab5 14.7 73 278,770 0.26 219 Sindh 24.1 50 117,149 0.43 345 Khyber Pakhtunkhwa6 16.7 24 99,292 0.24 175 Balochistan 35.4 15 28,987 0.54 358 Total 15-497 18.2 162 524,197 0.31a 251a Azad Jammu and Kashmir 11.1 16 81,048 0.20 179 Gilgit Baltistan 19.8 16 56,225 0.28 196 1 A pregnancy-related death is defined as the death of a woman while pregnant, during childbirth, or within 42 days after delivery. 2 Woman-years lived in that age group during the 36 months before the survey. For example, for the 15- 19 age group, this is calculated by taking ½ of the number of women age 15, plus 1½ times the number age 16, plus 2½ times the number age 17, plus 3 times the number age 18, plus 3 times the number age 19, plus 2½ times the number age 20, plus 1½ times the number age 21, plus ½ times the number age 22, plus 1½ times the number of deaths among women age 15-49 in the previous 36 months. 3 Expressed per 1,000 woman-years of exposure 4 Expressed per 100,000 live births; calculated as the age-adjusted pregnancy-related mortality rate times 100 divided by the age-adjusted general fertility rate 5 Punjab includes ICT. 6 Khyber Pakhtunkhwa includes the merged districts of former FATA. 7 Total excludes Azad Jammu and Kashmir and Gilgit Baltistan. a Age-adjusted rate 48 • Adult and Maternal Mortality Table 3.8 Pregnancy-related mortality ratio (PRMR) using live births as the denominator (pregnancy-related deaths divided by live births reported in the household survey) Pregnancy-related mortality ratio for the 3 years preceding the survey, by residence and region, Pakistan MMS 2019 Background characteristic Pregnancy- related deaths1 Live births Pregnancy- related mortality ratio2 Residence Urban 45 20,333 220 Rural 118 43,290 272 Region Punjab3 73 31,753 230 Sindh 50 13,786 364 Khyber Pakhtunkhwa4 24 14,075 170 Balochistan 15 4,010 383 Total5 162 63,623 255 Azad Jammu and Kashmir 16 8,501 188 Gilgit Baltistan 16 7,712 202 1 A pregnancy-related death is defined as the death of a woman while pregnant, during childbirth, or within 42 days after delivery, regardless of the cause of death. 2 Expressed per 100,000 live births 3 Punjab includes ICT. 4 Khyber Pakhtunkhwa includes the merged districts of former FATA. 5 Total excludes Azad Jammu and Kashmir and Gilgit Baltistan. Table 3.9 Maternal mortality Direct estimates of maternal mortality rates and ratios for the 3 years preceding the survey, by 5-year age groups, residence, and region, Pakistan MMS 2019 Background characteristic Percentage of female deaths that are maternal Number of maternal deaths1 Weighted number of woman-years2 Maternal mortality rate3 Maternal mortality ratio4 Age 15-19 13.0 12 117,365 0.10 194 20-24 17.4 19 100,449 0.19 99 25-29 23.4 24 90,591 0.26 115 30-34 29.1 30 68,283 0.44 263 35-39 20.5 31 61,286 0.50 481 40-44 2.7 4 44,828 0.08 286 45-49 0.6 1 41,395 0.03 331 Residence Urban 11.4 32 199,897 0.16 158 Rural 14.5 88 324,300 0.27 199 Region Punjab5 10.5 52 278,770 0.19 157 Sindh 15.7 33 117,149 0.28 224 Khyber Pakhtunkhwa6 15.8 23 99,292 0.23 165 Balochistan 29.2 13 28,987 0.45 298 Total 15-497 13.5 120 524,197 0.23a 186a Azad Jammu and Kashmir 6.4 9 81,048 0.11 104 Gilgit Baltistan 15.8 12 56,225 0.22 157 1 A maternal death is defined as the death of a woman while pregnant, during childbirth, or within 42 days after delivery for which there was a verbal autopsy that classified deaths as being either direct or indirect maternal deaths. 2 Woman-years lived in that age group during the 36 months before the survey. For example, for the 15-19 age group, this is calculated by taking ½ of the number of women age 15, plus 1½ times the number age 16, plus 2½ times the number age 17, plus 3 times the number age 18, plus 3 times the number age 19, plus 2½ times the number age 20, plus 1½ times the number age 21, plus ½ times the number age 22, plus 1½ times the number of deaths among women age 15-49 in the previous 36 months. 3 Expressed per 1,000 woman-years of exposure 4 Expressed per 100,000 live births; calculated as the age-adjusted maternal mortality rate times 100 divided by the age-adjusted general fertility rate 5 Punjab includes ICT. 6 Khyber Pakhtunkhwa includes the merged districts of former FATA. 7 Total excludes Azad Jammu and Kashmir and Gilgit Baltistan. a Age-adjusted rate Adult and Maternal Mortality • 49 Table 3.10 Maternal mortality ratio Total fertility rate, general fertility rate, maternal mortality ratio, and lifetime risk of maternal death for the 3 years preceding the survey, by residence and region, Pakistan MMS 2019 Residence Region Total3 Azad Jammu and Kashmir Gilgit Baltistan Urban Rural Punjab1 Sindh Khyber Pakhtun- khwa2 Balochistan Total fertility rate (TFR) 3.2 4.3 3.7 3.9 4.4 5.1 3.9 3.6 4.8 General fertility rate (GFR)4 102 137 120 124 139 152 124 110 141 Maternal mortality ratio (MMR)5 158 199 157 224 165 298 186 104 157 MMR (95% CI, lower bound) 91 136 79 148 84 130 138 23 53 MMR (95% CI, upper bound) 225 263 235 299 246 466 234 185 261 Lifetime risk of maternal death6 0.005 0.009 0.006 0.009 0.007 0.015 0.007 0.004 0.007 CI: Confidence interval 1 Punjab includes ICT. 2 Khyber Pakhtunkhwa includes the merged districts of former FATA. 3 Total excludes Azad Jammu and Kashmir and Gilgit Baltistan. 4 Age-adjusted rate, expressed per 1,000 women age 15-49 5 Expressed per 100,000 live births; calculated as the age-adjusted maternal mortality rate times 100 divided by the age-adjusted general fertility rate 6 Calculated as 1-(1-MMR)TFR, where TFR represents the total fertility rate for the 3 years preceding the survey Table 3.11 Maternal mortality ratio using direct method Maternal mortality ratios for the 3 years preceding the survey, by residence and region, Pakistan MMS 2019 Background characteristic Maternal deaths1 Live births Maternal mortality ratio Residence Urban 32 20,333 159 Rural 88 43,290 203 Region Punjab2 52 31,753 165 Sindh 33 13,786 237 Khyber Pakhtunkhwa3 23 14,075 161 Balochistan 13 4,010 317 Total4 120 63,623 189 Azad Jammu and Kashmir 9 8,501 108 Gilgit Baltistan 12 7,712 162 1 A maternal death is defined as the death of a woman while pregnant, during childbirth, or within 42 days after delivery for which there was a verbal autopsy that classified deaths as being either direct or indirect maternal deaths. 2 Punjab includes ICT. 3 Khyber Pakhtunkhwa includes the merged districts of former FATA. 4 Total excludes Azad Jammu and Kashmir and Gilgit Baltistan. Causes of Death • 51 CAUSES OF DEATH 4 Key Findings  All-cause mortality: Circulatory disease, infectious and parasitic disease, and neoplasms are the most common causes of death among women of reproductive age.  Maternal causes of death: 12% of deaths are due to complications of during pregnancy, childbirth, and the puerperium. Ninety-six percent of maternal deaths are due to direct obstetric causes and 4% are due to non- obstetric (indirect) causes.  Direct maternal causes of death: Obstetric haemorrhage is the most frequent cause of maternal death (41%), followed by hypertensive disorders (29%).  Treatment for deceased women: The majority of deceased women of reproductive age (15-49 years), including those dying as a result of pregnancy and childbirth complications, sought care from health facilities in the public sector. his chapter presents results of the verbal autopsy interviews conducted to determine causes of death among women age 15-49 in the 3 years preceding the 2019 PMMS. Knowing the causes of maternal deaths can help in formulating policies to prevent those deaths. However, obtaining data on deaths requires robust vital registration systems, which are lacking in most developing countries, including Pakistan. In the absence of such systems, the verbal autopsy method (an interview with family members or caregivers of the deceased person) has been developed as a way of collecting information that can be used to ascertain the cause or causes of death. 4.1 VERBAL AUTOPSY QUESTIONNAIRE The Verbal Autopsy Questionnaire (VAQ) used in the 2019 PMMS was based on the 2016 WHO standard VAQ for adults age 15-49 (version 1.5), with adaptations to address the country-specific context and to preserve comparability with the 2006-07 PDHS instrument. The questionnaire collected information on the deceased woman’s socioeconomic background, whether she experienced any accident or violence leading to death, specific diagnoses she may have received, signs and symptoms of a disease or a complication of pregnancy or childbirth in the period preceding death, and treatment by or contact with health care providers before death. 4.2 VERBAL AUTOPSY FIELDWORK During the household listing phase, every household in every cluster was asked if any household member had died in the 3 years preceding the survey (i.e., since January 2016). If the answer was yes, information on the name, sex, age at death, and year of death was collected. Later, fieldwork teams in each cluster were provided with a list of all households where a female in the 15-49 age group had died. Interviews were completed only in cases in which deaths occurred on or after 1 January 2016, the deceased woman was age T 52 • Causes of Death 15-49 at the time of the death, and the household could furnish a respondent who had knowledge of the circumstances preceding the death of the woman. During fieldwork, 1,182 verbal autopsies were conducted for women age 15-49 who died on or after 1 January 2016. Of these 1,182 VAQs, 1,177 (unweighted) were completed for women who died in the 3 years preceding the survey. 4.3 CAUSE OF DEATH CERTIFICATION AND ICD-10 CODING For the verbal autopsy process to be complete, each questionnaire must be reviewed so that a death certificate with the immediate and underlying cause(s) of death for the deceased person can be completed. Coding the cause(s) of death recorded in the death certificate according to the International Statistical Classification of Diseases and Related Health Problems, 10th revision (ICD-10; WHO 2016a), is a further step that produces internationally comparable data on the final underlying cause of death. Three panels of three reviewers each were created (with two obstetricians/gynaecologists and one specialist physician on each panel). The three doctors filled out death certificates according to WHO guidelines, and ICD-10 coding was done by another two physicians who had been trained in coding at a workshop facilitated by ICF. Death Certification The WHO standard death certificate (shown below) requires certifiers to identify a sequence of conditions and/or events leading to death that make both chronological and pathological sense. Conditions and events (for example, road traffic accidents) that have a causal relationship to death are listed in Part 1 of the certificate. The immediate (direct) cause of death is listed on the first line (a) of Part 1. There must always be an entry on line 1(a), but it is possible that in Part 1 only line 1(a) is filled out. When there are two or more conditions and/or events that form part of the sequence leading directly to death, each condition/event should be recorded on a separate line in Part 1. The last-listed condition/event in Part 1 is considered the underlying cause of death (the “tentative starting point” in the process of ICD-10 coding described below). If the tentative starting point had not taken place, death would have been averted. In the example below, line 1(c) is the underlying cause of death. Conditions and events that are significant and contributory but not causally related to death are listed in Part 2 of the certificate. Administrative Data Sex Female Date of birth 3 0 0 1 1 9 7 2 Date of death 3 1 0 1 2 0 1 8 Death Certificate: Part 1 and 2 1 Report disease or condition directly leading to death on line a Report chain of events in due to order (if applicable) State the underlying cause on the lowest used line Cause of death Time interval from onset to death a Haemorrhagic shock 1 hour b Due to: Abdominal haemorrhage 2 hours c Due to: Passenger in car that hit another car 2 hours d Due to : 2 Other significant conditions contributing to death (time intervals can be included in brackets after the condition) Diabetes mellitus (5 years) Causes of Death • 53 The reviewers assessed the 2019 PMMS VAQs and recorded their comments in the form developed from the WHO guidelines. The form provided the following information on the deceased woman according to the VAQ:  Tentative underlying cause of death.  Final underlying cause of death.  Reviewer’s decision on the category of death (direct obstetric death, indirect obstetric death, probable obstetric death, coincidental obstetric death, late maternal death, non-obstetric death, undecided on category of death).  Timing of death in relation to pregnancy, delivery, or the postpartum period.  Delay in seeking treatment (only required for maternal deaths).  Reviewer’s assessment of the quality of the information in the VAQ. Each reviewer assessed the quality of data on a scale of 1-5 (whereby 1 was considered as the poorest quality and 5 as the highest quality). Scoring was based on the following criteria: missing information, discrepancy within objective data, and discrepancy between objective data and the verbatim history of the fatal illness. ICD-10 Coding The coding of causes of death was done by applying the standard ICD-10 coding procedures. This process is essential for correct application of the ICD-10 instructions and for selection and modification of the tentative starting point. The ICD-10 categorises diseases and health conditions into four-digit codes. The hard copy version consists of three books. Volume 1 is a tabular list (by ICD-10 code) of diseases and conditions; Volume 2 provides guidance on the use of the ICD-10, including coding procedures; and Volume 3 is an alphabetical index of diseases and health conditions. To code a death certificate and determine the final underlying cause of death, coders first look up each entry on each line of the death certificate in the alphabetical index (Volume 3) to obtain the ICD-10 code for the entry and then look up each code in Volume 1 to verify that it is the correct one and see if there are any special instructions on how to use the code (which may lead to modification of the tentative starting point). After the code for each entry on the death certificate has been obtained, coders determine the tentative underlying cause of death by applying steps SP1 through SP8 to the codes in the death certificate as described in Volume 2. Next, coders apply the special instructions for modifications of the starting point (steps M1 to M4) as described in Volume 2, check for age and sex inconsistencies, and arrive at the final underlying cause of death. It is this final underlying cause of death that has ultimately been tabulated in the 2019 PMMS final report. When the three panel reviewers completed the review and returned the completed 2019 PMMS assessment of verbal autopsy forms, the information from these three forms was aggregated by one of the two ICD-10 coders on another (fourth) yellow coloured 2019 PMMS assessment of verbal autopsy form to determine if there was a consensus among the three reviewers about the category and underlying cause of death. 54 • Causes of Death Figure 4.1 provides the physician review process for the verbal autopsies. If at least two of the three reviewers agreed on the category and underlying cause of death, it was accepted as the category and underlying cause of death for that VAQ. If all three reviewers did not agree on the underlying cause and category of death, VAQs were reviewed by two original reviewers and a third lead reviewer. Figure 4.1 Physician review process for verbal autopsies The fourth assessment form was filled out in green if there was a consensus among the reviewers. However, if there was no agreement, the fourth form was filled out in red and categorised as “undetermined.” 4.4 CHARACTERISTICS OF DECEASED WOMEN The prime focus of the 2019 PMMS was identifying female deaths in survey households in an effort to gather detailed information regarding causes of death and estimate maternal mortality. The VAQ was completed and a final death certificate written for 1,177 women (unweighted) who died at age 15-49 in the 3 years preceding the survey. Table 4.1 shows that 38% of deceased women were age 40 or above at the time of their death and 72% were married. A majority were rural residents (69%), had no education (63%), and were not working (82%). One in 10 (11%) deceased women had a secondary education or higher. Twenty-one percent had never been married. Female deaths are distributed almost uniformly across the five wealth quintiles. Thirty-three percent of deceased women’s husbands had no education, while 23% had a secondary education or higher. In Azad Jammu and Kashmir, 37% of deceased women were age 40 or older at the time of their death, 73% were married, 85% lived in rural areas, 33% had no education, and 91% were not working. Forty-four percent of the husbands of deceased women in Azad Jammu and Kashmir had a secondary education or higher. Thirty-one percent of deceased women in Gilgit Baltistan were age 40 or older at the time of their death, 69% were married, 87% resided in rural areas, 64% had no education, and almost all were not working (98%). Twenty-nine percent of deceased women’s husbands had no education. Causes of Death • 55 4.5 RESPONDENTS TO THE VERBAL AUTOPSY QUESTIONNAIRES Verbal autopsy interviews were conducted with the deceased woman’s next of kin (one or more members of her household who were present during the fatal illness and/or at the time of death and who knew the most about her personal life). These interviews were conducted for all deaths of women age 15-49 identified during the first round of the survey to ascertain causes of death and identify maternal deaths as per the ICD-10 classification. Table 4.2 shows the characteristics of respondents to the verbal autopsy interviews. At the national level, 82% of verbal autopsies were conducted with more than one respondent, and in 94% of cases at least one respondent was present at the time of death. The corresponding percentages were 83% and 91% in Azad Jammu and Kashmir and 94% and 95% in Gilgit Baltistan. Brothers-in-law or sisters-in- law were the most common respondents for the verbal autopsy interviews (52% in Pakistan, 65% in Azad Jammu and Kashmir, and 69% in Gilgit Baltistan). Other common respondents were husbands, sons or daughters, parents, and brothers or sisters. 4.6 CAUSE-SPECIFIC MORTALITY Underlying cause of death The disease or injury that initiated the sequence of morbid events leading directly to death, or the circumstances of the accident or violence that produced the fatal injury (WHO 2016a). Sample: Women who died at age 15-49 in the 3 years preceding the survey The most common causes of death among women of reproductive age are circulatory disease (20%), infectious and parasitic diseases (14%), and neoplasms (14%). While the percentages of deaths caused by circulatory disease and neoplasms generally increase with age, deaths from infectious and parasitic diseases fluctuate somewhat erratically by age group (Table 4.3). Maternal causes such as complications of pregnancy, childbirth, and the puerperium are responsible for 12% of deaths among women of childbearing age in Pakistan. Deaths due to transport accidents and other external factors also account for 12% of the total (Figure 4.2). Patterns by background characteristics  Older women (age 35-49) and teenagers (age 15–19) are more likely to die from infectious and parasitic diseases than women age 20-34. Almost two-thirds of deaths from maternal causes occur at age 25-39. Deaths from circulatory disease are most common among women age 40-49.  Deaths from conditions related to the nervous system, digestive system, and respiratory system are more common among younger women (age 15-34) and women in rural areas (11%).  The percentage of deaths from infectious and parasitic diseases is highest in Sindh (18%). Figure 4.2 All-cause mortality Maternal 12% Infectious and parasitic disease 14% Other disease 61% Transport accident and other external 12% No cause determined 1% Percent distribution of deceased women age 15-49 in the 3 years before the survey Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan 56 • Causes of Death  Balochistan has a higher percentage of deaths (18%) due to complications of pregnancy, childbirth, and the puerperium than other provinces (14% each in Sindh and Khyber Pakhtunkhwa and 10% in Punjab).  Deaths due to transport accidents are higher in Punjab (5%) than in any other province. 4.7 MATERNAL CAUSES OF DEATH Ninety-six percent of maternal deaths were due to direct obstetric causes and 4% were due to non- obstetric (indirect) causes. The most frequent causes of maternal death were obstetric haemorrhage (41%) and hypertensive disorders (29%). Pregnancies with abortive outcomes were the cause of 10% of maternal deaths, and another 6% of women died due to pregnancy-related infections (Table 4.4 and Figure 4.3). Trends: Some causes of death increased as a percentage of all obstetric causes between 2006-07 and 2019. Obstetric haemorrhage increased from 33% to 41%, hypertensive disorders increased from 10% to 29%, and pregnancies with abortive outcomes increased from 6% to 10% (Figure 4.4). Other causes of death decreased as a percentage of all obstetric causes over that period. Pregnancy-related infections fell from 14% to 6%, and non-obstetric causes decreased from 13% to 4%. While iatrogenic causes of death contributed to 8% of maternal deaths in 2006-07, this specific cause was not identified in the 2019 PMMS; however, surgical and medical negligence were the trigger point for many maternal deaths, especially deaths due to obstetric haemorrhage. Figure 4.3 Maternal causes of death Figure 4.4 Trends in obstetric-coded deaths Pregnancy with abortive outcome 10% Hypertensive disorders 29% Obstetric haemorrhage 41% Pregnancy- related infection 6% Other obstetric 10% Non-obstetric 4% Percent distribution of deceased women age 15-49 in the 3 years before the survey Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan 6 10 33 14 7 13 8 3 10 29 41 6 10 4 0 Pregnancy with abortive outcome Hypertensive disorders Obstetric haemorrhage Pregnancy-related infection Other obstetric Non-obstetric Iatrogenic Unspecified/undetermined Percent distribution of deceased women age 15-49 in the 3 years preceding the survey whose death was due to a maternal cause 2006-07 PDHS 2019 PMMS Notes: Iatrogenic was not used as an underlying cause of death in the 2019 PMMS. Excludes Azad Jammu and Kashmir and Gilgit Baltistan Causes of Death • 57 4.8 DECEASED WOMEN AND HEALTH CARE Among women who died at age 15-49 (both maternal and non- maternal deaths) in the 3 years preceding the survey, the majority received care from a public sector health facility (Table 4.5). Forty- two percent of women who died due to maternal causes received treatment from only the public sector, whereas 36% received care from the private sector only (Figure 4.5). Patterns by background characteristics  Women in all of the age groups other than age 20-24 were more likely to receive treatment from public than private facilities.  A larger percentage of urban than rural women received treatment only from the private sector (33% versus 23%).  The percentage of women who received treatment only from a public sector facility was highest in Khyber Pakhtunkhwa (42%) and lowest in Balochistan (29%). Place of Death One in two women (51%) died at a hospital regardless of whether the death was maternal or non-maternal, while 36% died at home. Thirteen percent of women died on their way to the hospital or while returning home from the hospital (Table 4.6). A higher proportion of maternal deaths than non-maternal deaths occurred in a health facility (58% versus 50%). The proportion of women who died on the way to or while returning home from a health facility was twice as high among those dying from maternal causes as among those dying from non-maternal causes (22% versus 11%). A higher proportion of non-maternal deaths (38%) occurred at home than maternal deaths (18%). Patterns by background characteristics  Sixty-two percent of women in urban areas died at a health facility, compared with 45% of women in rural areas. The percentage of women dying on the way to the hospital or while returning home is higher in rural areas than in urban areas (15% and 7%, respectively).  Khyber Pakhtunkhwa had the highest proportion of hospital-based deaths (60%), followed by Balochistan (54%), Sindh (51%), and Punjab (47%). Figure 4.5 Treatment received for deceased women 42 3736 25 17 24 1 105 5 Maternal cause Non-maternal cause Percentage of deceased women age 15-49 in the 3 years preceding the survey who received medical care Public sector only Private sector only Public and private sector Home and health facility Home only Note: Excludes Azad Jammu and Kashmir and Gilgit Baltistan 58 • Causes of Death LIST OF TABLES For more information on data from verbal autopsy, see the following tables:  Table 4.1 Background characteristics of deceased women  Table 4.2 Respondents to Verbal Autopsy Questionnaires  Table 4.3 All cause-specific mortality  Table 4.4 Causes of maternal deaths  Table 4.5 Treatment received for deceased women  Table 4.6 Place of death Table 4.1 Background characteristics of deceased women Percent distribution of women age 15-49 who died in the 3 years before the survey by selected background characteristics at the time of their death, Pakistan MMS 2019 Pakistan Azad Jammu and Kashmir Gilgit Baltistan Background characteristic Weighted percent Weighted number Unweighted number Weighted percent Weighted number Unweighted number Weighted percent Weighted number Unweighted number Age 15-19 10.2 89 98 10.3 15 12 11.7 9 7 20-24 12.3 107 105 13.0 19 15 12.9 10 12 25-29 11.1 97 98 13.8 20 16 16.8 13 11 30-34 11.8 103 96 12.3 18 17 11.1 9 7 35-39 16.4 144 135 13.4 19 18 16.8 13 14 40-44 15.5 135 133 12.7 18 21 9.6 7 10 45-49 22.7 198 191 24.5 35 41 21.1 16 16 Marital status Married 72.1 629 617 73.2 105 100 68.7 52 51 Divorced/separated 2.8 24 26 2.2 3 3 3.4 3 3 Widowed 4.4 38 38 6.4 9 9 1.9 1 1 Never married 20.8 181 175 18.2 26 28 25.9 20 22 Residence Urban 31.5 275 376 14.9 21 62 13.3 10 17 Rural 68.5 598 480 85.1 122 78 86.7 66 60 Education No education 63.3 552 544 33.4 48 48 63.5 49 46 Primary1 16.0 140 132 17.5 25 25 8.3 6 9 Middle2 9.2 80 68 22.4 32 26 7.7 6 4 Secondary3 5.4 48 55 18.8 27 27 11.3 9 8 Higher4 6.0 53 54 7.9 11 14 9.2 7 10 Don’t know 0.0 0 3 0.0 0 0 0.0 0 0 Husband’s education Woman was never married 20.8 181 175 18.2 26 28 25.9 20 22 No education 32.9 287 257 11.7 17 15 29.1 22 21 Primary1 14.9 130 128 15.6 22 22 15.0 11 7 Middle2 8.4 73 81 10.6 15 19 10.3 8 9 Secondary3 13.5 118 126 33.1 47 36 12.5 10 9 Higher4 9.2 80 84 10.7 15 19 7.1 5 9 Don’t know 0.3 3 5 0.2 0 1 0.0 0 0 Employment status Working 17.3 151 152 8.9 13 13 2.5 2 1 Not working 82.4 719 702 91.1 130 127 97.5 74 76 Don’t know 0.3 3 2 0.0 0 0 0.0 0 0 Wealth quintile Lowest 21.0 183 186 5.8 8 6 31.3 24 22 Second 21.6 189 170 27.2 39 25 37.5 29 25 Middle 21.4 187 171 29.7 42 41 20.6 16 21 Fourth 18.4 160 163 21.9 31 40 8.9 7 7 Highest 17.6 154 166 15.4 22 28 1.6 1 2 Region Punjab5 55.7 487 330 na na na na na na Sindh 23.4 204 232 na na na na na na Khyber Pakhtunkhwa6 15.9 139 181 na na na na na na Balochistan 5.0 43 113 na na na na na na Total 15-49 100.0 873 856 100.0 143 140 100.0 76 77 na = Not applicable 1 Primary refers to classes 1-5. 2 Middle refers to classes 6-8. 3 Secondary refers to classes 9-10. 4 Higher refers to classes 11 and above. 5 Punjab includes ICT. 6 Khyber Pakhtunkhwa includes the merged districts of former FATA. Causes of Death • 59 Table 4.2 Respondents to Verbal Autopsy Questionnaires Percentage of respondents by their relationship to the deceased woman, Pakistan MMS 2019 Relationship to deceased woman Relationship Pakistan Azad Jammu and Kashmir Gilgit Baltistan Husband 38.2 35.9 42.4 Son or daughter 38.4 27.4 22.9 Son-in-law or daughter-in-law 10.5 3.7 9.4 Grandchild 0.2 1.4 0.0 Parent 27.5 22.5 36.3 Parent-in-law 20.8 17.8 25.3 Brother or sister 27.4 23.9 32.0 Brother-in-law/sister-in-law 52.3 64.8 68.6 Niece/nephew 5.4 10.3 6.0 Grandparent 0.7 2.5 4.0 Aunt/uncle 10.5 9.3 7.3 Other relative 12.3 11.1 29.1 Adopted/foster/stepchild 0.3 0.0 0.0 Not related 4.0 0.5 0.9 Domestic servant 0.3 0.0 0.0 Percentage with more than one respondent 81.8 82.6 93.7 Percentage with at least one respondent who was present when the deceased fell ill 96.1 95.0 93.7 Percentage with at least one respondent who was present when the deceased died 94.1 90.9 95.1 Number of deceased women 873 143 76 Table 4.3 All cause-specific mortality Percent distribution of deceased women age 15-49 in the 3 years preceding the survey by probable underlying cause of death, according to background characteristics, Pakistan MMS 2019 Background characteristic Maternal causes1 Infectious and parasitic disease2 Neoplasms3 Circulatory disease4 Transport accidents5 Other external causes6 Nervous system, digestive, respiratory7 Other causes (classified)8 No cause deter- mined9 Total Number of deceased women Age at death 15-19 12.9 17.9 7.7 6.2 4.2 11.6 15.0 24.4 0.0 100.0 89 20-24 14.5 10.8 11.0 15.3 1.2 13.7 12.5 17.6 3.6 100.0 107 25-29 22.5 8.1 13.3 16.7 0.0 8.6 13.4 16.1 1.3 100.0 97 30-34 22.3 3.9 14.2 18.6 5.6 12.5 10.8 12.0 0.0 100.0 103 35-39 19.9 15.8 14.9 16.9 2.2 6.6 4.3 17.3 2.1 100.0 144 40-44 2.3 15.8 19.6 27.9 3.4 5.5 8.8 16.7 0.0 100.0 135 45-49 0.6 20.4 14.6 29.6 4.3 7.6 7.0 15.8 0.0 100.0 198 Residence Urban 12.0 13.6 13.3 22.8 3.1 8.9 7.3 18.6 0.3 100.0 275 Rural 12.0 14.4 14.4 19.2 3.1 9.0 10.5 16.1 1.2 100.0 598 Region Punjab10 10.3 14.1 14.4 22.3 4.6 9.6 9.0 14.9 0.8 100.0 487 Sindh 13.7 17.5 16.5 15.4 1.9 10.2 9.6 15.2 0.0 100.0 204 Khyber Pakhtunkhwa11 13.7 9.9 10.9 20.2 0.5 6.0 10.9 25.0 3.0 100.0 139 Balochistan 18.3 12.7 9.4 23.0 0.6 5.8 9.8 20.3 0.2 100.0 43 Total 12.0 14.2 14.1 20.4 3.1 9.0 9.5 16.9 0.9 100.0 873 Note: Table excludes Azad Jammu and Kashmir and Gilgit Baltistan. 1 ICD-10 codes O00-O994 2 ICD-10 codes A010-B24 3 ICD-10 codes C069-D434 4 ICD-10 codes I081-I802 5 ICD-10 codes V892, V878 6 ICD-10 codes S068-T909, W34-Z915 7 ICD-10 codes G039-G948, J188-J969, K027-K922 8 ICD-10 codes D561-D70, E059-E669, L100, L899, M069-M629, N049-N390, R100-R99 9 ICD-10 codes Q223, Q249 10 Punjab includes ICT. 11 Khyber Pakhtunkhwa includes the merged districts of former FATA. 60 • Causes of Death Table 4.4 Causes of maternal deaths Percent distribution of deceased women age 15-49 in the 3 years preceding the survey who died from maternal causes, by cause of death, Pakistan MMS 2019 Cause of death Maternal deaths Maternal death: direct 95.9 Pregnancy with abortive outcome1 9.8 Hypertensive disorders in pregnancy, childbirth, and the puerperium2 29.0 Obstetric haemorrhage3 40.7 Pregnancy-related infection4 6.4 Other obstetric complications5 10.0 Maternal death: indirect 4.1 Non-obstetric complications6 4.1 Total 100.0 Number of deceased women who died from maternal causes 105 Note: Table excludes Azad Jammu and Kashmir and Gilgit Baltistan. 1 ICD-10 codes O00-O75 2 ICD-10 codes O101-O159 3 ICD-10 codes O441-O469, O670, O711, O72, O720, O721, O722 4 ICD-10 codes O223, O85 5 ICD-10 codes O211, O639, O669, O731, O759, O871, O882, O900, O909 6 ICD-10 codes O244, O990, O994 Table 4.5 Treatment received for deceased women Percent distribution of deceased women age 15-49 in the 3 years preceding the survey who received medical care by place of care, according to background characteristics, Pakistan MMS 2019 Background characteristic Public sector only Private sector only Public and private sector Home only Home and public/private sector Total Number of deceased women Cause of death Maternal 41.9 35.6 16.5 4.9 0.9 100.0 90 Non-maternal 36.8 25.2 23.6 4.6 9.8 100.0 678 Age at death 15-19 34.0 22.5 27.6 7.7 8.2 100.0 76 20-24 35.7 36.9 17.4 3.9 5.3 100.0 95 25-29 37.4 27.9 26.5 5.0 2.8 100.0 84 30-34 47.7 22.4 17.6 7.1 5.2 100.0 89 35-39 41.4 27.4 19.5 6.0 5.7 100.0 132 40-44 35.9 20.1 27.4 1.2 15.3 100.0 122 45-49 32.5 27.3 23.6 3.4 13.2 100.0 171 Residence Urban 34.6 32.6 22.1 2.2 8.2 100.0 249 Rural 38.8 23.4 23.1 5.7 9.0 100.0 519 Region Punjab1 39.4 17.7 23.9 6.5 12.4 100.0 426 Sindh 31.5 43.9 16.9 3.0 4.5 100.0 184 Khyber Pakhtunkhwa2 41.7 28.6 25.6 1.5 2.6 100.0 122 Balochistan 29.4 32.4 30.4 1.0 6.9 100.0 35 Total 37.4 26.4 22.8 4.6 8.7 100.0 768 Note: Table excludes Azad Jammu and Kashmir and Gilgit Baltistan. 1 Punjab includes ICT. 2 Khyber Pakhtunkhwa includes the merged districts of former FATA. Causes of Death • 61 Table 4.6 Place of death Percent distribution of deceased women age 15-49 in the 3 years preceding the survey by place of death, according to background characteristics, Pakistan MMS 2019 Background characteristic Hospital Home Way to hospital or returning home Other Total Number of deceased women Cause of death Maternal 57.6 17.8 22.0 2.5 100.0 90 Non-maternal 49.6 38.4 11.2 0.8 100.0 678 Age at death 15-19 46.8 37.1 16.1 0.0 100.0 76 20-24 60.8 25.4 13.4 0.4 100.0 95 25-29 47.1 35.6 14.6 2.7 100.0 84 30-34 50.3 28.1 20.4 1.3 100.0 89 35-39 44.5 45.5 8.9 1.0 100.0 132 40-44 56.1 34.1 9.2 0.6 100.0 122 45-49 48.9 39.7 10.3 1.2 100.0 171 Residence Urban 62.1 31.0 6.6 0.3 100.0 249 Rural 44.9 38.4 15.3 1.4 100.0 519 Region Punjab1 47.2 39.7 12.1 1.0 100.0 426 Sindh 51.3 34.8 12.3 1.5 100.0 184 Khyber Pakhtunkhwa2 59.8 26.4 13.9 0.0 100.0 122 Balochistan 53.8 30.7 13.5 2.0 100.0 35 Total 50.5 36.0 12.5 1.0 100.0 768 Note: Table excludes Azad Jammu and Kashmir and Gilgit Baltistan. 1 Punjab includes ICT. 2 Khyber Pakhtunkhwa includes the merged districts of former FATA. Maternal Health Care • 63 MATERNAL HEALTH CARE 5 Key Findings  Antenatal care: 91% of women who gave birth in the 3 years before the survey received antenatal care (ANC) from a skilled provider. Fifty-two percent of women had at least four antenatal care visits.  Components of antenatal care: Among women who received ANC for their most recent live birth or stillbirth, 89% had their blood pressure measured, 71% had a blood sample taken, and 65% had a urine sample taken.  Advice during antenatal care: 67% of women received advice on the importance of maintaining a balanced diet during pregnancy, and more than half received advice on the importance of exclusive breastfeeding and early initiation of breastfeeding.  Protection against neonatal tetanus: 70% of the most recent births or stillbirths to women in the 3 years before the survey were protected against neonatal tetanus.  Delivery: 71% of births were delivered in a health facility. Seventy-four percent of births were assisted by a skilled provider.  Postnatal checks: 69% of women received a postnatal check within 2 days of delivery. ccessible and quality health care services during pregnancy, childbirth, and the postnatal period are considered an essential factor in reducing maternal mortality in Pakistan. One of the targets of the Sustainable Development Goals (SDGs) 2015-2030 (

View the publication