Reproductive Health Survey Azerbaijan, 2001 Final Report

Publication date: 2003

REPRODUCTIVE HEALTH SURVEY AZERBAIJAN, 2001 FINAL REPORT Edited by: Florina Serbanescu, Ml) Leo Morris, PhD Shafag Rahimova, MD, PhD Paul W. Stupp, PhD Adventist Development and Relief Agency (ADRA) Azerbaijan Ministry of Health State Committee of Statistics (SCS) Mercy Corps (MC) BAKU, AZERBAIJAN Division of Reproductive Health, Centers for Disease Control and Prevention (DRH/CDC) ATLANTA, GEORGIA USA United States Agency for International Development (USAID) United Nations Population Fund (UNFPA) United Nations High Commissioner for Refugees (UNHCR) March, 2003 PRINTED BY: U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES Centers for Disease Control and Prevention (CDC) Atlanta, GA, 30333 Cover: Original painting comissioned by ADRA to Azeri artist, Javid Akhadov; cover design by Rose Pecoraro, visual communication specialist, DRH/CDC. Additional information about the AZRHS01 English Final Report may be obtained from: Division of Reproductive Health, Centers for Disease Control and Prevention (DRH/CDC), Mailstop K-35, 4770 Buford Highway, N.E., Atlanta, Georgia 30341- 3724, USA. Fax (770) 488-6242, phone (770) 488-6200, E-mail fserbanescu @ cdc.gov. Additional information about the AZRHS01 Azeri Final Report may be obtained from: Adventist Development and Relief Agency (ADRA), Azerbaijan; 68 Vidali St., PO Box 79. Baku, Azerbaijan. Fax: +(99 4 12) 47 43 59, phone: +(99 4 12) 47 43 57/58, E- mail office®adra.org.az. TABLE OF CONTENTS PREFACE .i ACKNOWLEDGMENTS . iii EXECUTIVE SUMMARY . v Chapter 1 INTRODUCTION. 1 (Florina Serbanescu, Ranee Seither, Shafag Rahimova) Chapter 2 METHODOLOGY. . 9 (Florina Sebanescu, Paul W. Stupp, Leo Morris) 2.1 Sampling Design . 9 2.2 Data Collection . 11 2.3 Response Rates . 12 Chapter 3 CHARACTERISTICS OF THE SAMPLE . 15 (Florina Serbanescu, Leo Morris) 3.1 Household Characteristics . 15 3.2 Characteristics of the Respondents . 18 Chapter 4 FERTILITY AND PREGNANCY EXPERIENCE. 21 (Florina Serbanescu, Leo Morris, Paul W. Stupp) 4.1 Fertility Levels and Trends . 21 4.2 Fertility Differentials . 27 4.3 Nuptiality. 29 4.4 Age at First Intercourse, Union, and Birth . 31 4.5 Recent Sexual Activity . 35 4.6 Planning Status of the Last Pregnancy . 37 4.7 Future Fertility Preferences . 40 Chapter 5 INDUCED ABORTION . 45 (Florina Serbanescu, Ranee Seither, Leo Morris) 5.1 Abortion Levels and Trends . 45 5.2 Induced Abortion Differentials . 51 5.3 Abortion Services . 55 5.4 Abortion Complications. 64 5.5 Reasons for Abortion. 67 Chapter 6 MATERNAL AND CHILD HEALTH.69 (Florina Serbanescu, Sharon Daves, Tonji Durant, Paul W. Stupp) 6.1 Prenatal Care .70 6.2 Intrapartum Care .79 6.3 Postnatal Care .85 6.4 Smoking and Drinking During Pregnancy .91 6.5 Pregnancy and Postpartum Complications . 91 6.6 Poor Birth Outcomes .96 6.7 Breastfeeding .98 6.8 Infant and Child Mortality . 101 Chapter 7 NUTRITIONAL STATUS OF MOTHERS AND CHILDREN.109 (Geraldine S. Perry, Florina Serbanescu, Paul W. Stupp, Abeda Hussain, Linda Fardy Hayes, Larry Grummer-Strawn) 7.1 Methodology .109 7.2 Changes in Levels of Chronic and Acute Malnutrition: 1996 to 2001 .111 7.3 Chronic Malnutrition .112 7.4 Acute Malnutrition .112 7.5 General Malnutrition .115 7.6 Prevalence of Anemia in Children Aged 12-59 Months.117 7.7 Nutritional Status and Anemia Levels Among Mothers with Children Aged 3-59 Months . 119 Chapter 8 CONTRACEPTIVE AWARENESS AND KNOWLEDGE OF USE. 123 (Florina Serbanescu, Linda Fardy Hayes, Leo Morris) 8.1 Contraceptive Awareness and Knowledge of Use . 123 8.2 Knowledge About Contraceptive Source and Effectiveness . 129 8.3 First Source of Information About Contraception. 132 Chapter 9 CURRENT AND PAST CONTRACEPTIVE USE . 135 (Florina Serbanescu, Leo Morris, Lisa Flowers) 9.1 Current Contraceptive Prevalence . 135 9.2 Source of Contraceptive Methods . 141 9.3 Dissatisfaction with the Current Method and Preference for Other Methods 143 9.4 Users of Traditional Methods . 146 9.5 Reasons for Not Using Contraception . 150 9.6 Intention to Use Contraception Among Nonusers. 151 9.7 Recent Trends in Contraceptive Use . 153 9.8 Contraceptive Failure and Discontinuation . 156 Chapter 10 NEED FOR CONTRACEPTIVE SERVICES . 161 (Florina Serbanescu, Leo Morris, Linda Fardy Hayes) 10.1 Potential Demand and Unmet Need for Contraception .161 10.2 Potential Demand for Family Planning Services According to Fertility Preferences . 166 Chapter 11 CONTRACEPTIVE COUNSELING .171 (Florina Serbanescu, Leo Morris, Lisa Flowers) 11.1 Communication With Family Planning Providers.171 11.2 Postabortion Counseling .173 Chapter 12 OPINIONS ABOUT CONTRACEPTION AND ABORTION .177 (Tonji Durant, Florina Serbanescu, Shafag Rahimova) 12.1 Interest in More Information on Contraception .178 12.2 Opinions About the Most Reliable Source of Information on Contraceptionl81 12.3 Opinions on the Advantages and Disadvantages of the Pill and IUD.182 12.4 Opinions on Risks to Women's Health Due to Contraceptive Use .185 12.5 Opinions on Risks to Women's Health Due to Abortion .190 Chapter 13 REPRODUCTIVE HEALTH KNOWLEDGE AND ATTITUDES.193 (Ranee Seither, Florina Serbanescu, Leo Morris) 13.1 Ideal Family Size .193 13.2 Knowledge of the Menstrual Cycle .195 13.3 Knowledge of the Fertility Effect of Breast-feeding .195 13.4 Attitudes Toward Abortion .197 13.5 Attitudes and Perceptions About Reproductive Norms and Gender Roles . 203 Chapter 14 HEALTH BEHAVIORS . 207 (Florina Serbanescu, Ranee Seither, Linda Fardy Hayes) 13.1 Prevalence of Routine Gynecologic Visits . 208 13.2 Breast Self-Examination . 210 13.3 Cervical Cancer Screening. 212 13.4 Prevalence of Selected Health Problems . 215 13.5 Impaired Fecundity . 217 13.1 Cigarette Smoking . 221 Chapter 15 FAMILY LIFE EDUCATION . 225 (Florina Serbanescu, Azam Buzurukov, Shafag Rahimova) 15.1 Opinions about Family Life Education in School . 226 15.2 Discussions About Family Life Education Topics with Parents . 231 15.3 Family Life Education Instruction in School . 233 15.4 Sources of Information on Sexual Matters . 238 15.5 Impact on Knowledge About Fertility Issues and Contraception . 238 Chapter 16 SEXUAL AND CONTRACEPTIVE EXPERIENCE OF YOUNG ADULTS . . 241 (Afua Appiah-Yeboah, Leo Morris, Shafag Rahimova) 16.1 First Sexual Intercourse. 241 16.2 Current Sexual Activity. 246 16.3 Opinions and Attitudes About Condoms and Condom Use . 249 16.4 Regional Comparisons . 253 Chapter 17 KNOWLEDGE AND EXPERIENCE OF SEXUALLY TRANSMITTED INFECTIONS. 255 (Florina Serbanescu, Shafag Rahimova, Lisa Flowers) 17.1 Awareness of STIs and Knowledge of STI Symptoms . 256 17.2 Most Important Source of Information and Mass Media Messages About STIs . 260 17.3 Self-Reported STI Testing and Diagnostic . 264 17.4 Self-Reported STI Symptoms . 266 17.5 Perceived Risk of STIs. 268 Chapter 18 KNOWLEDGE OF AIDS TRANSMISSION AND PREVENTION . 271 (Ranee Seither, Florina Serbanescu, Leo Morris) 18.1 Knowledge of HIV/AIDS . 272 18.2 Knowledge of HIV/AIDS Transmission . 274 18.3 Knowledge of HIV/AIDS Prevention . 280 18.4 Beliefs about the Risk of HIV/AIDS and Self-Perceived Risk of HIV/AIDS .284 Chapter 19 PHYSICAL AND SEXUAL ABUSE. 289 (FLorina Serbanescu, Shafag Rahimova, Leo Morris) 19.1 Comparative Findings on Intimate Partner Violence in Eastern Europe . . . 290 19.2 History of Witnessing or Experiencing Parental Physical Abuse . 291 19.3 Verbal, Physical and Sexual Abuse by a Partner or Ex-Partner. 293 19.4 Discussions of Physical Abuse with Others . 298 19.5 Prevalence of Physical Abuse . 300 REFERENCES . 303 GLOSSARY . . 313 ANNEX A: SAMPLING ERROR ESTIMATES . Al ANNEX B: INSTITUTIONS AND PERSONS INVOLVED IN AZRHS01 . B1 ANNEX Q: SURVEY QUESTIONNAIRE. Q1 Preface During the 10 years since regaining its independence, Azerbaijan has faced considerable difficulties resolving some of the problems that linger from the previous system and grappling with the challenges of the transition period. The war—which caused the largest refugee and internally displaced population in the region— and ethnic, social, and economic problems have hindered reconstruction of governmental and administrative systems, including the health sector. Reforms in public health care first require improvement of the health information system, particularly the statistics concerning maternal and child health. Population-based nationwide health surveys serve as a significant source of information in this field and assist in clarifying several health-related issues not covered by the official statistics. This report reflects preliminary results of the first nationwide reproductive health survey conducted among Azeri women. The research was funded by the U.S. Agency for International Development, the United Nations Population Fund, and the United Nations High Commissioner for Refugees, through Mercy Corps, and was carried out by the Adventist Development and Relief Agency Azerbaijan with the technical assistance of the U.S. Centers for Disease Control and Prevention. The research provides data on the reproductive health behavior of Azeri women as well as information on their fertility, planning status of pregnancies, abortions, use of women's health services, contraceptive knowledge and attitudes, and knowledge about STI and AIDS transmission and prevention. The survey results assist in revealing high-risk groups and provide a focus for future reproductive health programs. I believe that the results of the survey will be useful for monitoring and evaluating the current reproductive health and family planning programs being implemented in the country as well as for planning effective new studies and projects in this field. Academician Ali Insanov Minister of Health Azerbaijan Republic i ii Acknowledgments The 2001 Azerbaijan Reproductive Health Survey (AZRHS01) was conducted by the Adventist Development and Relief Agency (ADRA) Azerbaijan in collaboration with the Azerbaijan State Committee for Statistics and Mercy Corps. Technical assistance in survey design, sampling, questionnaire development, training, data processing, and report writing was provided by the Division of Reproductive Health of the United States Centers for Disease Control and Prevention (DRH/CDC). Principal investigators of the study were Shafag Rahimova M.D., national director of the AZRHS01 (for ADRA), and Florina Serbanescu M.D. and Leo Morris Ph.D., of DRH/CDC. Most of the funding for the AZRHS01 was provided by the U.S. Agency for International Development (USAID PASA DPE-3038-X-HC-1015-00), the United Nations Population Fund (UNFPA), and the United Nations High Commissioner for Refugees (UNHCR). We wish to thank the 7,668 women who made such a major contribution to our knowledge of women's and children health in Azerbaijan through their participation in the AZRHS01. We thank our dedicated interviewers, supervisors, and field work coordinators, Saida Ismaylova and Mahbuba Khalilova, for their commitment, dedication, and discipline during the survey data collection. This project could not have been completed without the collaborative efforts of the ADRA's survey headquarters team—Shafag Rahimova, survey director; Farid Agamaliyev, survey manager; Linda Fardy Hayes, survey consultant, Tamilla Rashidova, data entry supervisor, Gushan Karimova, secretary—ADRA's personnel—Wagner Kuhn,country director, Conrad Vine and Teymur Musayev, health coordinators, Mark Castellino, programs officer, and Kirill Kravchenko, director of finance—Mercy Corps team—William R. Holbrook, chief of party, Craig Redmond, program director, Jamila Kerimova and Javanshir Hajiyev, program Officers, and Muhammed Amer Mir, director of finance—and the panel of experts of the Azerbaijan Ministry of Health headed by Alexander Umnyashkin, adviser to the Minister of Health, and Oktay V. Akhundov, head of the information and statistics bureau. Special thanks are also extended to the USAID staff in Azerbaijan—William D. McKinney, country coordinator, Khalid H. Khan, CTO, Catherine Fischer, regional health specialist, and Gulnara Rahimova, project development assistant. We also thank the UN representatives— Ramiz Alekperov, UNFPA national program officer, and William Brady, UNHCR reproductive health coordinator—for their assistance in design, planning, and financial management. Many thanks to Mary Ann Micka, Mary Jo Lazear, and Willa Pressman, USAID/Washington, for their continued support of the survey. iii iv EXECUTIVE SUMMARY With the dissolution of the Soviet Union at the beginning of the 1990s, Azerbaijan regained independence and moved toward a democratic society with a free market economy. While dealing with the difficulties of developing a new functional autonomous government, the nation also faced war with its neighbor and former Soviet Republic, Armenia, over the region of Nagorno-Karabakh. A large number of Azeris have been displaced by Armenia's ongoing occupation of 20% of the country. Azerbaijan has had to deal with the health impact of war, displacement, incoming refugees, and economic disruption at the same time that it has lost the resources of the large Soviet health system. The health of women has suffered under these recent circumstances. The 2001 Azerbaijan Reproductive Health Survey (AZRHS01), the first population-based national survey of its kind conducted in Azerbaijan, documented significantly poorer reproductive health indicators than those in other countries of Eastern Europe and the former Soviet Union. The AZRHS01, conducted by the Adventist Development and Relief Agency (ADRA), Baku, with technical assistance from the Division of Reproductive Health, Centers for Disease Control and Prevention, Atlanta (DRH/CDC), employed a national probability sample of 7,668 women aged 15-44, including an oversample of conflict-affected areas with larger concentrations of internally displaced persons and refugees (IDP/Rs). The response rate was 93%. By collecting information from the general population as well as from those who are internally displaced, survey data can document specific needs associated with displacement, account for differences in reproductive health status between the two populations, and provide a useful tool for evaluating reproductive health programs and activities that specifically address displaced women and children. The survey was designed to collect reproductive health information from a representative sample of reproductive-age women throughout Azerbaijan. The questionnaire covered a wide variety of topics related to reproductive health for all women regardless of marital status and included questions on family life education and sexual behavior for women aged 15-24 as well as questions on the nutritional status of women and their children. Almost three out of five women (59%) who completed interviews were married or in a consensual union. About a quarter of the women had more than a secondary education, and just below 25% had not completed secondary school. Ninety percent of the survey population was Azeri; 4% were Talish; and the remainder were of other ethnic groups, including Lezgi, Tat, Avar and Russian. Azeri was the language spoken in 90% of households, followed by Russian (4%), Talish, (3%), and others (4%). Average household size was 4.9 persons. Ninety-one percent of households owned a television v set, but only 38% had a flush toilet, and a little more than one-third (35%) had electricity 24 hours a day. Marriage and Fertility Survey results show that Azeri women marry in their early 20s (the median age at first union was 22.3 years), report having first sexual intercourse at marriage, and have their first child soon after marriage (median age at first birth was 23.7 years). The age at first union is falling, a situation that has potential implications for future fertility patterns. As with women in other countries of the region, Azeri women initiate and complete childbearing at an early age and have a desired family size of two children. The highest fertility levels were among 20- to 24-year-old women and 25- to 29-year-old women, accounting for 36% and 32%, respectively of the total fertility rate. Compared with its Caucasus neighbors, Azerbaijan exhibits the lowest adolescent (age 15-19) fertility but the highest fertility rates for 20- to 24-year-old and 25- to 29-year-old women. Pregnancy Intention Status Fewer than half of the women who had been pregnant in the past 5 years (42%) reported that their most recent pregnancy was intended at the time of conception; 9% reported it as mistimed (i.e., wanted it at a later time); and 48% reported it as unwanted. Thus, 57% of women reported that their last pregnancy was unintended, and the majority of those (84%) reported it as unwanted rather than mistimed. The AZRHS01 found that the total induced abortion rate was 3.2 abortions per woman, 1.5 times the total fertility rate of 2.1 births per woman. Although abortions are legal in Azerbaijan and most abortions reported in the survey were performed in clinical settings, a substantial proportion are not reflected in the official statistics. The decreased ability of the reproductive health system to document the true magnitude of abortion levels is probably due to the underreporting of abortions performed in the private sector, inherent problems related to data registration in state-run medical facilities, and the persistence of abortion performed outside clinical settings. More than one in three women (35%) reported having had at least one abortion. Among those, 65% have had more than one abortion. Almost all abortions in the past 5 years were performed in a hospital or government clinic (70% and 26%, respectively); 3% took place outside of a medical facility, and only 1% were performed in a private clinic. Forty percent of abortions were performed in the first 6 weeks of gestation by vacuum aspiration (i.e., mini-abortions). One in five women reported early or late complications as a result of their abortions. Among married women (legal or consensual marriages), 70% reported that they did not want to have vi any more children, including 77% of those with two living children, 88% of those with three living children, and 93% of those with four or more living children. Prenatal Care and Breast-Feeding Just over two-thirds of Azeri women (70%) who gave birth in the past 5 years had received some prenatal care; of those, about two-thirds initiated their prenatal care in the first trimester. Pregnant women who received any prenatal care averaged 4.3 prenatal visits. According to the adequacy of Prenatal Care Index (Kotelchuck Index). Only 6% of births in the previous 5 years had received adequate care. The principal source of prenatal care was a women's consultation clinic (46%), followed by maternity hospital (36%) or a village hospital (13%). Two-thirds of women who attended prenatal clinics received some counseling about nutrition during pregnancy; about half of the women surveyed received information about breast-feeding (55%), delivery (54%), early signs of pregnancy complications (49%), and postnatal care (47%), but only two in five (38%) were counseled on the negative effects of smoking or alcohol. The majority of women with births in the past 5 years reported having their blood pressure measured as a part of routine prenatal care (82%); of those, one-fifth (19%) were identified as having high blood pressure. One-fourth of women reported pregnancy complications that required medical attention, including 1% who required hospitalization for these complications. Three-fourths (76%) of births took place in maternities or hospitals; however, one in four births in the past 5 years occurred outside of a medical facility. The prevalence of Cesarean deliveries reported between 1996 and 2001 was only 3%. Postpartum care was substantially less utilized than prenatal care (25% vs. 70%). Almost two fifths (38%) of births in the past 5 years resulted in at least one postpartum complication. The total stillbirth rate was 21 per 1,000 births. Twelve percent of all live births weighed less than 2,500 g at birth, and 5% of births took place before the 37th week of gestation. Almost all children born in the past 5 years were breast-fed (95%). The mean duration of breast-feeding was 11.6 months, but full breast-feeding averaged only 3.3 months, including exclusive breast-feeding, which lasted less than 1 month, on average. Nutrition Data from the 1996 National Health and Nutrition Survey of the Internally Displaced and Resident Population of Azerbaijan allow comparison of the change in prevalence of chronic and acute malnutrition in young Azeri children between 1996 and 2001. The nutritional status of the IDP/R population was similar to that of the non-IDP/R population on all the indicators of the survey. The prevalence of low height for age, or stunting, which is considered evidence of chronic malnutrition, vii was 13% among children under 5 years old. In general, the problem of low weight-for-height, an indicator of acute malnutrition, was no greater among children under age 5 than expected, although the rate among children under age 2 was more than 3 times greater than the rate for children aged 2-5 (4% vs. 1%). Low weight-for-age, an indicator of general malnutrition, was found in 7% of children aged 3-59 months. This value reflects the greater prevalence of chronic malnutrition because the rate of acute malnutrition is much lower. Hemoglobin samples revealed a 32% rate of anemia among children aged 12-59 months. Among mothers of children aged 3-59 months, overweight was more common than underweight (38% and 6%, respectively). The prevalence of anemia was 40% among mothers of children aged 3-59 months, and the prevalence among pregnant mothers was similar (38%). Infant Mortality Infant mortality rates were directly calculated from responses to survey questions on the following topics for each live birth: the date of occurrence, sex of the child, survival status and, for children who had died, the age at death. Survey data were used to calculate mortality levels among respondents' children in the following categories: infant mortality (deaths before the first birthday), child mortality (deaths between 12 and 59 completed months of age), and child-under-5 mortality (i.e., deaths before the 5th birthday). Infant mortality was further divided into two periods: neonatal (0-27 days) and postneonatal (28-364 days). The total infant mortality rate was 81 deaths per 1,000 live births. Neonatal and postneonatal mortality were 38 and 43 deaths per 1,000 live births. Child mortality was calculated as 11 deaths per 1,000 live births. The mortality rate for children under age 5 was found to be 92 per 1,000 live births. Contraceptive Awareness and Use Azeri women demonstrated a relatively high level of awareness of the existence of some family planning methods. Eighty-seven percent of Azeri women had heard about at least one contraceptive method, generally the IUD, condoms, or the pill (83%, 58%, and 53%, respectively). On average, they recognized fewer than three modern methods. For the most widely known modern contraceptive methods, however, respondents had a serious gap between their awareness of a method and knowledge of how that method is used; the gap ranged from 17 percentage points for condoms to 32 percentage points for the IUD. A gap of similar magnitude was obvious between respondents' awareness of a contraception method and knowledge of where the contraceptive could be obtained, ranging from 14 percentage points for the condom to 24 percentage points for the IUD. Correct knowledge about the effectiveness of modern methods was also generally lacking. No modern viii method was recognized as being very effective by a majority of women, partly because a substantial number of women lacked knowledge about how modern methods are used. Even when women who had never heard of a specific method were excluded, no method with high effectiveness (e.g., tubal ligation or IUD) was correctly recognized as highly effective by a majority. The first source of information about contraception was a friend or colleague (40%), followed by a relative other than a parent (19%), a physician (16%), a partner or boyfriend (11%), and the mass media (4% audiovisual media, 3% print media, and 3% books). These findings explain, in part, the poor quality of contraceptive information, which is often acquired through rumors, and argue for increasing the public health efforts to educate women about the benefits of contraception through official channels (school, mass media, and health providers). Nearly three out of four women stated that they want more information about contraception (including 85% of women aged 20-24 years). Of those women, 67% said that a gynecologist would be the most reliable source of information and 10% said that mass media would be the most reliable source of information. Contraceptive prevalence among Azeri women in union is among the lowest of that reported by any former Soviet republics having survey data. Only 55% of married women reported using any method of contraception in the month preceding the interview, and less than a quarter of those women used a modern method, mainly the IUD (6%). Just 1% of married women reported that they have been surgically sterilized. Although more than 90% of women with three or more children did not want any more children, only 7% had had a surgical sterilization. The public medical sector was generally the largest source of modern contraceptive methods in Azerbaijan (54%), followed by commercial outlets (35%). Private clinics and NGOs are an emerging source of oral contraceptives. Most women using non-supplied methods (withdrawal and the rhythm method) stated that the major factors influencing their decision not to use a modern method were fear of side effects (90%), lack of or little knowledge (71%), cost (61%) or lack of access (53%), and partner preference for traditional methods (49%). More than two-thirds of women using traditional methods consider their method of contraception to be of equal or greater effectiveness than modern methods; only 25% recognized that the IUD or the pill are more effective methods in preventing pregnancy than the method they currently use. Conversely, the one-year failure rates for withdrawal and calendar method users were 26% and 30% respectively, compared with 21% for condom users, 15% for pill users, and just 1% for IUD users. In addition to higher than average method-specific failure rates (excluding users of IUD), the survey data showed considerably high discontinuation rates: 44% to 58% for traditional methods, 67% for condom, and 82% for pill users. A total of 53% of married women were estimated to have an unmet need for modern contraceptive methods, the highest proportion among Eastern European and former Soviet Union countries. Most of these women (84%) need methods ix to help them effectively limit fertility while fewer need contraception to postpone childbearing. About two of every three women who have used a modern method in the past 5 years were advised by a health care provider to use the current or most recent method, but only 40% of them received general information about other methods and only one-third were told about the method's effectiveness. Like other countries in the region, only a minority of Azeri women received family planning counseling (32%) or were offered contraceptive supplies (2%) pre- or postabortion. Only 27% of women who gave birth in the past 5 years received information during prenatal care about family planning after birth, and just 34% received such information as part of postpartum care. Young Adults In stark contrast to other Eastern European countries, virtually all sexually experienced young adult women had their first sexual experience after marriage (95%); only 1% of all young adult women reported any premarital sex. Almost none reported using any contraception at first intercourse; 85% cited a desire to become pregnant as their main reason for not using contraception. Use of contraception at most recent intercourse was reported by about one-third (36%) of sexually experienced young adults (women aged 15-24 years). Of those not using contraception, three out of five (62%) were either already pregnant (37%) or wanted to become pregnant (25%). Almost all (98%) sexually experienced young women reported only one sex partner in their lifetime. Almost no sexually experienced young women had ever used condoms. Only one in five sexually experienced young women had ever discussed condom use with a partner. Nearly half of women (47%) stated that they would feel protected against pregnancy if a partner suggested using a condom, and more than one-third (37%) would feel protected against sexually transmitted infections (STIs), but about one in five (18%—22%) reported negative feelings about such a suggestion. Women's Health Only about one in two (57%) sexually experienced women had ever been examined by a gynecologist during a routine exam, including 22% who had been examined in the previous 12 months, and 21% who had their last exam more than 3 years ago. Just 2% of sexually experienced women had ever had a pap smear, and less than 1% had had their most recent test in the past 3 years. Fewer than one in three sexually experienced women (30%) had ever heard of breast self- examination, and only 10% had ever performed one. More than one-third of all women had been told by a physician that they had anemia, and more than half of this group (57%) had anemia occurring outside of pregnancy. A quarter of all women (27%) and 42% of currently married women reported that they had been diagnosed with pelvic inflammatory disease. Because extramarital sexual relations x are relatively uncommon among Azeri women, possible explanations for the high PID rates include: poor hygienic conditions and inadequate standards of care in abortion facilities that may increase the risk of postabortion infections, lack of routine gynecologic visits, and lack or inadequate treatment of STIs, possible acquired from an unfaithful partner. Additionally, one-third (32%) of sexually experienced women reported abnormal vaginal discharge and 2% reported genital sores or ulcers in the past 12 months. Attendance at infertility clinics was reported by 12% of women in union; 7% reported current fecundity impairment. Cigarette smoking was almost nonexistent among Azeri women. Family Life Education Three in four Azeri women of reproductive age supported family life education in school, and a majority believed that school-based courses on reproductive biology ("how pregnancies occur"), contraception, and STIs should start by age 16 (73%, 62%, and 60%, respectively). Less than 20% of women favored school-based education on those topics before age 14. A total of 56% of young adult women had talked about at least one sex education topic with a parent before age 18; however, only about one in five young women (22%) talked to a parent about abstinence, one in ten about how pregnancies occur, less than 6% discussed HIV/AIDS or other STIs, and just 4% talked about contraception. Only 40% of young women had had at least one school-based course or class on family life education before age 18, and few had had courses related to HIV/AIDS (7%), other STIs (3%), or contraceptive methods (2%). Survey results indicate that the quality of teaching of family life education should be improved: only 8% of young adults knew the time during the menstrual cycle when conception is most likely to occur, one in four (24%) were aware that decreases the risk of pregnancy, and 58% knew that a women could become pregnant at first intercourse. Most young adults either did not know or had misinformation about the effectiveness of most methods of contraception. HIV/AIDS and Other STIs Most (74%) Azeri women have heard of HIV/AIDS, but only one in five women (21%) knew that an HIV/AIDS infection could be asymptomatic; just 1% knew that HIV can be asymptomatic and cannot be spread by kissing or by medical or dental treatment (UNAIDS Knowledge Indicator 2). Two in five women identified both monogamy and condom use as prevention measures (UNAIDS Knowledge Indicator 1). Although awareness of HIV/AIDS was high, a lower proportion of women had heard of syphilis (41%), gonorrhea (35%), trichomonas (12%), chlamydia (8%), bacterial vaginosis (7%), genital warts (6%) or genital herpes (5%). Almost two-thirds (62%) of respondents xi mentioned mass media as the most important source of information about STIs, but recent mass media messages were more often reported to be about HIV/AIDS, whereas messages on other STIs were considerably less common (54% vs. 13%). Friends and peers were the next most important source of information (14%). The National Working Group on STI Management and Guidelines, established in 1998 to increase STI knowledge among Azeri men and women of reproductive age, requires close collaboration between public health organizations, NGOs, and audiovisual media. In designing such educational campaigns, the working group needs to ensure that no misconceptions or needless threats are disseminated, because media imagery may be difficult to offset. Some groups of women, particularly those who are less educated or are younger than age 25, may require specialized interventions. Physical and Sexual Abuse Lifetime experience of spousal physical abuse was reported by 20% of Azeri women, comparable to reproductive health survey results in most other former Soviet bloc countries. One in four women (26%) reported witnessing abuse between parents as a child, and one in three reported experiencing abuse at the hands of a parent. Almost one in three (30%) women reported that they had been verbally abused by a partner or ex-partner, one in five reported some form of physical abuse, and one in ten reported spousal sexual abuse (i.e., being forced to have intercourse against their will). Only 1% of physically abused women reported the violence to police or discussed it with a health care provider, and even fewer sought legal counsel for recent domestic abuse. Most women failing to report domestic violence cited reasons related to social perception, such as bringing the family a bad reputation (48%), or personal embarrassment (13%). Six percent of women stated that they had been forced to have sexual intercourse; in 95% of those cases, the perpetrator was the woman's husband, consensual partner, or boyfriend. Conclusion The AZRHS01 showed that the women of Azerbaijan have inadequate knowledge of and access to diverse contraceptive methods. Doctors and nurses need preservice and inservice education and training in contraceptive technology as well as in contraceptive counseling skills. The public needs appropriate education through social marketing and family life education in schools. Researchers in the United States have developed an "informed-choice" strategy for people to make contraceptive decisions on the basis of well-informed choices about family planning and protection against HIV/AIDS and other STIs. This strategy targets five areas: government policies, xii communication programs, access to contraception, family planning program management and leadership, and counseling (Upadhyat U et al., 2001). Informed choice means that individuals and couples can make their own personal decisions on spacing and limiting children when given accurate information along with access to services and supplies to carry out their decisions. This principle has long been fundamental to family planning programs around the world; unfortunately, its implementation has been uneven. Azeri culture supports monogamy and delay of sexual intercourse until a woman's marriage to an extent rarely found in other countries, even those of the former Soviet Union. These behaviors that have clear public health benefits and should be encouraged. However, given the difficulties in obtaining valid data from young people about their sexual behaviors, particularly when sexuality is a taboo topic (which seems to be the case in Azerbaijan), survey results on this subject should be interpreted with caution. In Azerbaijan, fertility control has been predominately achieved through the practice of induced abortion. Abortion complications and their treatment burden an already struggling health system. Postabortion care activities, including emergency obstetric care, family planning counseling and services, and appropriate referral for other health care needs (such as those related to nutrition or violence), would seem a particularly useful way to prevent recurrent abortions and redirect funds toward preventive activities. The national family planning program in Azerbaijan, a collaboration between the Azerbaijan Ministry of Health, the United Nations Population Fund, and several international and local NGOs, is in the early stages of development. It will take time to set official policies and budgets for programs throughout the country. Areas of emphasis—those requiring immediate attention—should be based on population-based data, such as that in this survey or other related surveys. To accurately judge continued needs and progress, public health officials will need more complete vital statistics, which require improvements in the system of immediate registration of births, deaths, abortions, immunizations, and other health events. A national policy on family life curriculum could reach most young people in the country because almost everyone attends at least some secondary school. Family life education courses promoting delayed initiation of first intercourse and knowledge of human sexuality, contraception, and disease prevention have been shown to promote increased use of contraceptives, which protect against unintended pregnancy and STIs. Because Azeri schools currently lack such comprehensive education, a clear statement of government policy and dedication of resources will be required to bring it into existence. xiii Accurate information campaigns in the mass media and in the community can inform people of their right to make their own decisions, explain their options, and direct them to appropriate health care providers. Such campaigns should include information not only about HIV/AIDS but also about other STIs, such as syphilis, whose prevalence in Azerbaijan is much higher than that of HIV/AIDS and therefore poses a greater risk for Azeri men and women. Improving access to reproductive health care should include improving the availability of a range of contraceptive methods and provision of a network of women's health clinics. The entire spectrum of women's health issues should be addressed because they ultimately affect reproductive health and the health of the nation's children. Program management that improves the quality of care enhances clients' choices and improves contraceptive efficacy and continuation rates. Family planning programs must take into consideration that one or two methods of contraception will not be appropriate for every woman. Programs that offer a variety of methods, adequate information and treatment as needed empower women to use contraception correctly, consistently and to their best advantage. Counseling can be crucial to helping couples think through their decisions. Couples who are unaware of the variety of family planning methods that are available may incorrectly assume that contraception is not for them. A knowledge of costs, effectiveness in preventing pregnancy and STIs, proper usage, and side effects are all necessary for couples to choose methods which meet their needs. Evaluation is an important component of education and program efforts. Future population-based surveys and monitoring of vital statistics will allow public health planners to examine progress toward achieving reproductive health program objectives in terms of knowledge, attitudes, and practices of Azeri women. They will also help measure progress toward broad goals of reducing mortality and morbidity among children and reproductive-age women. Plans can then be adjusted on the basis of an understanding of which programs have been effective and which objectives require enhanced or alternative efforts. Once baseline data are available, continual implementation of needs assessment, policy setting, identification of targets, program planning, and objective evaluation can be used to achieve steady improvements in reproductive health. xiv CHAPTER 1 INTRODUCTION With the dissolution of the Soviet Union in 1991, Azerbaijan again became an independent country. A nation of about 86,600 km2 in the Caucasus region, it borders on Iran to the south, Armenia to the west, Georgia and Russia to the north, and the Caspian Sea to the east. The Great Caucasus Mountain range passes though the northeastern part of the country. Azerbaijan has only 0.2 hectares of arable land per capita. The industrialized Absheron Peninsula suffers from pollution of its air, soil, and water, including the Caspian Sea. The Azerbaijan State Committee on Ecology has classified the city of Sumgait, which is on the Peninsula, an ecological disaster area, a legacy of its years as a major industrial center of the Soviet Union. Approximately one-third of the population inhabits the Absheron area, including 1.8 million in Baku, the capital of Azerbaijan. Outside Baku (which consists of 11 administrative units or rayons), the country is divided into 65 administrative units (59 rayons and 6 cities) and the autonomous republic of Nakhchivan (consisting of 6 rayons and the capital, Nakhchivan city) (State Committee of Statistics of the Azerbaijan Republic [SCS], 1996; SCS, 2001)'. Approximately 51% of the 8.1 million people of Azerbaijan live in urban areas. About 30% of the total population is younger than age 15, and 6% are older than age 65. Women make up 51% of the population (SCS, 2001). The life expectancy at birth has been increasing recently; in 2000 it was 75.1 years for women but only 68.6 years for men (SCS, 2001). Azeri, the predominant ethnic group, represented 83% of the total population at the 1989 Census (Goskomstat, 1989; Ministry of Health [MOH] and SCS, 2001). Before 1991, Russians (6%), Armenians (6%), and Lezgis (2%) were the largest ethnic minorities; most Armenians were clustered in the Nagorno-Karabakh region. After the disintegration of the Soviet Union and the war with Armenia, the ethnic composition was substantially affected by external migration and an influx of Azeri refugees from Armenia. Currently, Azeris represent more than 90% of the total population (SCS, 2001). Eighty-nine percent of the population speak Azeri, a language of Turkic origin; 3% speak Russian, 1 For enumeration purposes, these administrative units were grouped into 10 regions according to their contiguous geographical location. See Table 1.1 at the end of the chapter includes the listing of rayons in each region, as defined by their enumeration statistical code. 1 2% speak Armenian, and 6% speak other languages (United Nations Population Fund [UNFPA], 1999b). In pre-Soviet times the Azeri language was written with the Arabic alphabet. The Soviet policy forced a change to Latin (in 1926) and later to Cyrillic lettering (in 1940) but the written language has recently returned to its Latin form (Gurbanov, 1967; Elliot, 1999). Literacy is estimated to be about 97% for the population over age 14, although rates are somewhat higher for men than for women (Central Intelligence Agency [CIA], 2001). More than 93% of the population identify themselves as Muslim, 2.5% as Russian Orthodox, and 2.3% as Armenian Orthodox (MOH and SCS, 2001). Most people report that they do not practice their religion, and both the culture and the government are secular, not theocratic. Upon the collapse of the U.S.S.R. in 1991, Armenia and Azerbaijan, both former republics of the Soviet Union, engaged in a protracted war over Nagorno-Karabakh, the predominately Armenian- populated region within the Azerbaijan territory. In 1994,, the two countries reached a cease-fire agreement, but Armenia still occupies about 20% of Azerbaijan. The fighting has left Azerbaijan with some 790,000 internally displaced persons and refugees (IDP/Rs)—570,000 Azeri ethnics displaced from Nagorno-Karabakh and the surrounding occupied territories and 220,000 Azeri residents who fled Armenia and relocated in Azerbaijan when the war started—about 10% of its entire population (United Nations High Commissioner for Refugees [UNHCR], 1999). They constitute the largest proportion of IDP/Rs concentrated in one country of the Caucasus region. It is estimated that about two-thirds of IDP/Rs live in improvised housing conditions (e.g., public buildings, shelters, railroad wagons, mudhouses, dugouts, and tents) (United Nations Development Programme [UNDP], 1999). The constitution of Azerbaijan, ratified in 1995, established the government as a democratic republic. Citizens age 18 and older are eligible to vote and presidential elections are held every 5 years. The president appoints the prime minister and the Council of Ministers, contingent upon approval by the National Assembly. The National Assembly is elected every 5 years, by a combination of direct and proportional representation. The country's president appoints both the president and the vice-president of the Constitutional Court, which has the right of judicial review over the legislation of the National Assembly and presidential decrees. More than 20 political parties are active in Azerbaijan (UNDP, 1999). The shift from the command economy of the former Soviet Union to a new market economy has been a challenge for Azerbaijan. Three-quarters of Azerbaijan's exports are oil and gas; the rest consist mainly of machinery, cotton, and food products. Private sector employment increased to 56% in 1997, accounting for 46% of the gross domestic product (GDP). During the war with Armenia, inflation reached astronomical rates of growth—up to 1,664% annually in 1994—but it fell to an estimated 1.8% in 1999 (UNFPA. 1999a). 2 The transition to a market economy has had a negative impact on the welfare of the population. The unemployment rate is approximately 20%. The per capita GDP is just $537. In 1995, 68% of the population were classified as poor and 24% as very poor (MOH and SCS, 2001; World Bank, 1997). Food expenditures absorb an increasing percentage of the average family income (about 70% in 1997), yet consumption has fallen well below that needed to maintain health, especially consumption of meat, fish, and dairy products (UNDP, 1999). According to a recent United Nations Children's Fund (UNICEF) study, 17% of children younger than age 5 are underweight, about 20% are stunted, and 8% suffer from wasting (UNICEF, 2000). During the Soviet Union years, the Central Ministry of Health in Moscow oversaw the Azerbaijan MOH. The Soviet Union handed down all plans and standards that the Republic should meet, even information sheets for patient education. The emphasis was on medical treatment and funds were allocated according to the number of hospital beds. The Soviet system provided various kinds of health care facilities. Hospitals existed at several levels: small rural hospitals; rayon, or district, hospitals for more severe cases; and republic-level hospitals for the most seriously ill and injured. Pediatric and adult polyclinics staffed with specialists served urban areas. Businesses with numerous employees sometimes had their own polyclinics. Rural areas also had polyclinics but without specialized physicians. Feldsher-ackucher points (FAPs) also served as the primary level of care for residents of the most rural regions. Educational institutions placed greater emphasis on producing physicians rather than nurses, physician assistants, or health practitioners from other disciplines (UNFPA, 1999b). All the former Soviet-bloc countries have inherited the same centralized, government-supported health system (Semashko model), well known for its relative inefficiency in terms of structure, management, and resource allocation and for unresponsiveness to patients' needs. Recently, the system has been particularly affected by the acute lack of resources characterizing all social sectors in the newly independent states. Common features of this system are a massive reliance on hospital- based health care services, which generates too many hospitals and hospital-based specialized physicians, and an inadequate supply of primary health care services. In the transition to a market economy, the costly hospital-based curative system became impossible to maintain; most hospitals lacked minimum equipment, drugs, and supplies and could not afford the maintenance costs. In most former Soviet Union countries, health care deteriorated rapidly, particularly in the area of reproductive health services as reflected in the worsening of several outcome indicators (e.g., maternal and infant mortality, sexually transmitted infection [STI] prevalence, and utilization of preventive services). In many countries of the region, health reforms are currently in various stages of development but are hampered by limited resources, thus leaving many segments of the population uninsured or with minimum health benefits. 3 In Azerbaijan, the government health care reform efforts are ongoing, but challenged by the recent economic problems and territorial disputes. Although Moscow no longer has a role in health operations or planning, the organizational skeleton it created remains. During the Soviet era, health care was free. Today, even though physicians are still employed by the state, people generally end up paying for health care services out of already strained household budgets. Even though patients are required to pay for care that was previously free of charge, the physical infrastructure of the health care system has been allowed to deteriorate and equipment has gone without necessary repairs or has become obsolete. Physicians are unlikely to have the most up-to-date information and skills and pharmaceutical dispensaries lack necessary medicines. It has been 10 years since Azerbaijan gained independence, yet copies of old Soviet health education pamphlets are still distributed. In addition to MOH-operated clinics, a few private providers of health care exist, as do clinics operated by international organizations that reach out to displaced populations (UNFPA, 1999b). After the collapse of the Soviet Union, GDP and wages in Azerbaijan declined. Government expenditures for health in 1999 represented 1.6% of the GDP, down almost 50% from the 1990 level of 2.9% (World Health Organization [WHO], European Public Health Information Network for Eastern Europe [EUPHIN/EAST], 2000). In reality, the decline has been even more dramatic, because as the total GDP declined, the population increased substantially in the aftermath of the Nagorno-Karabakh war with Armenia (Bladen et al., 1998). The real governmental expenditures on health are estimated to be less than one-quarter of the pre-independence level, amounting to only about US$7.00 per capita (1997 dollars). Most of the care is provided on a fee-for-service basis, with patients paying the largest share of costs; however, some categories, including pregnant and post- partum women, continue to receive free health coverage (Public Health Protection Law 360-IQ, 1997; Presidential Decree 62, 1997). According to World Bank estimates, the out-of-pocket costs for health accounted for more than 80% of spending in the health sector in 1995 (World Bank, 1997). Even when out-of-pocket costs are factored in, the health expenditures per capita (the sum of public and private expenditures on health divided by the country's population) amount to only US$36.00 per capita, lower than in Georgia and Kazakhstan (US$46.00 and US$86.00 per capita) but higher than in Armenia (US$27.00 per capita) (2001 dollars) (World Bank, 2001a). During the 1990s, Azerbaijan experienced a steep increase in maternal mortality (UNFPA, 1999; SCS, 2001). However, after an initial surge in the maternal mortality rate (MMR) between 1991 and 1994, culminating in 43.8 deaths per 100,000 live births in 1994 (a rate almost 5 times higher than the 1990 level), MMR stabilized in the 1995-1999 period and started to decline in 2000. The official estimate of maternal mortality for 2000 was 37.6 deaths per 100,000 live births, almost 4 times higher than the 1991 level of 10.4 deaths per 100,000 live births (MOH and SCS, 2001). According to the official statistics, about 12% of the maternal deaths in the past 3 years were due to abortion, including 1% due to "artificial medical abortion" (WHO-EUPHIN/EAST, 2000; MOH, 2001a). 4 However, a recent UNFPA country assessment report suggests that the actual MMR is considerably higher (UNFPA, 1999b). Several population-based studies conducted after 1990 documented that a substantial proportion of Azeri women deliver at home with the assistance of local midwives instead of government-employed physicians. The 1996 National Health and Nutrition Survey found that as many as one-third of all children younger than age 1 were born at home (Branca et al, 1996). Regional household cluster surveys of women with children under age 5 conducted in Northwest Azerbaijan in 1997 and 1999 documented that 25% of women had no prenatal care visits, and the proportion of newborns born at home increased from 37% in 1997 to 44% in 1999 (Buchholz, 1999). According to the Azerbaijan MOH, the proportion of women who delivered at home increased between 1990 and 2000 from 2.3% to 8.6% of the total labors "observed in establishments of the Ministry of Health" (MOH and SCS, 2001). Abortion is the most common form of birth control; three-fourths of sexually active women reported at least one lifetime induced abortion in a small area sample survey of Relief International clinics (Posner et al., 2001). Almost two-thirds of women reported recent symptoms suggestive of STIs, and the prevalence of pelvic inflammatory disease due to (STIs) is estimated to be high (Claeys et al., 2001; Kerimova et al, 2000). The fertility rate started to drop prior to 1990, but the decline during the past decade has been at a faster pace. From a level of 3.3 births per woman in 1980, the total fertility rate (TFR) decreased slowly to about 2.7 in the period 1981-1993, then fell abruptly to slightly below a replacement level of 2 births per woman in 1998 (MOH, 2001a). The absolute number of births decreased by 36% between 1990 and 2000 (from 182,989 births to 116,994 births) (SCS, 2001). Women typically marry and begin families at a young age. Most do not have premarital sexual relationships and births out-of-wedlock are rare—less than 5% of all births, according to the MOH (MOH and SCS, 2001). Unmarried women generally do not live on their own, no matter what their age, education, or professional status. Although the reported fertility rate has declined, the population is expected to continue to increase because of the great number of women of childbearing age who were born during the earlier periods of high fertility. The population is expected to stabilize at about 9.5 million in 2025. A net emigration from the country is taking place, mostly men from the larger cities going abroad to look for better economic opportunities. Infant mortality and under-5 mortality were officially reported as 12.8 deaths per 1,000 live births and 25.8 deaths per 1,000 live births, respectively, as of 2000 (MOH, 2001a; SCS, 2001). Both rates have declined considerably from those reported in the late 1950s and early 1960s. However, the official statistics may be plagued by problems such as the non- registration of births of infants who die shortly after birth and the misclassification of neonatal deaths and early deaths of premature infants as stillbirths. International agencies suggest that the actual 5 fertility rate is 2.2 births per woman (CIA, 2001) and that the mortality rates may be much higher. Mortality rates are reportedly higher for children in rural locations or with poor socioeconomic status (UNFPA, 1999b). These statistics should be interpreted with caution. The health information system during the Soviet times was often flawed by overreporting of "positive" results, which could bring rewards, and underreporting of undesired statistics, which could lead to disciplinary actions (Bladen et al., 1998). Even when the data collected were reliable, they usually satisfied the needs of "centralized" decision- making but were not always useful for describing the health status and the burden of disease on the population at subnational levels. Although the old system is no longer in place, some of its characteristics are likely to have been retained. In addition, with the emerging private health sector and the shifting of health costs from the state to the individual, official data are unlikely to be complete (Bladen et al., 1998). The UNFPA Country Population Assessment of 1999 identified the need to assess the health situation related to reproduction, including but not limited to maternal morbidity and mortality, abortion, prenatal and postpartum care, reproductive tract infections, STIs and HIV/AIDS, and contraceptive awareness and use, and it called for a comprehensive nationwide survey to collect the necessary data (UNFPA, 1999a). In addition to those topics, other recommended areas of study were popular beliefs about fertility, ideal age of marriage, pregnancy and birth intervals, how fertility decisions are made, and how much money women are willing and able to pay for their reproductive health care. In conclusion, Azerbaijan has undergone major socioeconomic and political changes: the war with Armenia, forced migration and population displacement, economic hardships, and deterioration of health and social services. These changes have affected practically all aspects of life for its people. The reported flaws associated with official statistics have prohibited any meaningful attempts at informed decision making, planning, and program evaluation in reproductive health. A nationwide survey was recommended to assess the reproductive health status of the population during this transition period, a period of profound changes in health needs and access to health care services. The national reproductive health survey conducted in Azerbaijan in 2001 (AZRHS01) is the first nationwide population-based survey aimed at providing a wide array of information about the current status of women's health in that country. The survey will aid in identifying unmet programmatic needs and will serve as a baseline for future studies and evaluations. The AZRHS01 was specifically designed to meet the following objectives: • To assess fertility, abortion, contraception, and various other reproductive health issues in Azerbaijan. 6 • To enable policy makers, program managers, and researchers to evaluate existing reproductive health programs and develop new strategies. • To study factors that affect fertility, contraceptive use, and maternal and infant health, such as geographic and sociodemographic factors, breast-feeding patterns, use of induced abortion, and availability of family planning services. • To identify characteristics of women at risk for unintended pregnancy. • To identify high-risk groups and focus additional reproductive health studies on them. • To obtain data on the knowledge, attitudes, and behavior of young adults 15-24 years of age. • To provide data on the level of reported STI symptoms and knowledge about transmission and prevention of AIDS. • To provide data on women living in prolonged displacement. Similar to the survey conducted in Georgia, completed in 2000, the AZRHS01 included an oversample of refugee women and women internally displaced by war and ethnic cleansing to document their specific health needs (Serbanescu et al, 2001). The disruption associated with living in improvised settings makes safe motherhood difficult, limits contraceptive access and use, increases the risks of HIV/AIDS and other STIs, neglects the special needs of adolescents, and may increase the risk of violence against women. Public health surveillance systems often exclude data collection and analysis essential to addressing the specific issues of IDP/Rs. To our knowledge, no country or organization has attempted parallel documentation of the reproductive health status of a nation and an internally displaced group within the country. By collecting information from the general population and from IDP/Rs, the AZRHS01 can document specific needs associated with displacement, account for differences in reproductive health status between the two populations, and provide a useful tool for evaluating existing reproductive health programs and activities that specifically address displaced women and children. The Division of Reproductive Health, U.S. Centers for Disease Control and Prevention (CDC), Atlanta, Georgia, coordinated survey activities and provided technical assistance to the local implementing agency, the Adventist Development and Relief Agency (ADRA), Baku, Azerbaijan. 7 Funding was provided by the United States Agency for International Development (USAID)—through the umbrella agreement managed by Mercy Corps (MC)—the United Nations Population Fund (UNFPA), and United Nations High Commissioner for Refugees (UNHCR). 8 CHAPTER 2 METHODOLOGY 2.1 Sampling Design The AZRHS01 is based on face-to-face interviews with 7,668 women at their homes. The survey was designed to collect information from a representative sample of women of reproductive age throughout Azerbaijan. The universe from which the respondents were selected included all females between the ages of 15 and 44 years, regardless of marital status, who were living in households in Azerbaijan when the survey was carried out (excluding the autonomous region of Nakhchivan and the occupied territories of Nagorno-Karabakh and surrounding areas). The questionnaire included information on each woman's education, employment, l iving arrangements, and other background characteristics as well as histories of marriage, divorce, cohabitation, sexual activity, pregnancy, and contraceptive use. Additional questions investigated health risk behaviors that may affect reproductive health (e.g., smoking and drinking habits), women's health screening practices, and intimate partner violence. The questionnaire was developed in English, translated into Azeri and Russian, and translated back to ensure accuracy and linguistic equivalency. The household survey used a stratified multistage sampling design using the recent 1999 census as the sampling frame (State Committee of Statistics of the Azerbaijan Republic [SCS], 2000). For the AZRHS01, the geographic area of the Azerbaijan Republic was divided into four independent sampling strata. The strata were created by grouping regions with a similar concentration of IDPs and refugees (IDP/Rs), as recorded by the United Nations High Commissioner for Refugees (UNHCR, 2000). The sample was selected with probability proportional to the population size (PPS) within each stratum. Stratum 1 included six rayons that each consisted of more than 30% of their population constituted by IDP/Rs: Fizuli (53%), Xanlar (51%), Barda (44%), Naftalan (40%), Aghjabedi (32%), and Bilasuvar (31%). Stratum 2 included five rayons in which the IDP/Rs represented 20%-30% of the population: Imishli (25%), Saatli (23%), Belagan (22%), Mingechevir (21%), and Terter (20%). Stratum 3 included only the Baku district, which also had a relatively high concentration of IDP/Rs (14%). Stratum 4 included all other rayons, except those in Nakhchivan and the occupied territories of Nagorno-Karabakh and surrounding areas. 9 Regions with high concentrations of IDP/Rs (Strata 1 and 2) were oversampled for programmatic reasons. The oversampling in regions heavily populated by IDP/Rs was needed to include enough displaced women in the sample to allow independent estimates of their reproductive health status. This technique illustrates how surveys may be designed and integrated in the development, monitoring, and evaluation of targeted reproductive health programs. The oversampling of IDP/Rs was specifically designed to assess the reproductive health status of these women and measure the impact of the Azerbaijan Humanitarian Assistance Project (AHAP) funded by USAID and various projects targeting the IDP population supported by UNHCR and UNFPA. These projects aim to reduce the reliance on induced abortion by increasing access to and availability of effective contraceptive methods and by reducing the prevalence of STDs through the promotion of healthy behaviors among women (e.g., routine gynecologic exams) and child survival activities. These projects encompass various interventions, such as the establishment of modern health clinics for women; training of health professionals; development of information, education, and communication messages; social marketing; and provision of high-quality contraceptive supplies. Figure 2.1 compares the distribution of households in the sample (shown with bars) and the distribution of households in the 1999 Census (line graph) by the four strata. Stratum 1, one of the 10 smallest of the strata according to the Census (areas with more than 30% IDP/Rs) actually contains the largest number of households in the sample (3,415 households). These 3,415 households represent 111,876 households in the Census, such that the sampling fraction for this stratum is 1 in 32.7 households. Similarly, Stratum 2, consisting of regions with 20%-30% IDP/Rs, contains 1,707 households corresponding to 102,466 households in the Census (sampling fraction is 1 in 60 households). In contrast, the sampling fraction for the stratum labeled "Other" is only 1 in 273.2 households. The first stage of the three-stage sample design was a selection of Census sectors with probability proportional to the number of households in each sector, after the sectors were grouped into four strata. This stage was accomplished by using a systematic sample with a random start in each stratum. During the first stage, 300 census sectors were selected and became primary sampling units (PSUs), as follows: Baku (80 PSUs), regions with more than 30% of the population being IDP/Rs (100 PSUs), regions with 20%-30% of the population being IDP/Rs (50 PSUs), and all other regions (70 PSUs). In the second stage of sampling, clusters of households were randomly selected in each census sector chosen in the first stage. The cluster size was based on the number of households required to obtain an average of 20 completed interviews per cluster. The total number of households in each cluster took into account estimates of unoccupied households, average number of women aged 15-44 per household, the interview of only one respondent per household, and an estimated response rate of 90% in urban areas and 92% in rural areas. Finally, in each of the households selected, one woman between age 15 and 44 was selected at random for interview (ifthere was more than one woman was in the household). Because only one woman was selected from each household containing women of reproductive age, all results have been weighted to compensate for the fact that some households included more than one eligible female respondent. Survey results were also weighted to adjust for oversampling of households in the regions with a high concentration of IDP/R population and the undersampling in regions in which less than 20% of the population consisted of IDP/Rs. Except for Table 2.3.1, all tables in this report present weighted results. The unweighted number of cases, used for variance estimation, is also shown in each table. Thus, the survey can be used to make national and subnational estimates because of the elaborate process used to weight the data—that is, to determine how many women in the population were represented by each woman in the sample. 2.2 Data Collection The interviews were performed by 30 female interviewers, who were specially trained in interview techniques, survey procedures, and questionnaire content before the beginning of fieldwork. 11 Interviewer training was managed by the Adventist Development and Relief Agency Azerbaijan (ADRA), with the involvement of Shafag Rahimova, survey director; Conrad Vine, health coordinator; Farid Agamaliyev, project manager; Linda Fardy Hayes, survey consultant; and the U.S. Centers for Disease Control and Prevention (CDC) team (Florina Serbanescu and Natalia Melnikova for the reproductive health component and Geraldine Perry for the nutrition component). Interviewer training took place at the Ministry of Health International Training and Service Center just before data collection began; it consisted of 1 week of classroom training in fieldwork procedures and proper administration of the questionnaire and 1 week of practical training in the field with close monitoring by the trainers. At the end of the training period, six teams were selected, each consisting of four female interviewers, one nutritionist, and one supervisor. ADRA staff managed the fieldwork with technical assistance from the Division of Reproductive Health of the CDC. Two fieldwork coordinators (Saida Ismaylova and Mahbuba Khalilova) supervised the fieldwork implementation. Fieldwork lasted from April through July 2001. Each team was assigned to visit a number of primary sampling units in all regions of the country and traveled by car throughout the country on planned itineraries. Interviews were conducted at the homes of respondents and lasted, on average, about 40 minutes (79 interviews are missing information about the duration of the interview). Although most interviews were conducted in Azeri, a Russian-language questionnaire was also available. All interviewers were bilingual. Completed questionnaires were first reviewed in the field by team supervisors and then were taken by the fieldwork coordinators to the national State Committee of Statistics (SCS) headquarters for data processing. 2.3 Response Rates Of the 11,162 households selected in the household sample, 8,246 included at least one eligible woman (aged 15-44 years). Of those, 7,668 women were successfully interviewed, yielding a response rate of 93%. About 5% of women were absent and could not be interviewed during several revisits. Virtually all respondents who were selected to participate and who could be reached agreed to be interviewed (the individual refusal rate was only 1.2%). Response rates were lower in Baku and its environs (86%) than in other urban areas (94%) and rural areas (96%) (Table 2.3.1). 12 The distribution of women in the sample by 5-year age groups differs slightly from the official estimates for the year 1999: the survey sample slightly overrepresents adolescent women (15- to 19- year-olds) and underrepresents women aged 25-29 by 2 percentage points, after confidence intervals are taken into account (see top panel of Table 2.3.2). The sample retains the same over- and underrepresentation for women aged 15-19 and 25-29 for both urban and rural residents. At least two factors may have contributed to the differences observed: (1) official estimates reflect the age composition recorded in 1999, 2 years before the survey took place, and (2) lower response rates occurred among 25- to 29-year-old women, who are most likely to be employed and not at home. The distribution of women in the sample by marital status (by 5-year age groups), however, does not differ significantly from the Census estimates (see bottom panel of Table 2.3.2). 13 14 CHAPTER 3 CHARACTERISTICS OF THE SAMPLE 3.1 Household Characteristics Similar to the definition used in other surveys and in the 1999 Azerbaijan Census, the AZRHS01 defined household as a person or group of persons who shared the dwelling and the related living expenses. Visitors were not counted in the household composition and were not included in the number of eligible respondents. After all eligible respondents in the household were listed, only one woman aged 15-14 was randomly selected for the individual interview. The survey found that a typical Azeri household containing an eligible respondent consisted of almost five persons (Table 3.1.1). Households in Baku contained, on average, one person fewer than did rural households (4.4 persons per household vs. 5.3 persons per household). The larger household size in rural areas can be partially explained by higher fertility levels (see Chapter 4). One-or two-person households (presumably childless women or couples) were very uncommon (1% and 5%, respectively); these types of households were more frequent in Baku and other urban areas (8% and 6%) than in rural areas (4%). Households consisting of eight or more members were also relatively uncommon; they were least prevalent in urban areas (4%) and most prevalent in rural areas (11%). Socioeconomic well-being is an important determinant of reproductive health status. Information on household amenities (i.e., electricity, flush toilet, telephone line, and central heat) and ownership of various goods or properties (i.e., television; refrigerator; private car; video recorder; mobile phone; vacation home; and vegetable garden, orchard, or vineyard) collected in AZRHS01 is shown in Table 3.1.2. Response options to each of these items were "yes" and "no". In addition, information on the average hours of electricity available per day and on household crowding were obtained for each respondent. Crowding was determined by the total number of persons living in the household divided by the total number of rooms in the house (not including the kitchen or bathroom); if the result was being greater than one, respondents were classified as living in crowded conditions. 15 Only 38% of respondents lived in households with flush toilets, and about 40% had a telephone line at home. Only one-third of respondents had an uninterrupted power supply, and 1 in 5 had central heating. The proportion of households with such amenities varied significantly by residence. For example, Baku households were 12 times more likely than rural households to have flush toilets, 8 times as likely to have central heating and 24 hours of electricity daily, and 5 times as likely to have a telephone. The average Azeri household has 15 hours of electrical power per day (data not shown). The power supply is most limited in rural areas, where 50% of households have only 10 hours of electricity per day; compared to in other urban areas and Baku,50% of households have 17 and 23 hours, respectively, of electricity daily). Generally, households of people who are not internally displaced persons and refugees (IDP/Rs) living in conflict-affected areas were the least likely to have a flush toilet, central heating, or uninterrupted power supply. Televisions were available in almost every household surveyed,with women of reproductive age (91%); the proportion was highest in urban areas (95%-98%). Almost all households had refrigerators (78%), especially in urban areas, particularly in Baku(96%). As expected, almost all households in rural areas had a vegetable garden (89%), whereas only 1 in 2 and 1 in 5 urban and Baku households, respectively, had such gardens. Video recorders were not widespread: almost 2 . in 3 households in Baku but only 37% of respondents in other urban areas and 22% in rural areas owned a video recorder. Only 1 in 4 families owned a car in Azerbaijan, with little variation by 16 residence. Few families owned a vacation home or a secondary residence (14%). The use of mobile phones was low (19% of women reported that they had one) and was concentrated in Baku and other urban areas (41% and 17%). Crowding did not substantially differ between urban and rural households, although the average number of persons per household was lower in urban areas than in rural areas (data not shown). The IDP/R households were the most crowded. 17 All of these household amenities and goods, including living in uncrowded conditions and having electricity 24 hours per day, were summed to create a score to classify the socioeconomic status (SES) of the household. Equal values were assigned for possession of each amenity or good. For each household, this inventory yielded a score whose reliability was assessed using the Cronbach coefficient alpha. Based on this initial evaluation only 10 items were selected for use in the SES score (alpha coefficient=.70) Possession of a vegetable garden, orchard, or vineyard and having electricity 24 hours per day were not included in the final score because the score is based exclusively on possession of items that are associated with high SES. Possession of a garden is inversely correlated with SES, and the electricity shortage among Azerbaijan households is common and relatively uninfluenced by household SES. The score ranged from 0 to 10, where 0 represented the lower end (no amenities and goods included in the score) and 10 represented the higher end (all 10 items included in the score). The score was further divided into terciles to create three levels for the SES variable. Respondents with a score of 0-3 amenities were classified as living in households with low SES; those with scores between 4 and 6 were classified as having middle SES; and those with scores of 7 or higher were considered as having high SES. The same methodology to assess the socioeconomic distribution of the population has been used in other reproductive health surveys in Eastern Europe and former Soviet Union countries. According to this computation, almost half (49%) of reproductive age women in Azerbaijan live in households with a low SES, 39% have a middle SES, and only 12% are classified as high SES. 3.2 Characteristics of the Respondents A total of 39% of the survey respondents in the sample were young adults (15-24 years of age) (Table 3.2.1). The age distribution was slightly younger in rural areas, where 41% of the women were young adults. Age distribution varied little by region; the only notable exceptions were in the South region, where the population was slightly younger than the country average (43% young adults) and the Central region, where the proportion of young adults was significantly lower (32%). A slight majority of women were legally married (58%); additionally, less than 1 percent were in a consensual union (i.e., an unregistered marriage or living with a partner "as husband and wife" but not legally married). No significant urban-rural or regional differences in marital status were found. Divorce and separation appeared to be uncommon; only 3% of women reported that they had been previously married. Two percent of the women surveyed were widowed. More than 1 in 3 women (36%) had never been married or lived with a partner. Official statistics show that Azerbaijan has a fertility rate slightly above the replacement level of two children per woman (2.2 births per woman in 1999, according to the State Committee of Statistics of the Azerbaijan Republic (SCS). Fertility levels reported in AZRHS01 were similar to the official estimates. Although the percentage of 18 19 childless women was lower in urban than in rural areas (41% vs. 46%), more women in rural areas reported three or more children than did those in urban areas (26% vs. 22%); 27% of the women in the South and Central regions reported three or more children. Baku residents were the least likely to have three or more children (18%). Most of the respondents (48%) had completed secondary education or attended some years of postsecondary school. The proportion who had received formal education beyond the secondary level was 1.5 times higher among urban than among rural residents (37% vs. 15%). Respondents residing in Baku were significantly more likely to have some postsecondary education (45%) than were respondents of other regions, particularly than those residing in the South region (13%). Azeri was the predominant ethnic group (90%), followed by Talish (4%), Lezgi (1.5%), and Tat and Avar (1.3% and 1%). Most of the Talish population was concentrated in the South; Lezgis and Tats were more likely to reside in the North and Northeast regions. Russians represent less than 1% of the population; most Russian respondents reside in Baku (3%). Most respondents (90%) spoke Azeri at home; 4% spoke Russian, and 3% spoke Talish. Women in Baku were significantly more likely to speak Russian in the family than were women outside Baku, even if they had an Azeri background. Most women (80%) reported that they did not work outside the house. Because of low job availability, rural women were even less likely to work outside the house (14% vs. 25% in urban areas). The proportion of women holding a job (including part-time work) was the lowest in the South region (13%) and the highest in Baku (27%). Approximately 10% of all women in the sample were IDP/Rs because of the war in Nagorno- Karabakh and the surrounding territories, including 2% of women who fled Armenia when the war started. Most of the internally displaced families in Azerbaijan resided in the South-West region, where they constituted more than 30% of the population, and in Baku, where they accounted for 13% of the population. About 1 in 4 IDP/R families (26%) resided in temporary housing conditions in public buildings and other government facilities (e.g., hotels, schools, factories, sanatoria, health camps, farms, and other state-owned facilities); about 1 in 5 IDP/R families resided either in mudhouses (16%) or shelters built by nongovernmental organizations (4%); a few IDP/Rs were living in train wagons, dugouts, or tents (3%) (data not shown). 20 CHAPTER 4 FERTILITY AND PREGNANCY EXPERIENCE One objective of the AZRHS01 was to assess the current levels of and trends in risk factors associated with reproductive behaviors and to identify factors that might change such behaviors. Policy makers and program managers may use the findings presented here to design programs that respond to the reproductive behavior of the population and tailor them to meet the needs of key subgroups. To obtain information about reproductive patterns, the questionnaire included a series of questions about marriage, divorce, sexual activity, contraceptive use, childbearing, the use of induced abortion, infertility, desired family size, planning status of all pregnancies in the past 5 years, and information about prenatal care for all births during the past 5 years. Information about pregnancies (i.e., births, abortions, and fetal losses) was collected through a complete lifetime pregnancy history for each woman up to the time of the interview. This information represents an important addition to vital statistics routinely compiled at the local and state level because it allows comparisons of fertility and abortion data by background characteristics and behaviors. 4.1 Fertility Levels and Trends During the past decade, scientifically designed nationwide population-based surveys of reproductive health have been conducted in many countries of eastern Europe and the former Soviet Union with support from the U.S. Agency for International Development and the United Nations and technical assistance from the Centers for Disease Control and Prevention (Reproductive Health Surveys) or Macro Incorporated (Demographic Health Surveys). These surveys have used similar methodology and questionnaires, thus allowing for good comparability across countries. Such surveys have been implemented in several The countries of Eastern Europe and the former Soviet Union countries share a common history as well as recent social, political and economic changes following since the fall of communism; they have inherited the same state-subsidized health care system modeled after the Russian centralized system). Demographically, most of these countries have much in common in the areas of fertility and fertility regulation practices. Nonetheless, at the last USSR Census, the total fertility rate among the Soviet republics differed sharply, an expression of their socioeconomic and cultural characteristics and 21 ethnic diversity. At the forefront were the Central Asian republics where, in all but one country (Kazakhstan), the total fertility rates (TFRs) ranged from 3.9 births per woman in the Kyrgyz Republic to 5.2 births per woman in Tajikistan (Brackett, 1993). The European Soviet republics (Ukraine, Russia, and Belarus) and the Baltic countries reported the lowest fertility levels— about 2 births per woman. Among the Caucasus countries, Azerbaijan had the highest fertility rate (2.8 births per woman), similar to the Kazakhstan rate of 2.9 births per woman. Although fertility levels started to decline throughout the Soviet republics before 1989, the decline continued in the 1990s, in some cases at a faster pace, resulting in fertility levels far below the replacement level of 2.1 births per woman in the European former Soviet countries and in a loss of about 1 birth per woman in the Central Asian republics. Current levels of fertility based on survey responses were estimated with the use of age-specific fertility rates calculated from information collected through the respondents' lifetime pregnancy histories (Table 4.1.1). The TFR was computed by accumulating the 5-year age group-specific fertility rates and multiplying the sum by 5. The TFR is thus defined as the average number of live 22 births a woman would have during her reproductive lifetime (15-44) if she experienced the currently observed age-specific fertility rates (ASFRs). Numerators for the ASFRs were calculated by selecting live births that occurred during the 36-month period preceding the survey and grouping them (in 5- year age groups) by the age of the mother at the time of pregnancy outcome (calculated from the mother's reported date of birth). The denominators for the rates represent the number of woman- years lived in each specified 5-year age group by those mothers during the 3-year period preceding the survey. As shown in Table 4.1.1, the TFRs were the lowest in Eastern Europe, particularly in Romania, Russia, and Ukraine (1.3, 1.3, and 1.4 births per woman, respectively). The TFRs in the Caucasus region, with the exception of Azerbaijan (whose rate of 2.1 births per woman is at replacement level), were higher than in Eastern Europe but slightly lower than the replacement level of 2.1 births per woman. Only the Central Asian republics (except Kazakhstan) have fertility rates above replacement level. Under the assumption of replacement fertility levels held constant and zero net migration, the population of Azerbaijan will continue to grow to more than 10.5 million by the year 2025 and more than 12 million by 2050 (Kingkade, 1994). Table 4.1.1 also shows that survey estimates in most countries were either identical or within survey sampling errors, compared with official estimates for the same time period. For example, the TFR for the 3 years preceding the survey in Azerbaijan was within the sampling error compared with the most recent official rate available (1997-1998), but it was probably higher than the current official reporting. (Ministry of Health [MOH], 2001). The only clear exceptions were Georgia and Armenia, countries where the survey estimates exceeded the official estimates by 30%. Although underreporting of births by civil registries may have played a role, particularly in Georgia, overestimates of the female population were more likely to explain lower official rates in both countries. Denominators used to calculate official fertility rates were derived from the 1989 Census projections without adjusting for the substantial outmigration experienced during the 1990s as a result of war (Armenia, Georgia) and territorial secession (Georgia), produced artificially lower ASFRs and TFRs than the survey estimates (Khachikyan et al., 2001., Serbanescu et al., 2001). During the past 10 years, fertility rates declined in all countries presented in Table 4.1.2. Except for Uzbekistan and Kyrgyz Republic, most countries have relatively low fertility rates with high levels of childbearing among women in their 20s, followed by sharp declines at age 30 and older (Table 4.1.2). Like women in other countries of the region, Azeri women initiate and complete childbearing at an early age. The highest fertility levels are among 20- to24-year-old and 25- to 29-year-old women, accounting for 36% and 32%, respectively, of the TFR. Fertility among adolescent women (44 births per 1,000 women aged 15-19) is the fourth highest, contributing 11% of the TFR. Women aged 35-39 and 40-44 make minimal contributions to total fertility; their ASFRs account for only 5% and 2%, respectively, of total fertility. 23 Compared with its Caucasus neighbors, Azerbaijan exhibits the lowest adolescent fertility but the highest fertility rates for 20-24 -year-old and 25-29-year-old women. As a result, its young-adult fertility rate of 195 births per 1,000 women aged 15-24 years is similar to the rate in Armenia (199 births per 1,000) and higher than that in Georgia (178 births per 1,000). Lower adolescent fertility rates in Azerbaijan were also observed in the 1989 USSR Census, which documented fertility of 29 births per 1,000 women aged 15-19, the second lowest rate in USSR at that time, after Turkmenistan (Brackett, 1993). Lower-than-average teenage fertility rates are probably the result of the higher mean age at first marriage in Azerbaijan (the highest mean among former Soviet republics at the most recent USSR Census) and low rates of nonmarital fertility for this age group. 24 25 Using the findings from AZRHS01, fertility trends can be estimated over several 3-year periods (Table 4.1.3 and Figure 4.1). These trends had to be restricted to women aged 15-39 years since the rates for older women (i.e., age at pregnancy outcome) either represent partial fertility rates (due to the inherent truncation of the time exposure) or cannot be assessed. For example, some women aged 40-14 in 1996-1998 would be more than 44 years old in 1999-2001, but only those aged 44 or younger would have been selected to participate in the survey. Similarly, all women aged 40-44 in 1994-1996 would have been more than 44 years old in 2001 and thus not included in the sample. Compared with previous years (1994—1996), fertility declines in 1998-2001 are notable in all age groups. Given the peak in fertility at age interval 20-24 and 25-29, declines of these ASFRs would have a higher impact on the total fertility decline than would declines in all other ASFRs. For example, between 1994-1996 and the most recent 3-year period (1998-2001), the ASFRs for 20- to 29-year-old women (contributing to 68% of the TFR in both time periods) declined by 27% and 21%, respectively, and the TFR declined by 25%. Fertility decline among women aged 35-39 was even higher (a more than 60% decline), but because their contribution to the total TFR is quite low (5% in 1998-2001 and 9% in 1994-1996), the impact on the TFR decline was less substantial. Similarly, the general fertility rate for 1998-2001, defined as the number of births per 1,000 women of reproductive age (15-44), was 71 births per 1,000 women aged 15-44 and 77 births per 1,000 women aged 15-39. General fertility rate for 1994-1996 was 107 births per women aged 15-39. Thus, between the two time periods general fertility rate declined by 28% (data not shown). The cumulative past fertility of women interviewed in the AZRHS01 was calculated as the percent distribution of women by the number of live births and stratified by the current age of each woman at the time of the interview (Table 4.1.4). A total of 43% of all women aged 15-44 were childless at the time of the interview, but only 9% of women currently in union had not had their first child. Although few women reported a birth before age 20, by age 29, 70% of all women had given birth. Just 1 in 8 women (12%) remained childless by age 44. Among currently married women, 40% of adolescents have already had their first child, 80% of 20- to 24-year-olds have given birth, and more than 90% of women at least 30 years old have had their first child. Only 4% of married women remained childless by age 40-44. A minority of women had four or more children (12% of all women and 19% of currently married women). 26 4.2 Fertility Differentials Table 4.2 shows the ASFRs and TFRs among different subgroups. Fertility among women living in urban areas, including Baku, was almost 20% lower than among rural women in the 3-year period preceding the interview. Most of the difference between rural and urban fertility rates was the result of higher ASFRs among rural residents aged 15-29; fertility at age 30 and older was similar in both rural and urban areas. By region, women living in the Central areas and in Baku had the lowest levels of fertility (1.8 and 1.9 births per woman, respectively). Again, most differences in ASFRs by region were among young adults. The TFR was inversely related to the educational level; it decreased from a high of 2.3 births per woman among those with less than a complete secondary education to 1.8 births per woman among women 27 who attended university. Fertility differences according to education were more pronounced among younger women. Generally, women with the highest educational attainment had the peak of their fertility at ages 25-29, whereas women with the least education reached their highest fertility at age 20-24. Women with low socioeconomic status (SES) had, on average, 2.3 births per woman, 28 compared with 1.9 and 1.6 births per woman, respectively, among women with middle and high SES. Fertility rates were similar for IDP/R and non-IDP/R women. 4.3 Nuptiality Because the main exposure to the risk of pregnancy occurs among women who are married or in a consensual union, reproductive health behaviors are greatly influenced by marital status. Survey results showed that the median age at first marriage among women aged 15-14 (15-49 in Central Asian republics and Armenia) is between 20 and 22 years of age in all countries mentioned in Table 4.1.1. Because the probability of having a child is much higher among married women and couples typically have a strong desire to initiate childbearing soon after marriage (first birth was typically within 2 years after first marriage), most countries of the region exhibit the highest fertility rates among currently married young adults (data not shown). Thus, it is important to know the marital distribution by age group and the changes over time in age at first union and at first birth. The proportion of currently married women in Azerbaijan (58%) was comparable to that of other countries of the region (ranging from 54% in Russia to 68% in Uzbekistan). In addition, a small proportion of women were living in consensual unions, particularly in Eastern Europe (10% of women in Russia, 6% in Romania, and 4% in Ukraine) but much less so in Central Asia (Table 4.3). At the time of the AZRHS01, less than 60% of women aged 15-44 were currently married (58%) or living in a consensual union (1%). Five percent of women were widowed, divorced, or separated (from a spouse or from a partner in a consensual union), subgroups that collectively constitute the category of "previously married." More than 1 in 3 women (36%) had never been married or lived with a partner. The proportion of currently married (either legal or consensual marriage) women did not vary significantly by residence or region. Marital levels increased rapidly with age, from 10% among 15- to 19-year-olds, to 44% among women aged 20-24, and to 73% among 25- to 29-year-olds; the rate reached a maximum of about 84% for women aged 35-39 and started to decline thereafter. Consensual unions were uncommon across all age groups. Separation, divorce, and widowhood increased with age, reaching a maximum of 12% among women aged 40-44. The proportion of never-married women decreased abruptly with age, from 90% among 15- to 19-year-olds to 53% among women aged 20-24, 24% among women aged 25-29, and 11 % among women aged 30-34. Among women aged 35 or older, about 8% of women had never been married. The proportion of women married or in union was significantly lower among women who did not complete high school (47%) than among women with a completed secondary or technicum education 29 30 (61% and 70%, respectively) and those with university or postgraduate education (59%). In studying the impact of education on marital levels, it should be kept in mind that the youngest women are less likely to marry because they are still in school. Because at least some secondary education was compulsory in former Soviet-bloc countries, many young women could not marry before at least age 16 (the youngest age for official marital eligibility), resulting in lower marital rates among those with less than complete secondary education. Among women aged 20-24, however, the likelihood of being in a marital relationship, either consensual or formal, was inversely correlated with education. For example, 47%-52% of 20- to 24-year-old women with high school education or less were in union, compared with 23% of 20- to 24-year-olds with postsecondary education (data not shown). 4.4 Age at First Sexual Intercourse, Union, and Birth Age at first union and age at first sexual intercourse plays an important role in determining fertility. Delays in these events decrease the number of reproductive years that a woman spends at risk of getting pregnant and increase the likelihood of having fewer children. Age at first birth also has a direct impact on fertility because postponing the first birth may contribute to the decline of the TFR. 31 32 In Azerbaijan the median ages at first union and first birth were 22.3 and 23.7, respectively (Figure 4.4 and Table 4.4). Thus, most fertility for the youngest women is typically marital. Out-of-wedlock births are rare in Azerbaijan, and unmarried women contribute little to total fertility (less than 5% of births were out-of-wedlock, according to official records). Information on age at first sexual intercourse, first union, and first live birth for all women are presented by age of the respondent at the time of interview in Table 4.4.1. The left side of the table shows the proportion of respondents within each age cohort (5-year age group) who have ever had sexual intercourse (top panel), ever been in formal or consensual marriage (middle panel), and ever had a live birth (bottom panel) before reaching specific ages. The overall median age (age by which 50% of women aged 15-44 have experienced the event) and the median age within each age group are also displayed for each event. By comparing the proportion of women within different cohorts who experienced various events before age 20, it is possible to detect whether the age of occurrence of each event has changed over time. For example, the proportion of women who had sexual intercourse before age 20 has increased from 26% among 40- to 44-year-olds to 34% among 20- to 24-year-olds; however, the proportion who reported premarital sexual experiences remained essentially unchanged in the two cohorts, because this increase coincides with an identical increase in the proportion of married women among younger cohorts. In Azerbaijan sexual abstinence before marriage is a common practice. Apparently, traditional norms are strong and have not been altered by recent changes that have influenced young adult reproductive behaviors in the industrialized world and in some of the Eastern European former Soviet-bloc countries. As shown in Table 4.4.1, median ages at first intercourse and first marriage for each cohort are virtual identical. Although young women aged 25-29 were initiating sexual activity about a year earlier than older women (e.g., women aged 40-44), they also marry a full year earlier. Thus, essentially no differences exist across subgroups in the time interval between the first intercourse and the first union, and premarital sexual intercourse is uncommon for all cohorts. As mentioned previously, a higher proportion of women in the younger cohorts had their first marriage before age 20 (34% among 20- to 24-year-olds) than in the older cohorts (26% among 40- to 44-year-olds). Consequently, the median age at first union has decreased by a full year, from 22.6 to 21.5, between the two cohorts. This trend is particularly interesting and has potential implications for future fertility patterns and fertility control measures. Given that more Azeri women currently marry at younger ages than older cohorts (a return to some extent to traditional practices of the pre- Soviet era) and that they have a relatively early start (1-2 years after the first marriage) and end to childbearing, their fertility is likely to remain at the replacement level. However, compared with their counterparts in older cohorts, they will spend a greater number of their reproductive years at risk of unintended pregnancy and have a greater need for long-term, effective contraception; in the absence 33 of effective birth control methods, they will be more likely to rely on induced abortion to avoid unwanted births. The most notable change between cohorts is manifested in the patterns involving age at first birth. The age at first birth has also decreased for younger cohorts, paralleling the decrease in the age at first union. A substantially larger proportion of women aged 20-24 had their first birth before age 20 than women aged 40-44 (22% vs. 12%). The time interval between the first union and the first birth within each cohort, however, has gradually increased. For example, the median age at first birth among 25- to 29-year-olds was 1.8 years later than their median age at first union, whereas median age at first birth among 40- to 44-year-olds was 1.4 years later than their median age at first union. These findings suggest that younger cohorts tend to marry younger than older cohorts but have a slightly later onset of childbearing than in the older cohorts. Among all reproductive age women, 85% had their first union by age 30 and 79% had their first live birth by that age (data not shown). Urban women initiate sexual activity, union, and childbearing at a slightly older age than rural women do, but the difference is not significant (Table 4.4.2). The intervals between these events are similar for urban and rural residents, a fact that may explain the lack of significant differences in 34 cohabitation and fertility rates by residential area. Differentials in median age of experiencing these events are significantly affected by education. The median age at first intercourse, first marriage, and first birth was 3.1 years older for women with university education than for those who had not completed secondary education. 4.5 Recent Sexual Activity Current sexual activity is an essential indicator for estimating the proportion of women who are at risk of having an unintended pregnancy and are therefore in need of contraceptive services. It also has major implications for the selection of a contraceptive method that best suits the reproductive behavior and fertility preferences of each individual. As shown in Table 4.5 and Figure 4.5, 36% of all women aged 15-44 who were interviewed in the AZRHS01 reported that they had never had sexual intercourse. Additionally, 4% of all women were pregnant, and 4% reported postpartum abstinence at the time of the interview. For all women with sexual experience who were not currently pregnant or postpartum (56%), only 44% were currently sexually active (i.e., had intercourse within the month preceding the interview). Thus, if we exclude respondents who have never had intercourse, 79% of sexually experienced women were currently sexually active (44% of 56%). 35 36 Among women who were married or living with a partner, 75% reported having had intercourse at least once within the past month and 7% had had intercourse 2 or 3 months previously (Table 4.5). Conversely, only 5% of previously married women had had intercourse within the past 3 months. Most previously married women (76%) reported that their last sexual intercourse occurred 1 or more years ago, presumably while they were still married. Less than 1 percent of never-married women reported having had any sexual experience. Only 1 in 4 young adult women (i.e., those aged 15-24) reported sexual intercourse; of those, 53% reported their last sexual encounter within the past 30 days, and 34% were pregnant or in early postpartum. More than 80% of women aged 25 or older, over 80% reported sexual experience. Of those, more than two-thirds had had intercourse within the past month. 4.6 Planning Status of the Last Pregnancy For each pregnancy ended since January 1996, all respondents were asked about the planning status of the pregnancy at the time of conception. Each pregnancy was classified as either planned (i.e., wanted at the time it occurred), mistimed (i.e., occurring earlier than intended), unwanted (i.e., the respondent wanted no more children), or unsure. Mistimed and unwanted pregnancies together constitute unintended pregnancies (Westoff, 1976). Considerable evidence indicates that women who are pregnant with an unintended pregnancy are more likely to seek an elective abortion, to enter prenatal care late or not at all, and to experience pregnancy or perinatal complications (Brown and Eisenberg, 1995). Data on pregnancy intendedness should be interpreted with caution, however, because they tend to underrepresent the level of unintended pregnancies. One common source of underreporting is induced abortions, which are not always reported; because most pregnancies ending in elective abortion are unintended, such pregnancies will be underreported to the extent that abortions are underreported. Abortion underreporting does not appear to be a major concern in AZRHS01 because abortion rates calculated from the survey exceeded recent officially reported levels. Another source of underreporting may stem from retrospective rationalization and ambivalence for unintended pregnancies ending in live births. Women are asked to report retrospectively their thoughts about the pregnancy intention at the time of conception, and retrospectively reported intentions after the child is born become more positive (Miller, 1994). Thus, data shown here represent conservative estimates of the true levels of intendedness for pregnancies ending either in abortions or in live births. Despite the potential underreporting of unintended conceptions, the data in Table 4.3 show some important differences in pregnancy intendedness by according to pregnancy outcome and background characteristics. Not surprisingly, given the high rates and ratios of induced abortion, the proportion of pregnancies 37 that are unintended is quite high. Fewer than 1 in 2 women of childbearing age (42%) said that her most recent pregnancy was intended at the time of conception, 9% reported it as mistimed, and 48% reported it as unwanted. Thus, 57% of women reported their last pregnancy as unintended; most of those women (84%) reported it as unwanted rather than mistimed. However, most women whose last pregnancies resulted in live births said those births were intended (85%). Conversely, all but a small percentage of women whose last pregnancy ended in induced abortion reported that their conceptions were unintended (96%). It should be noted that a relatively high proportion (35%) of women whose last pregnancy ended in miscarriage or stillbirth reported these pregnancies as unwanted, 4 times the proportion of women with live births who reported an unwanted pregnancy. These data suggest that either unintendedness had a negative influence on pregnancy development and outcome or that some of these outcomes may have been induced abortions, reported as spontaneous abortions or stillbirths because respondent bias toward giving a more socially desirable response. The relatively high unintendedness of pregnancies reported as spontaneous abortions or stillbirths was similar to that observed in other Eastern European reproductive health surveys (Serbanescu, 1995, 1998, 2000). Planning status of the last pregnancy did not vary significantly by residence, but unintended pregnancies increased with age and parity. Adolescents and women aged 20-24 were less 38 39 likely to report unintended pregnancies (13% and 36%, respectively) than were women aged 25-29 (53%), 30-34 (70%), or 35 and older (78%). The ratio between unwanted and mistimed conceptions also varied with age; among 15- to 19-year-olds, most unintended pregnancies were mistimed rather than unwanted (the unwanted-to-mistimed ratio for these women was 1:16). Among women aged 20 or older, more pregnancies were unwanted than were mistimed. The unwanted-to-mistimed ratio for these women ranged from almost 2:1 among 20- to 24-year-olds to 3:1 among 25- to 29-year- olds, 8:1 among 30- to 34-year-olds, and 50:1 among those aged 35 or older. Thus, mistimed pregnancies are rapidly replaced by unwanted pregnancies as maternal age increases, primarily because spacing failure is replaced by the failure to end childbearing. As a result, virtually all unintended pregnancies among respondents were unwanted at older ages. A similar pattern can be seen when the planning status of the last pregnancy is examined in light of the number of living children. Women who had never had a live birth and women with one child were less likely to report that their last pregnancies were unwanted than were women with two or more live births. The level of unintended pregnancy did not vary significantly with education, socioeconomic status, ethnic background, or IDP/R status. 4.7 Future Fertility Preferences Knowledge of reproductive intentions in a population is essential for helping couples choose the contraceptive method that will allow them to control when to have children. The preference among women for small families is reflected not only in declining fertility levels and high abortion rates but also in their stated desires not to have more children. Among women in union, more two-thirds of respondents (69%) reported that they did not want to have more children (Table 4.7.1). Only 22% of women currently in union said they intended to have a child in the future, including 14% who wanted a child right away or within 2 years and 8% who wanted to wait at least 2 years before having another child. An additional 2% were unsure whether they wanted to have more children, and 7% said that they could not have any (more) children. The intention to have any (more) children decreased rapidly with as the number of living children increased. By the time women had two children, most (77%-93%) were ready to terminate childbearing. Among those with no living children, almost 3 in 4 women (71%) wanted children; that proportion dropped to less than 16% among women with two or more children. Among women who wanted more children, the timing of the next birth was also influenced by parity: Most childless women wanted to have a child right away or within a year, whereas women with one or more children wanted to have another child after 2 or more years. 40 Women have had all the children they desire by a young age; consequently they have many years of exposure to the risk of unintended pregnancy. Younger women were much more likely than older women to want more children (see bottom panel of Table 4.7.1). The intention to have more children steadily decreased from 79% among the youngest age group to 50% for women aged 20-24, 34% among those aged 25-29,18% among women aged 30-34,9% among women aged 35-39, and only 4% for women aged 40 and older. Of those who desired additional children, most women wanted 41 to wait at least 1 year, except for the few women aged 35 or older who did not want to terminate childbearing and wanted to have a child right away. The desire to have a child within 1 year is lower among 15- to 19-year-olds (34%) and 25- to 29-year-olds (15%), presumably because they want to space the next pregnancy. These findings are important for the national family planning program, which should consider spacing methods for younger women and long-term or permanent methods for older women. Such low levels of desired childbearing, especially with limited availability of effective long-term contraception and the typically early start (and finish) of childbearing, increases the probability of unintended pregnancies and subsequent abortion. A total of 74% of Azeri women who can conceive reported that they do not want to have more children (Table 4.7.2). Only 23% of those with one living child wanted no more children, contrasting with 81% among two-child women and 95% among women with three or more children. The desire to terminate childbearing does not vary significantly by residence and education at any parity but is 42 directly correlated with age. Women younger than age 35 were less likely to report that they wanted to terminate childbearing at any parity. Despite substantial differences in fertility between the Eastern European and Central Asian countries examined, rates of childbearing have fallen substantially in all places and reproductive intentions, especially for couples with two or more children, are surprisingly similar (Figure 4.7). Although strongly influenced by different social norms, cultural values, and economic circumstances, reproductive intentions in these countries show a similar pattern. Among fecund women in union, between 50% and 77% want no more children. The desire to limit fertility is generally higher in Eastern European and Caucasus countries than in Central Asian countries. Regardless of the region, the desire for additional children decreases rapidly with the number of living children. By the time women have two children (or three children, in Central Asia), they generally are ready to terminate childbearing. In Eastern Europe and the Caucasus Region, more than 80% of women with two or more children (90% in Romania and Russia) report that they want no 43 more children. In Central Asian Republics, most women with three, four, or more children report that they want to terminate fertility (data not shown). As mentioned earlier, such low levels of desired childbearing, especially given the limited availability of effective long-term contraception and the typically early start (and finish) of childbearing, enhances the probability of unintended pregnancies and subsequent abortion. Public health officials and health care providers should always consider fertility preferences in their efforts to help couples satisfy their contraception needs. 44 CHAPTER 5 INDUCED ABORTION As discussed below, induced abortion, not contraception, has been the main method of fertility control in the 15 independent countries that emerged from the collapse of the USSR. In most of those countries, the abortion-to-live-birth-ratios in 1989 were greater than one abortion for every live birth, although systematic underreporting of induced abortion was very likely (Popov, 1996). For the entire Soviet Union in 1989, the abortion-to-live-birth-ratio was 1.3:1, the abortion rate was 96 per 1,000 women aged 15-49, and the lifetime induced abortion rate was 3.3 abortions per woman. Economic, social, and cultural differences among the countries most likely have affected abortion reporting, making comparisons among countries difficult to interpret. Russia, Belarus, and Ukraine have consistently reported the highest rates of abortion, whereas the rates in Central Asia were substantially lower (Goskomstat USSR, 1990). Several factors are widely believed to have contributed to the widespread use of abortion and underutilization of modern contraception. The relative isolation of the USSR from the contraceptive advancements in Western countries affected both the knowledge about and the availability of high- quality contraceptive methods. In addition, misconceptions among both family planning clients and providers about the health risks associated with certain modern methods, fatalistic attitudes toward health issues, and a medical system that promoted curative rather than preventive care, compounded by easy access to and low cost of obtaining induced abortions, have contributed further to the high reliance on induced abortion (Remennick, 1991, Popov, 1996). These patterns were further shaped by a climate of strong moralistic principles, which condemned premarital and extramarital pregnancies, disapproved of sex education in school, and discouraged open discussions about sex- related issues. The extent to which these factors continue to play a role in the use of induced abortion varies from one country to another, now that each country is in the process of developing new reproductive health policies and programs. 5.1 Abortion Levels and Trends For several decades one of the most outstanding demographic features of most of the Eastern European countries has been the high reliance on induced abortion as a means of birth prevention. Induced abortion has been the single most important method of controlling fertility. In recent years, 45 abortion rates and ratios in many of these countries have been among the highest in the world. Factors frequently cited as contributing to widespread reliance on abortion include the limited availability of contraceptive methods; poor quality of the methods available; fears about possible side effects, particularly with hormonal methods; and easy access to and low cost of induced abortion. Before the Soviet Union's breakup, Azerbaijan had the lowest abortion rate in the Caucasus region (23 abortions per 1,000 women aged 15-49, compared with 31 per 1,000 in Armenia and 51 per 1,000 in Georgia), and it was significantly lower than that in the Slavic republics (Goskomstat USSR, 1990). Since the breakup of the former Soviet Union, the reported vital statistics indicate a steep decline in the abortion rate (from 23 per 1,000 women aged 15—49 in 1989 to 12.7 per 1000 in 1998, and 7.7 per 1,000 in 2000), but this decline is not supported by the AZRHS01 data (MOH, 2001a). The ability of official abortion statistics to document the true magnitude of the abortion levels is hampered, however, by the underreporting of abortions performed in the private sector, inherent problems related to registration data in state-run medical facilities, and the persistence of abortion performed outside clinical settings. Similar barriers have led to discrepancies between official and survey-based abortion rates in other countries, particularly in the Caucasus region (Table 5.1.1). 46 Survey estimates in most countries were either within sampling error (e.g., Moldova, Uzbekistan) or slightly higher compared with official estimates for the same time period. In the Caucasus countries, however, the survey estimates exceeded the official estimates by a considerable margin (survey-based estimates were 6-11 times higher than official estimates). It is worth noting that all three countries in the Caucasus region experienced massive internal or external territorial disputes, population displacements, or outmigration that may have caused significant disruptions in the collection of health statistics in recent years. Table 5.1.2 shows total and age specific abortion rates based on data from recent reproductive or demographic health surveys conducted in Eastern European countries and the Newly Independent states (Goldberg et al., 1993; KIIS and CDC, 2000; ORC/MACRO International 1995-2001; Serbanescu et al., 1995, 1998, 2001; VCIOM and CDC, 1998, 2000). With the exception of Romania, where abortion was illegal until 1990, most of the Eastern European countries have had some of the highest abortion rates in the world for several decades (Table 5.1.2). 47 The AZRHS01 found that abortion rates in Azerbaijan are higher than those reported in recent surveys in the Russian Federation (urban sample), Romania, and Armenia and much higher than in Moldova, Ukraine, and Central Asian republics, but lower than in Georgia. The age-specific abortion rates (ASIARs) shown in Tables 5.1.2, 5.1.3, and 5.1.4 represent the proportion of women in a specific age group who terminated pregnancy by induced abortion within the 3-year period preceding the survey. The rates were calculated by using the age of the woman at the time of the pregnancy's termination. The total induced abortion rate (TIAR) was calculated by summing the ASIARs for the same 3-year period used in the analysis of fertility levels. Similar to the total fertility rate (TFR), the TIAR describes the number of abortions a woman would have in her lifetime under the current ASIARs. The AZRHS01 data indicate that the general abortion rate in the 3 years before the survey (May 1998-April 2001) was 116 abortions per 1,000 women aged 15-44, the total abortion rate was 3.2 abortions per woman, and the abortion-to-live-birth ratio was three abortions for each live birth (3:1). National sample surveys on reproductive health, which could have provided information about induced abortion levels based on women's self-reports, have never been carried out before in Azerbaijan, so comparisons with similar data on abortion are not possible. 48 The TIAR was 1.5 times higher than the TFR during the 3 years prior to the survey (3.2 vs. 2.1). Unlike fertility, the age pattern of abortions in Azerbaijan is concentrated at age groups 25-29 (177 induced abortions per 1,000 women) and 30-34 (176 per 1,000), which together account for 50% of the TIAR. The third highest ASIAR occurred among women aged 20-24. Except for the youngest age groups, ASIARs are significantly higher than ASFRs (Figure 5.1.1). These findings suggest that Azeri women achieve their desired family size at young ages, after which most pregnancies are unintended and are intentionally terminated. The official statistics do not routinely calculate lifetime total abortion rates. The USSR's statistics, however, show a pre-independence TIAR of 0.8 abortions per woman aged 15-49, the lowest lifetime abortion rate among former Soviet Union countries (Brackett, 1993). Based on the most recent ASIARs for abortions performed in governmental facilities reported by the Azerbaijan Ministry of Health (MOH), the estimated TIAR for the period 1998-2000 was 0.3 abortions per woman, two-thirds lower than the pre-independence level (Table 5.1.3). Compared with the survey estimates, the official ASIARs are generally low, but they are substantially lower for the two age groups that contribute to 50% of abortions: the ASIAR among women aged 25-29 (20 induced abortions per 1,000 women vs. 177 per 1,000), and the ASIAR among women aged 30-34 (16 induced abortions per 1,000 women vs. 176 per 1,000). It is likely that the underreporting of abortions among women aged 35-39 is of the same magnitude, but the official statistics do not allow for a separate ASIAR for this age group. 49 50 AZRHS01 data permit estimation of abortion trends over several 3-year periods (Table 5.1.4 and Figure 5.1.2). These trends had to be restricted to women aged 15-39 because the rates for older women (age at outcome) represent partial abortion rates (due to the inherent truncation of the time exposure) or cannot be assessed. For example, some women aged 40-44 in 1996-1998 would be older than age 44 in 1999-2001, but only those aged 44 or younger would have been selected to participate in the survey. Similarly, all women aged 40-44 in 1994-1996 would have been older than 44 years old in 2001 and thus excluded from the sample. Abortion rates were higher in 1998-2001 than in previous years (1994-1998) in all age groups. Given the peak in abortion at age intervals 25-29 and 30-34, changes in these ASIARs would have a greater impact on abortion trends than changes in other ASIARs. For example, between 1994- 1996 and the most recent 3-year period (1998-2001), the ASIARs for 25- to 34-year-old women (contributing to more than 50% of the TIAR in both time periods) increased by 30% and 20%, respectively, while the TIAR for women aged 15-39 increased by 25% (from 2.3 to 2.9 abortions per woman aged 15-39). Increases in abortion rates among women aged 15-24 were even higher (200% and 40%, respectively), but because their contribution to the total abortion rate is quite low(16%in 1998-2001 and 14% in 1994-1996), their impact on the TIAR increase was negligible. 5.2 Induced Abortion Differentials As shown in Table 5.2.1, the abortion rates among all women were equally high and varied little by background characteristics, except for internally displaced persons and refugees (IDP/Rs), who reported substantially higher rates. The TIAR in rural areas was about 20% higher than in urban areas (3.4 vs. 2.8 abortions per woman). Abortion rates in the South-West (where most of the IDP/R population resides) were substantially higher than in the rest of the country; the rates in the South region were the lowest. The TIAR was lowest for women with a university education; on average, women with lower levels of education reported 0.7 abortions more than women with postgraduate education did (3.2 vs. 2.5 abortions per woman). Most of the variation in abortion rates by education was the result of higher ASIARs among women aged 15-24 who had less than a university education. Women of Azeri ethnic background reported, on average, one abortion more than did those of other backgrounds. IDP/Rs reported the highest TIAR (4.7 abortions per woman) and higher ASIARs than non-IDP/R women. One way to reduce unintended pregnancies that result in abortion is through the provision of family planning services. In countries around the world, increases in the use of modern contraceptives have, over time, been associated with decreases in the numbers of abortions (Cohen, 1998). 51 As shown in Chapter 10, a large share of the potential demand for family planning services is among subgroups of women who have also reported higher rates of induced abortion (i.e., rural women, those who are less educated, women with two or more children, and IDP/R women), a finding indicating that access to services is not equal and that the family planning program needs to expand its reach. 52 53 As shown in Table 5.2.2, 1 in 3 women of reproductive age (35%) reported having had at least one induced abortion. The likelihood of having an abortion is positively associated with age because exposure to pregnancy, particularly unintended pregnancy, increases with age. Although few adolescents reported any abortions (0.4%), by ages 20-24 the percentage rises to 12%; it increases to more than one-third among 25- to 34-year-olds and 60% among women aged 35 and older. The likelihood of having an abortion is also positively associated with the number of living children, which is also a strong predictor of unintendedness because women in Azerbaijan achieve their desired family size of one or two children fairly rapidly. The likelihood of having at least one abortion was somewhat greater among urban women, IDP/Rs, and women who have at least completed secondary education. In every region except the South, more than 1 in 3 women had had at least one abortion; just more than one-quarter of women in the South reported ever having an abortion. As shown in Figure 5.2, the use of abortion was also heavily influenced by pregnancy order, which refers to all prior pregnancies, including live births, induced abortions, miscarriages, and other outcomes. Women with no prior pregnancies were the least likely to have pregnancies ending in abortion (1%) and the most likely to have a live birth (86%). The likelihood of abortion increases 54 rapidly among women who had any prior pregnancies. Although a woman with one prior pregnancy has a likelihood of abortion lower than that of having a live birth, once she has two or more prior pregnancies the likelihood of resorting to abortion is significantly higher than that of carrying the pregnancy to term. Thus, the induced-abortion-to-live-birth ratio is directly correlated with pregnancy order, increasing from 0.01:1 among women with no prior pregnancy, to 1.1:1 among women with two prior pregnancies, 5.4:1 among women with four or more prior pregnancies. Because not all women were exposed to the risk of an unintended pregnancy and a subsequent abortion, in the right panel of Table 5.2.2 we restricted the denominator to include only women who have ever had an abortion. More than 1 in 3 women (35%) reported they had only one abortion, 24% had two abortions, 16% had three abortions, and 26% had four or more abortions, including 2% who had 10 or more lifetime abortions. Women who reported multiple abortions were more likely to be older, to have high parity, to be IDP/Rs, and to live outside the South or North-Northeast regions. 5.3 Abortion Services As is the case with all the former Soviet republics, Azerbaijan was subject to the liberal abortion legislation and regulations issued by the former USSR. Abortion on request has been available within the first 12 weeks of gestation since the Soviet Supreme Council decree issued in November 1955. The decree, entitled "On the Elimination of Induced Abortion Prohibition," reinstated the first Soviet abortion law, which was issued in November 1920 and revoked in June 1936. With several additions and modifications, this 1955 law remained in force essentially unchanged. In 1987, early abortions by electric vacuum aspiration after obligatory pregnancy testing were authorized by the Order of the Ministry of Health of the USSR No. 757 (June 5, 1987). These procedures were called "mini-abortions" because they are performed in the earliest stages of gestation (in women whose menstrual period is no more than 20 days overdue, roughly corresponding to a maximum of 6 weeks of pregnancy), involve minimal cervical trauma (i.e., do not require cervical dilatation and anesthesia), and use electrical vacuum aspiration rather than sharp curettage. The same order permitted mini-abortions to be performed outside hospitals in ambulatory clinics. Starting in 1989, early pregnancy termination by vacuum aspiration was officially recognized as a legal abortion procedure, but it was reclassified as "menstrual regulation" and reported separately from the induced abortion statistics. Such reclassification, however, can be misleading because menstrual regulation does not require a pregnancy confirmation and is not regarded legally as an abortion (WHO, 1997). In all the former Soviet Union countries, menstrual regulation by vacuum aspiration is performed after pregnancy has been confirmed; its primary intent is to terminate an unwanted pregnancy, so it must be reported in the total abortion statistics. 55 Additional regulations were issued to permit induced abortion during the first 28 weeks of gestation on medical and social grounds (USSR MOH, Order No. 234 of March 1982 and Order No. 1342 of December 1987) and to briefly legalize "commercial" abortions in private clinics and "for-fee" sections of state hospitals (legalized in March 1988 by the USSR MOH and outlawed in December 1988 by a decree issued by the Council of Ministers) (USSR MOH; USSR Council of Ministers). Under the current law, induced abortion can be performed only by ob/gyns by either vacuum aspiration or sharp curettage; abortion procedures are permitted only in medical facilities that have been state-certified for performing abortion. Outpatient medical facilities (e.g., women's consultation clinics [WCCs] and private clinics) can perform induced abortion only by vacuum aspiration. The AZRHS01 collected information on respondents' last four abortions performed since January 1996 in a detailed abortion history that included questions about the reason for abortion; the place where the procedure was performed; abortion registration and payments; use of local or general anesthesia and antibiotic prescriptions; number of nights, if any, spent in the hospital after the procedure (abortion patients are released in the same day of the intervention if they do not have postabortion complications); and the presence or absence of early and late postabortion complications. Data were collected starting with the most recent procedure in an attempt to minimize recall biases. The data presented here are from detailed abortion histories of abortions that took place from July 1996 to June 2001. Almost all abortions (90%) were reported to be completed in the first trimester of gestation. However, respondent reports on this issue are subject to several possible biases, including irregular menses, problems in recalling the event, and reluctance to admit abortions beyond the legal gestational limit. Almost 1 in 2 abortions (46%) were reported to be performed between 7 and 12 weeks of gestation, 43% were performed before 7 weeks, and 11% were reported as late abortions (13 weeks or more). The numbers are too small to draw any statistical conclusions, but late abortions were reported more often by rural women, women with less than complete secondary education, and women with low socioeconomic status (SES). Late abortions were more common among women with no prior induced abortions than among those with one or more prior abortions (data not shown). Of all abortions reported by survey respondents in the past 5 years, approximately 40% were mini- abortions (Table 5.3.1). Mini-abortions were twice as prevalent among urban respondents as among rural residents (52% vs. 24%) and were most common among women living in Baku (62%). The proportion of abortions classified as mini-abortions decreased somewhat with woman's age and increased directly with education and SES. Mini-abortions were least prevalent among women in the South-West (14%), whereas in other regions they constituted one-quarter to three-fifths of the 56 57 abortion procedures reported (27%-62%). Mini-abortions were slightly less prevalent among IDP/Rs than non-IDP/R women. Although ambulatory clinics are not licensed to perform D&C (dilation and curettage) abortions, mini-abortions represented only 60% and 47%, respectively, of induced abortions performed in WCCs and private clinics. Thus, D&C abortions performed in ambulatory clinics, along with abortions performed outside medical facilities, are likely to substantially contribute to the underregistration of abortions reported by the Ministry of Health. The proportion of induced abortions terminated by vacuum aspiration did not vary significantly by the year of pregnancy termination. The percentage of pregnancies terminations by vacuum aspiration was slightly lower among first abortions than among repeated abortions because first-order abortions were more often performed at gestational ages of 7 weeks or later (39% of first-time abortions were performed before 7 weeks of gestation compared with 45% of abortions of rank two or higher) (data not shown). By law, all abortions must be performed in hospitals or ambulatory clinics or cabinets (offices) by ob/gyns. As shown in Table 5.3.2, most survey respondents' induced abortions occurring in 1996 or later were performed in gynecological wards (70%). About a quarter of them (26%) were performed in state-run ambulatory units, such as WCCs, and only 1% were performed in private clinics. Abortions performed in WCCs were more prevalent in urban areas (34%) than in rural areas (14%). In Baku, abortions performed in WCCs were still outnumbered by those performed in hospitals (40% vs. 58%). Abortions performed in private clinics and in WCCs increased with education and SES. Early abortions (i.e., mini-abortions) performed by vacuum aspiration were more likely to occur in hospital wards and WCCs (59% and 38%, respectively). Although most induced abortions at 7 weeks or later were performed in hospital wards (77%), 18% were reported to be performed in ambulatory units (17% in WCC and 1% in private clinics), and 4% were performed outside medical facilities. The distribution of location of abortions did not change between 1996 and 2001. Only 3% of pregnancy terminations were reported to take place outside the health system; however, about two-thirds of those abortions (68%) were performed by either D&C or vacuum aspiration, suggesting that they were performed by qualified physicians at either their homes or the respondents' homes (data not shown). Because abortions performed outside medical facilities (either self-induced, performed by lay persons, or performed by doctors outside the health system) are illegal, it is likely that women were reluctant to admit these outcomes, in spite of the interviewer's assurance of anonymity and that this figure is probably an underestimate of the proportion of abortions performed outside the health facilities. Rural women were more likely to report such abortions than were urban women (4% vs. 2%); women residing in the South and Central regions (7%), those with less than 58 59 complete secondary education (6%), and those with low SES (4%) were slightly more likely to report abortions performed outside certified health facilities (data not shown). Only 5% of abortions were preceded by testing for sexually tranmitted infections (STIs) (data not shown). STI screening was more slightly likely among women in Baku and the North-Northeast region and among those with a university education, high SES, and abortion taking place after the first trimester. No difference in screening was found between urban and rural areas. Tests for pregnancy confirmation were more likely to be performed when abortion procedures took place in ambulatory settings (either WCC or private clinics) than in hospital gynecologic wards. Nearly 3 out of 5 abortions were reportedly due to method failure (Table 5.3.3). Almost all women claimed to be using traditional methods of contraception; just 4% of abortions were for pregnancies that occurred while a woman was using a modern method. Rural women, women living in the South region, and women with low SES were most likely to report contraceptive use (mostly traditional methods) before the aborted pregnancy. In Azerbaijan, almost all abortions are performed for a fee (which varies from one facility to another). Reports of abortion payments were lower among rural women than urban women, outside of Baku, and increased directly with education and SES. At the time of the survey, mean charges for an abortion procedure were about 47 thousand manat (about US$10.00). The amount paid for an abortion ranged from no payment to, in one case, 900,000 manat (Table 5.3.4). Only 4% of abortions were performed at no charge; 11% of abortion payments were 20,000 or less, 38% were between 21,000 and 40,000 manat, 39% were between 41,000 and 100,000 thousand manat, and 3% were more than 100,000 manat. Less than 5% of women reported that abortion payments were only gifts of unknown amount or could not remember the amount paid. Women in urban areas, including those living in Baku, those with university training, and those with high SES, were more likely to make, on average, larger abortion payments than other women. The cost of late abortions was 31% higher than abortions performed in the first 12 weeks of pregnancy. The average abortion payments were highest for WCCs and private clinics and lowest for procedures performed outside a medical facility. Generally, abortion performed after 6 weeks of gestation is an inpatient procedure, but patients are released within the same day and do not have to spend the night in the hospital. Survey results confirmed that virtually all women who had abortions since 1996 (98%) had been released within the same day of the abortion procedure (data not shown). Only 2% of women with abortions had to be hospitalized for at least one night; the length of hospital stay varied with gestational age and the presence or absence of abortion complications. 60 61 62 Nearly 3 in 5 abortions (59%) between 1996-2001 were performed without any anesthesia; one-third were performed with local (cervical) anesthesia and 5% involved intravenous anesthesia (Figure 5.3). The likelihood of receiving anesthesia was higher in urban areas than in rural areas, increased with the respondent's education and SES, and was directly influenced by gestational age (data not shown). Early abortions (i.e., under 7 weeks) were the most likely to be performed without anesthesia (62%), whereas about one-half (55%) of abortions performed at a gestational age of 13 weeks or higher received anesthesia. The likelihood of anesthesia for abortions performed by D&C was not significantly different from the likelihood for anesthesia with vacuum-aspiration abortions (43% vs. 39%). Women having hospital-performed abortions (which are more likely to be performed after 6 weeks and by D&C) were slightly more likely to receive anesthesia than were those who obtained abortions in a WCC (41% vs. 35%). 63 5.4 Abortion Complications Legally induced abortions are associated with a certain risk of postoperative complications, whose incidence and severity are strongly correlated with age of gestation, parity, woman's age, surgical procedure, operator's skill, type of anesthesia, and preexisting pathology (Henshaw, 1990). Abortions performed at 7 to 9 weeks of gestation have significantly fewer complications than those performed between 10 and 14 weeks. Similarly, abortions performed by vacuum aspiration have fewer complications than the classic D&C procedure. First-trimester abortion complication rates from studies performed in developed countries ranged from 0.9 per 100 abortion procedures in the United States (Hakim-Elahi et al., 1990), to 3 per 100 in France (Thonneau et al., 1998), and 6.1 per 100 in Denmark (Heisterberg and Kringlebach, 1989), but in the absence of an international standard definition of abortion morbidity, comparisons between countries are difficult to interpret. Survey estimates of postabortion complications are usually based on symptoms or conditions reported by respondents and therefore may be less accurate than hospital based statistics. As shown in Table 5.4.1, 21% of all abortions performed since 1996 were followed by immediate complications (17%) or late sequelae (4%). This finding is consistent with the level of postabortion complications documented by other reproductive health surveys conducted in Eastern European countries with high abortion rates, as shown in Figure 5.4. 64 65 Early complications were most prevalent among women living in the South-West region (23%) and among women with late abortions (23%) or abortions performed outside of public medical facilities (24%-25%). As expected, abortions with early complications were more likely to be followed by late sequelae (at 6 months or more after the abortion was performed) than were abortions without any immediate health problems (13% vs. 2%). Just 21% of abortions were followed with antibiotic treatment (Table 5.4.1). Women were more likely to receive antibiotic treatment if they lived in an urban area (especially Baku), attended postsecondary education, were of higher SES, had a late abortion, or had the procedure performed at a private clinic. Women who suffered early complications were nearly twice as likely to receive antibiotics (39%). Most of the early complications involved prolonged pelvic pain (78%), severe or prolonged bleeding (42%), high fever (38%), and pelvic infection (27%); less than 1% of complicated abortions had perforations of the uterus (Table 5.4.2). With the exception of uterine perforation, it is difficult to assess how serious the other early complications might have been. An indirect approach to measure their severity is to consider early complications as serious when they required overnight hospitalization or were followed by late complications. As shown previously, few immediate complications required one or more nights of hospitalization, and 22% were associated with late complications. The prevalence of early complications increased by 50% after 13 weeks of gestation. 66 5.5 Reasons for Abortion Most induced abortions (64%) take place because of the woman's desire to not have a child (Table 5.5 and Figure 5.5). Nearly 1 in 5 (18%) abortions were obtained because of economic or social reasons (e.g., low income, unemployment, fear of losing a job, or crowded living conditions), 14% because the woman wanted to space childbearing, and 1% for partner-related reasons (e.g., the partner objected to the pregnancy).Only 3% of abortions took place for maternal health reasons (i.e., pregnancy was threatening the woman's physical or mental health), and 1% took place because of fetal defects or potential risks for the baby. The use of abortion for fertility control was mentioned slightly more often by rural women (who already have a higher mean number of living children than urban women); women who reside outside of Baku, especially in the North-Northeast (71%) and Central (70%) regions of the country; and women over age 34 (76%), who also have more children. A woman's desire for no (more) children as a reason for abortion was strongly correlated with pregnancy order, from 11% among women pregnant for the second time to 56% among women with two previous pregnancies and 71 % among those with four or more previous pregnancies. Socioeconomic reasons were reported more often in urban areas, especially in Baku (27%), where the cost of living is more expensive and adequate housing is an increasing problem. 67 Partner's objection to pregnancy was an uncommon reason for the respondent's decision to not carry a pregnancy to term, regardless of the respondent's background characteristics, presumably because most women were married at the time of having the abortion and the couple was in agreement on the abortion decision. 68 CHAPTER 6 MATERNAL AND CHILD HEALTH Maternal and child mortality are measures of a nation's health and worldwide indicators of social well-being. The most recent World Health Organization (WHO) estimates for the newly independent states showed that the maternal mortality ratio (MMR) of 37 deaths per 100,000 live births in Azerbaijan in 1995 was higher than in Georgia and Armenia but substantially lower than in the Central Asian republics (Hill et al, 2001). According to the most recent official estimates, the MMR in 2000 was 37.6 deaths per 100,000 live births (State Committee of Statistics of the Azerbaijan Republic [SCS] , 2001). The same source places the infant mortality rate at 12.8 infant deaths per 1,000 live births. A recent nationwide UNICEF survey, however, estimated that both maternal and infant mortality are substantially higher (79 maternal deaths per 100,000 live births and 79 infant deaths per 1,000 live births, respectively) (UNICEF, 2000). Adequate perinatal care is an essential step in preventing, identifying, and addressing risk factors that may affect the health of mothers and their babies. Under the USSR health guidelines, women's access to perinatal care was free of charge and consisted of three components: preconception care, prenatal care, and postnatal care. Prenatal care visits included a comprehensive health assessment at the beginning of pregnancy and continuous surveillance of health status throughout the pregnancy. Preconception and prenatal care counseling was generally offered by primary care providers and consisted of provision of a wide array of information, including health risks associated with pregnancy itself and those that can affect the development of the fetus, such as tobacco and alcohol use, maternal infection (e.g., rubella, toxoplasma, and sexually transmitted infections [STIs]), and genetic conditions. Though very detailed, preconception counseling was offered only to young couples prior to marriage without any follow-up before their planned childbearing. Standard prenatal care (for uncomplicated pregnancies) required routine visits according to gestational age: monthly visits before 12 weeks of pregnancy; bi-monthly visits from 12 to 30 weeks of gestation; and weekly or bi-monthly visits thereafter. Prenatal care included a general health risk assessment consisting of medical examination and a series of laboratory tests (i.e., blood, urine, vaginal bacteriological exams, and screening for STIs and isoimmunization Rh) that were repeated periodically. Postpartum care was performed in parallel with infant care visits several times during the first year postpartum (Notzon et al., 1999). After its independence from the Soviet Union in 1991, Azerbaijan was no longer able to sustain a comprehensive perinatal care system, and many maternal and child health indicators started to deteriorate. 69 This chapter examines selected aspects of maternal and child care in Azerbaijan (e.g., sources of health care, utilization of maternal care services, breast-feeding), to identify subgroups with specific needs for care and to investigate maternal and child health outcomes that may be related to the availability and quality of maternity care services. All estimates reported here are based on respondents' reports recorded in the lifetime pregnancy history and a detailed birth history for all births carried to term since January 1996. 6.1 Prenatal Care Prenatal care is most effective when it is initiated in the early stages of pregnancy, is continued throughout gestation (according to recommended standards of periodicity), and is comprehensive (i.e., includes risk assessment, risk reduction or treatment of medical conditions, and counseling). This section describes the use of prenatal care among survey respondents for all pregnancies carried to term (either live births or still births) since January 1996. Women were asked in what week or month of gestation they had their first visit for prenatal care (not counting visits that were just for a pregnancy test or just for the delivery) and the number of prenatal care visits during pregnancy. Of the 3,430 births reported since January 1996, just over two-thirds of women (70%) had received some prenatal care; of those, about two-thirds (45% of 70%, or 64%) received their first prenatal care visit in the first trimester (Table 6.1.1). Approximately 1 in 5 women had the first visit during the second trimester; 6% had their first prenatal care visit during the third trimester. The level of any prenatal care within different subgroups varied sometimes by a considerable margin (between 53% and 89%). Rural women, residents of the South region, those who did not complete secondary education or had a low SES, and women who had already had two or more births were more likely not to have any prenatal care. Similarly, the percentage of mothers who entered prenatal care in the first trimester varied widely, from a low of 32% to a high of 65%. Fewer than 1 in 2 women (45%) reported early prenatal care. Women living in urban areas were more likely to start prenatal care earlier in pregnancy than women in rural areas were (55% vs. 35%). Early entry into prenatal care was highest among women living in Baku (61%) and lowest (34%) in the South. Early entry into prenatal care was highly correlated with the mother's education and SES; women who had not completed high school had a lower likelihood of initiating prenatal care early (32%) than did women with postsecondary education (60%). In addition, 47% of the women who had not completed high school reported receiving no prenatal care, whereas only 11% of women with a university education had no prenatal care. Similarly, women with low SES had a much lower likelihood of initiating prenatal care early in pregnancy than did women with high SES (35% vs. 65%). Internally displaced or refugee women (IDP/Rs) and women living in regions with a high concentration of IDP/Rs were slightly less likely 70 71 to report any prenatal care, probably reflecting the fact that these women have a lower SES than women living in areas not directly affected by the Nagorno-Karabach war. Mothers of low birth weight (LBW) babies were more likely than mothers who gave birth to normal weight babies to have no prenatal care (41% vs. 29%). Prenatal care should not only start early but also continue throughout pregnancy, according to recommended standards of periodicity. To assess the adequacy of prenatal care, it is necessary to monitor both the time of the first visit and the number of prenatal care visits once care has begun. Pregnancies ending in the 5 years prior to the survey averaged three prenatal visits; the range was 0 visits to 30 visits (data not shown). Among women with any prenatal care, the average number of prenatal care visits was 4.3. More than half of women with any prenatal care (39% of 70%, or 55%) had only 1-3 visits, and fewer than 1 in 10 women had 10 or more prenatal care visits (right panel of Table 6.1.1). A small proportion of women stated that they did not remember the number of prenatal care visits. Women who had 10 or more prenatal visits were generally the same women who started prenatal care early, because the number of visits was correlated with the month of initiation of care. The adequacy of prenatal care is assessed by using the Adequacy of Prenatal Care Utilization Index (APNCU), also known as the Kotelchuck index. This index combines the time of initiation of prenatal care (i.e., the month when prenatal care begins) with the number of visits received (according to American College of Obstetricians and Gynecologists recommendations). Inadequate care is defined as no or late prenatal care or less than 50% of recommended visits. The three remaining levels require early initiation of care (i.e., by the fourth month of gestation). Intermediate care requires 50%-79% of the recommended number of visits; adequate care, 80%-109%; and adequate care "plus," 110% or more of the recommended number of visits (Kotelchuck, 1994). By applying this index to data from the AZRHS01, only 6% of births within the past 5 years received adequate or adequate plus care, and 81 % received inadequate prenatal care (Figure 6.1.1). Standards of prenatal care for routine pregnancies in the era of the Soviet Union (still in effect in the Russian Federation but discontinued in many successor states) exceed U.S. standards, requiring early onset of prenatal care and an average of 14 prenatal care visits before delivery (Notzon et al., 1999). In the AZRHS01, the principal source of prenatal care was a women's consultation clinic (WCC) (46%). The second source of most prenatal visits was a maternity (36%) or a village hospital (13%). Rural dispensaries and private clinics provided prenatal care for 2% and 1%, respectively, of pregnant women; only 3% of women received prenatal care at home (data not shown). Thus, 97% of women reported having most of their prenatal care in a medical facility. Women's clinics were the principal source for prenatal care for all pregnancies, irrespective of women's background characteristics, except in rural areas and the South and South-West regions, where most prenatal care 72 was provided through hospitals. Generally, in WCCs and hospitals most care is provided by obstetricians (ob/gyns). Because WCCs and hospitals were the most used sources of prenatal care, most of the prenatal care visits (94%) were provided by ob/gyns; only 5% were provided by nurses or midwives (data not shown). Nurses provided as much as 38% of the prenatal care for the few women who received prenatal care at a rural dispensary or at home. Prenatal care is frequently inadequate in the countries of Eastern Europe and the successor states of the USSR. In recent reproductive health (RHS) and demographic and health (DHS) surveys conducted in the region, the proportion of pregnant women with no prenatal care was less than 1% in the Czech Republic, 1% in Moldova, 4% in Russia, 8% in Armenia, 9% in Ukraine and Georgia, between 2% and 5% in Central Asian Republics, 11% in Romania, and 30% in Azerbaijan. Late prenatal care is also common. With the exception of the Czech Republic, where more than 90% of women began receiving care in the first trimester, in all other countries less than three-fourths of women entered prenatal care early (Figure 6.1.2). Late prenatal care was more prevalent in the Caucasus region than in other regions. In the United States it is recommended that at least 90% of 73 pregnant women enter prenatal care early; in 2000, 83% of mothers began prenatal care in the first trimester while only 4% had no prenatal care or late care (third trimester) (CDC, 2002). Dissemination of health messages is an important component of prenatal care visits. In the absence of routine preconception care, the first prenatal visit is a critical opportunity to screen women for behavioral risk factors (e.g., tobacco and alcohol use), medical and genetic risks, and occupational risks and to provide comprehensive counseling. Counseling should include information about maternal behaviors and exposures that may affect the health of the fetus, nutrition, rest, and early signs and symptoms of pregnancy complications. In addition, approaching the time of delivery, counseling should prepare women for what they will face when giving birth, distribute accurate information regarding labor and delivery, and provide advice about techniques to reduce the pain and anxiety during labor. Also, counseling about breast-feeding and family planning after birth should be initiated during the prenatal period and reinforced during postpartum care. Because the initiation and frequency of prenatal care visits evaluate only the quantitative dimension of the prenatal care (i.e., adequacy of utilization of services), the AZRHS01 included additional questions aimed at assessing information received and measurements performed during the prenatal visits, that is, the adequacy of the content of prenatal care (Table 6.1.2). 74 75 76 Two-thirds of women who attended prenatal care clinics received some counseling about nutrition during pregnancy (66%); about 1 in 2 women received information about breast-feeding (55%), delivery (54%), potential complications during pregnancy and their early signs (49%), and postnatal care (47%); about 1 in 3 women received information about the negative effects of smoking and alcohol (38%); and only 27% of women received information about family planning after birth. Maternal characteristics that appear to be associated with lower levels of counseling for most of the topics include rural residence, residence in the South and South-West regions, less than complete secondary education, having three or more previous births, and receiving most of the prenatal visits in rural dispensaries or at home. The proportion receiving information during prenatal care visits was directly correlated with the number of prenatal visits. In addition to counseling, the first prenatal care visit should include a detailed medical history of the woman and her family, including information about risk, factors and genetic disorders; a detailed obstetrical history; a comprehensive physical examination; measurements of blood pressure; urinalysis; basic blood tests; ultrasound; and tests for various types of infection. Monitoring of mother's weight, blood pressure, and basic blood tests is extended during the follow-up visits, but ultrasound exams are carried out only two more times, at 16-28 weeks and at 36 weeks (Notzon et al., 1999). Despite these guidelines, prenatal care in the past 5 years did not always include required measurements (Table 6.1.3). About 1 in 6 women who had any prenatal care since January 1996 did not have her blood pressure measured at least once during prenatal care; 1 in 4 women did not have a blood exam, 1 in 3 women did not have an urine exam, and about 1 in 2 women never had her height or weight measured. Only about 1 in 3 pregnancies (37%) had had at least one ultrasound exam (Table 6.1.4). Maternal characteristics associated with higher levels of ultrasound exams include urban residence (51%), residence in Baku (73%), postgraduate education (62%), high socioeconomic status (SES) (66%), having seven or more prenatal care visits, and having most of prenatal visits in polyclinic. Lower prevalence of ultrasound exams was associated with rural residence (17%), livingin the South region (11%), and having most prenatal care at home (16%). Survey data do not allow us to differentiate between use of ultrasound for selected specific indications (e.g., confirmation of gestational age, assessment of fetal viability, fetal malformations, fetal growth, fetal presentation, multiple pregnancy, examination of the placenta, and assessment of amniotic fluid) and use for routine screening, either during early pregnancy (16-20 weeks) or in late pregnancy (after 20 weeks). Most women, however, had their first ultrasound exam before 20 weeks of pregnancy, suggesting the use of ultrasound for specific indications rather than for screening. Women in urban areas, including Baku; those with high educational attainment; those with seven or more prenatal care visits; and those whose primary source of prenatal care was a polyclinic or 77 78 maternity hospital were slightly more likely than other women to have their first ultrasound exam during the first 20 weeks of pregnancy. 6.2 Intrapartum Care All births should occur in medical facilities where adequately trained personnel can monitor the progress of labor and delivery. According to AZRHS01, most deliveries in the past 5 years took place in maternity wards (56%) or village hospitals with inpatient obstetrical care (17%) (Table 6.2.1 and Figure 6.2.1); however, 1 in 4 births were delivered outside medical facilities and less than 1% were delivered in a private clinic. Home deliveries were relatively high among rural residents (36%), those living in the Central, South, and South-West regions (35%, 36%, and 39%), those with low levels of education or low SES (39% and 36%, respectively), IDP/R women and non-IDP/Rs living in conflict-affected areas (41 % and 35%), those with four or more other births (42%), and those with no prenatal care (48%). 79 80 81 Seventy percent of births in the past 5 years were delivered by physicians, 19% by midwives or nurses, and the remaining 11 % by untrained birth attendants (data not shown). Virtually all deliveries that took place in maternities or private clinics were delivered by a physician (92%-94%), compared with two-thirds of deliveries in village hospitals and 22% of home deliveries. Deliveries not assisted by a health professional (i.e., a physician, midwife, or nurse) were more likely to occur in rural areas than in urban areas (17% vs. 6%), in the South and Central regions (19% and 16%), among women with less than complete secondary education (18%) and low SES (17%), and among IDP/R women (18%). The average time spent in a medical facility prior to delivery was about 7 hours; the range was from less than 1 hour to 4 days (Table 6.2.2). According to data published in the literature, the average duration of labor ranges from 6 hours (for multiparous women) to 10 hours (for nulliparous women) (Duig, 1975). Thus, many women, particularly those giving birth for the first time were admitted for delivery at or right after the onset of labor. The average time spent in the hospital prior to delivery did not vary greatly by mother's background characteristics. It was slightly shorter for less educated women, IDP/R women, and multiparous women. Women with any pregnancy complications and those who delivered by caesarean section (C-section) were more likely to report a long predelivery hospital stay, probably because they required closer monitoring of pregnancy, medical temporization of delivery, and use of C-section to end long labors. About half of women who gave birth in a medical facility were discharged in the first 4 days after delivery (54%), and 10% were discharged after 5 days (Table 6.2.2, right panel). One in four women (24%) was discharged after 6-7 days, and 12% of women spent 8 or more days in the hospital after delivery. Rural women, women with lower levels of education and low SES, non-IDP/R women living in conflict-affected areas, and women with three or more prior births were more likely to be discharged after a short postpartum hospital stay. As expected, women who delivered by C-section were more likely than women with vaginal deliveries to have hospital stays of 8 or more days (63% vs. 10%). The C-section rate varies considerably among countries, from about 5% to more than 20% of all deliveries. The optimal rate is not known, but little improvement in birth outcomes has been demonstrated if the rate is higher than 7%. In Azerbaijan, most births are delivered vaginally, and the prevalence of C-section among all deliveries between 1996 and 2001 was only 2.6% (Table 6.2.3). Before the dissolution of the USSR, the C-section rate for all Azerbaijan, though not routinely published, was estimated to range from 0.7% to 7% (Petrikovsky and Hoegsberg, 1990). 82 83 Women residing in urban areas were about 5 times as likely to have a cesarean delivery as women residing in rural areas. Women aged 35 or older reported C-section rates higher than those of women aged 34 or younger (see Table 6.2.3). The C-section rate increased directly with education and SES but was not significantly different among IDP/R and non-IDP/R women. Women who experienced prolonged labor were more likely to deliver by C-section than were women with labor of normal duration. Births with labor duration of more than 20 hours (more than 14 hours for multiparous women) were almost 5 times more likely to be delivered by C-section than were births with a shorter duration of labor. Almost half of C-sections, however, were performed before the onset of labor. Respondents were asked to identify the most important reason for they had delivered by cesarean section (Figure 6.2.2). The most frequent reasons given by the respondents included: fetal malpresentation (16%), previous C-section (15%), prolonged labor (12%), C-section performed on request (11%), obstetric hemorrhage (8%), fetopelvic disproportion (7%), and fetal distress (7%); 1 in 5 women (21%) reported that the C-section indication was due to "other" factors. Women who lived in rural areas were twice as likely to report having a C-section due to prolonged labor (10% vs. 84 24%) and to obstetric hemorrhaging (7% vs. 14%). Women from the urban areas had a significantly higher number of C-sections due to having previous C-sections than rural women did (17% vs. 3%) (data not shown). 6.3 Postnatal Care After delivery, it is important to assess the health of both the mother and the infant and to provide counseling regarding breast-feeding, proper child care, nutrition, and family planning. The postnatal period is a critical time that allows the health care provider to evaluate the physical and psychological health of a new mother and her infant, to detect and treat postpartum complications, and to provide the counseling and support needed to address any specific problems related to child care and family planning. Under the USSR health system, pregnant women were required to report to rural dispensaries, village hospitals, and WCCs for their monthly checkups; moreover, they were required to give birth in maternity wards or rural hospitals, where mothers with uncomplicated deliveries remained for about 5-7 days following delivery. Within a few days after hospital discharge, a health professional would make a house visit to examine the baby and counsel the mother. If the child were healthy, a nurse would return to the mother's home weekly during the first month postpartum. After that, both mother and child would continue to be seen regularly (at 1, 3, 6, 9, and 12 months) by a physician or a nurse at the polyclinic for physical examination, routine measurements, immunization, and postpartum counseling (Notzon et al., 1999). Perinatal care in Azerbaijan has changed since the dissolution of the Soviet-controlled system. Although the standards of care inherited from the Soviet system are still in effect, their application is less rigorously enforced. The AZRHS01 identified that 26% of births took place at home. The medical assistance at birth and location of delivery can greatly affect the health of mothers and the babies' chance of survival, particularly for LBW babies. Even in uncomplicated deliveries, giving birth at home reduces the likelihood of postnatal care for both the mother and her baby. The AZRHS01 provides information about the use of postnatal care and the content of postnatal counseling (Table 6.3.1). Postnatal care was substantially less utilized than prenatal care (25% vs.70%) (see also Table 6.1.1). Its use was very low among both urban and rural women (27% and 23% respectively), increased slightly with maternal education, and was the highest among women living in households with high SES. Postnatal care was highest among first-time mothers and lowest among women with two or more births. Lower utilization of maternal care services among high- parity women has long been recognized and explained through greater responsibilities within the household related to child rearing compounded with greater confidence and experience among this group of women. The use of C-section for delivery was associated with much higher rates of 85 86 87 postnatal care use (45%), probably because of the overlap with postsurgical care (data not shown). Most women who received postnatal care were counseled about child immunizations (71%), child care (66%), nutrition (66%), and breast care (63%). However, counseling about planning for future pregnancies and contraception (34%) and about breast-feeding (16%) were significantly lower. The type of health advice given during postnatal care did not vary significantly with maternal characteristics, except for women who had had their fourth child, which lowered the percentage of information presented during the postnatal care visit. A direct relationship was found between education level and the percentage of women reporting receipt of information during the postnatal period for each characteristic except breast care and breast-feeding. As mentioned previously, the timing of the first postpartum visit is supposed to be during the first week after the hospital discharge. Most women surveyed (72%) reported having a postnatal care visit within the first week after delivery (Table 6.3.2). This table also shows that 8% reported having a postnatal care visit 1 to 2 weeks after delivery and 18% having a postnatal care visit 2 or more weeks after delivery. Some women reported not remembering when they had their postnatal visit (1.5%). The questionnaire asked each mother if and when a health professional checked the baby's health after delivery (Table 6.3.3); 62% of the babies were seen by a health professional soon after hospital discharge. Urban women (75%), particularly those living in Baku (80%), were much more likely than rural women to receive those services (49%). Women from the South, South-West and West areas were less likely to have a high well-baby clinic's attendance. Women who had a university degree (75%) and women who reported a high SES (79%) were most likely to take their baby to a health professional to be examined. Women who delivered by a C-section were more likely to report taking their child for a examination than were women who delivered vaginally (84% vs. 62% respectively) (data not shown). Of the women who took their newborn to a health professional to be examined, 53% took their child within the first week of delivery, 25% took their child within 1 or 2 weeks after delivery, 21% went to visit the health care professional after 2 weeks, and a small percentage (1.5%) did not remember when they took the newborn to a well-baby clinic (Table 6.3.4). The proportion of women who registered their newborns was 86% (Table 6.3.4). Eighteen percent (urban) and 17% (rural) of women registered their child within the first week after delivery. Most women registered their child 1 or more weeks after delivery. Forty-six percent of women registered their child 1 to 4 weeks after delivery, and 37% registered their child more than 4 weeks after delivery. 88 89 90 6.4 Smoking and Drinking During Pregnancy The use of tobacco and alcohol during pregnancy is a major risk factor for poor pregnancy outcomes. Smoking during pregnancy has been linked to LBW infants, preterm deliveries, sudden infant death syndrome, and respiratory problems in newborns. The damaging effects of alcohol use during pregnancy include fetal growth retardation, mental retardation, physical abnormalities (especially in facial features), and altered neonatal behavior. Developmental abnormalities occur in approximately 35%-40% of infants born to alcoholic mothers and are associated with consumption of at least two drinks per day (Coles, 1993). Only 0.5% of births during the 5 years prior to the survey occurred to mothers who were smokers at the time they discovered they were pregnant. Only 0.4% continued to smoke during their pregnancy (data not shown). The proportion of women who smoked prior to getting pregnant or during pregnancy was slightly higher in the areas of Baku and the North-Northeast areas of Azerbaijan (1% and 0.9%). The highest smoking prevalence prior to pregnancy was reported by women aged 40-44 (4.8%); 1.3% of the women who reported having an LBW baby reported smoking prior to knowledge of the pregnancy and during the pregnancy (data not shown). Similar to the low prevalence of women smoking during pregnancy was the percentage of women drinking during pregnancy. Only 0.8% of women reported drinking during their pregnancy. Urban women were more likely to report drinking during pregnancy than rural women were (1.4% vs. 0.2%). Women with a university degree had higher prevalence of drinking during pregnancy than women in the other education categories (2.6%); women with high SES (3.3%) reported drinking more often during pregnancy than did the women of low and medium SES (0% and 1.2%, respectively) (data not shown). 6.5 Pregnancy and Postpartum Complications Routine measurement of blood pressure is an essential component of health risk assessment during prenatal visits. However, as is the case with other health measurements and diagnostics, self-reports of medical conditions may reflect a combination of risk factors and differences in reporting. In particular, the data suggest a higher likelihood of complete reporting of health problems from individuals with better access to medical care. As shown in Table 6.5.1, most women with births in the past 5 years (82%) reported measurement of their blood pressure during pregnancy, and one-fifth (19%) were identified as having high blood pressure (HBP). Only 1 % were hospitalized due to HBP. 91 Measurement of blood pressure was less common among women who lived in a rural area or in the South or South-West regions, had not completed secondary education, had a low SES, and had received prenatal care in a village hospital. The prevalence of reported HBP during pregnancy was highest among women who lived in the Southwest region (23%), who were age 35 or older (28%), had less than a university education, and had a medium SES (23%). Women who received prenatal care in a rural clinic or village hospital were more likely to report HBP than were women who received prenatal care at other sites. Twenty-five percent of women with recent births reported pregnancy complications requiring medical attention (Table 6.5.2).The conditions mentioned most often were anemia (8%), edema/water retention (8%), bleeding (96%), and HBP (5%). Pregnancy complications that required medical attention were slightly more prevalent among women residing in urban areas than in rural areas (27% vs. 22%) and were more prevalent among first-order births than among third or higher order births (29% vs. 21%). Women who lived in the southern region of the country (16%), were aged 20-24 (21 %), who had completed less than a university degree, or who were of low SES (23%) reported fewer pregnancy complications. In contrast to other former Soviet Union countries, hospitalization rates for pregnancy complications were low. Only 3% of Azeri women with pregnancy complications reported that they had been hospitalized for those conditions, whereas 30% of such women in Moldova, 32% in Ukraine, and 50% in Russia reported hospitalization (data not shown). A total of 38% of women reported at least one postpartum complication (Table 6.5.3). Reported postpartum complications ranged from 28% of women experiencing severe uterine pain to 1% of women experiencing breast infections. Reports of postpartum complications were less than 38% only among residents in the North-Northeast regions of Azerbaijan (29%). The highest reports of postpartum complications came from women with a maternal age of 15-24 (41%) and from women who had had any pregnancy complication (50%) or prolonged labor (48%). 92 93 94 95 6.6 Poor Birth Outcomes Of all births during the 1996-2001 period, 21.2 per 1,000 were stillbirths (Table 6.6). The stillbirth rate was highest among women living in urban areas, residents of the West and Central regions, women aged 35-44, women with postsecondary education, and women with three or more previous births. The stillbirth rate did not vary significantly by the IDP/R status. Complicated pregnancies that required hospitalization were significantly more likely to end in a stillbirth than were uncomplicated pregnancies (42 per 1,000 vs. 15 per 1,000). Compared with normal labor, prolonged labor (i.e., more than 20 hours for nulliparous women and more than 14 hours for multiparous women) was associated with a more than 2 times higher prevalence of stillborns (52 per 1,000 vs. 20 per 1,000). The total LBW rate (defined as the percentage of live births with birth weight less than 2,500 g) for infants born alive during the 1996-2001 period was 12%. However, higher rates were found among rural women; women in the South-West, South, and Central regions (18%, 15%, and 14%, respectively); women with low education (16%) or low SES (16%); IDP/R and non-IDP/R women living in conflict affected areas (16%); women with at least three prior births; women with no prenatal care (17%); and those who delivered at home (18%). Women who had prolonged labor (18%) or who delivered by C-section prior to the labor induction (24%) were more likely to report LBW, but their numbers are small. Interestingly, most LBW babies were delivered at term (data not shown). The reported prematurity rate (defined as the percentage of live births delivered before 37 weeks of gestation) for the same time period was 4.7%. Higher reported prematurity rate was associated with "no labor" (21%), pregnancy complications (8%), older age (10%), and high SES (9%). 96 97 6.7 Breast-Feeding The AZRHS01 included questions about breast-feeding patterns and duration for all children under 5 years of age. As shown in Table 6.7.1, most babies (95%) born during 1996-2001 were breast-fed for at least a short period of time. The percentage of babies ever breast-fed varied little by selected characteristics. Rates of breast-feeding were slightly lower among women living in urban areas, including Baku; women living in the Central region; and women living in households with a high SES. Infants who were delivered by C-section had a lower rate of breast-feeding than did those delivered vaginally (82% vs. 95%). LBW babies were slightly less likely to be breast-fed than were those with a birth weight of 2,500 g or more (92% vs. 95%), but the difference is not statistically significant. According to WHO recommendations, early suckling (i.e., within the first hour postdelivery) should be promoted after all spontaneous deliveries. Table 6.7.1 (right panel) includes the time elapsed between delivery and initiation of breast-feeding. Of infants who were breast-fed, only 13% began breast-feeding during the first hour after birth; Most children began breast-feeding between 2 hours after birth and the completion of the first day (38%) or during the second day of life (26%). About 1 in 5 babies (23%) began breast-feeding only after 48 hours of life. Breast-feeding initiation within the first hour was higher among rural than among urban women (17% vs. 9%), was inversely related to the SES of the mother, and increased with birth order. Caesarean delivery substantially reduced the likelihood of early breast-feeding. For infants delivered by C-section, breast-feeding was more likely to be initiated after 2 days, if ever. An infant is "exclusively" breast-fed if he or she receives only breast milk and is "almost exclusively" or predominantly breast-fed if he or she receives breast milk accompanied by water or other liquids (except other types of milk). Children with exclusive or almost exclusive breast-feeding are considered to be "fully" breast-fed (Labbok and Krasovec, 1990). These indicators are recommended by WHO to assess the adequacy of breast-feeding practices in a population and allow for comparisons with findings from other countries. The WHO recommendations state that "all infants should be fed exclusively on breast milk from birth to 4-6 months of age" and that some breast-feeding should be maintained until the child is at least 1 year old (WHO, 1991). The proportion of children under 5 years old still being breast-fed at the time of the survey was calculated by single month of age (0-59 months); the denominator included all live births in those 5 years, regardless of survival (Table 6.7.2). Those proportions were summed together to calculate the mean duration of breast-feeding. This method is known as the "current status mean" method (WHO, 1991). Durations of exclusive and full breast-feeding were calculated in the same way. 98 99 100 The mean duration of any breast-feeding was 11.6 months. For most of this time, however, breast- feeding was only partial. The mean duration of exclusive breast-feeding was 0.4 month and, with the exception of women residing in the Central region, did not vary greatly by maternal characteristics. Women in the South-West and West regions, those who gave birth after age 34, and IDP/Rs had lower mean durations. In addition, babies delivered by C-section, those with LBW babies, and babies who initiated suckling 48 hours or more after birth had lower mean durations of exclusive breast-feeding. Thus, few children in Azerbaijan were exclusively breast-fed for the minimum 4- month period recommended by WHO. 6.8 Infant and Child Mortality Although higher than in most of the former Soviet-bloc countries of the Central and Eastern Europe, the infant mortality rate (IMR) in Azerbaijan is comparable to the rates reported by Central Asian Republics (Table 6.8.1). Mortality rates from government sources, however, tend to underestimate 101 the real IMR, sometimes by a considerable margin. Recent RHS and DHS surveys conducted in the region estimate much higher rates than the official reports, particularly in the Caucasus region and Central Asian Republics. Survey-based infant mortality estimates exceeded the official rates by 10% in Ukraine; by 50%-70% in Romania, Uzbekistan, and Georgia; by more than 100% in Armenia, Kazakhstan, Kyrgyzstan and Turkmenistan; and by more than 300% in Azerbaijan. Concerns about the reporting of the IMR in Caucasus and Central Asia were raised by the former Soviet Central Statistical Administration, which felt the need to apply correction factors to the data reported by those regions in order to ensure comparability with the European republics of the USSR (Andreyev and Ksenofontova, 1991). The registration process of infant deaths in Azerbaijan is similar to that of other former Soviet Union countries. Currently two official counts of infant deaths exist: one through the Center for Medical Statistics and Information of the Ministry of Health (MOH) and one through the State Committee for Statistics of Azerbaijan (SCS). The MOH receives infant mortality data on aggregate data forms from medical facilities (monthly from maternity hospitals and annually from pediatric wards and polyclinics). The SCS receives mortality data from urban and rural civil registry bureaus; the medical death certificates filled out by physicians in hospitals or ambulatory units are submitted by relatives in order to obtain official death certificates needed for burial. The original medical death certificates are submitted to SCS, which processes demographic data and applies cause-of-death codes according to the 10th revision of the International Classification of Diseases (WHO, 1993). It is unclear whether one system is more accurate than another, because both are subject to potential misclassification and underreporting. For example, medical facilities may feel pressured to misclassify very premature babies as miscarriages, because IMR is a classic indicator of evaluating performance of health units. The SCS may be subject to underregistration of both births and deaths because the declaration of these events is performed by a third party (usually a parent), usually within the first 3 months after birth (Jone and Grupp, 1983). Thus, infants born alive who subsequently die before having birth certificates issued would remain unregistered as either live births or infant deaths. The likelihood of escaping registration may be the highest when the death occurs soon after birth and decreases with the increase in the infant's age. One of the principal objectives of the AZRHS01 was to estimate levels and trends in infant and child mortality. The survey included a series of questions to obtain for each live birth the date of occurrence, sex of the child, survival status and, for children who had died, the age at death. This information allows a direct calculation of infant and child mortality rates for precise periods of time, by means of life tables. Survey data were used to calculate mortality levels among respondents' children, namely, infant mortality (i.e., deaths before the first birthday), child mortality (i.e., deaths between 12 and 59 completed months of age), and child-under-5 mortality (i.e., deaths before the 102 fifth birthday). Infant mortality was further divided into two periods: neonatal (0-28 days) and postneonatal (29 days to 11 completed months). In the 10-year interval from January 1991 to December 2000, the IMR was estimated at 80.8 per 1,000 live births, and the neonatal and postneonatal mortality rates were 38.1 per 1,000 and 42.7 per 1,000, respectively (see Table 6.8.2 and Figure 6.8). In interpreting these results, the reader should bear in mind that survey data tend to underestimate neonatal mortality to a greater extent than child mortality at older ages. When the death occurs in the first few hours or days of life, some women, especially those with low levels of education and those who have had many births, do not always report their losses as infant deaths because they may not consider their births to be live births. For this reason, the estimated neonatal mortality rate of 38.1 and, implicitly, the IMR of 80.8 should be considered as minimum values for this period of time. The IMR of 80.8 per 1,000 live births estimated from AZRHS01 for the most recent 10-year period was 3.8 times higher than the average rate of 21.5 infant deaths per 1,000 live births reported by MOH (Table 6.8.2). The statistical standard error (SE) for the survey period estimate was 4.5 percentage points (calculated as SE= [rate/square root of number of deaths]*1.4 where 1.4 103 represents the design effect needed because the AZRHS01 used a cluster sampling design. Standard errors can be used to calculate confidence intervals around the IMR within which we can say with 95% confidence that the true value of the population IMR lies. Thus, the point estimate of 80.8 per 1,000 should not be considered as the exact value of the IMR; in theory, that would have been possible to calculate if all women of reproductive age had been interviewed. The true rate could be higher or lower; its value lies between a 95% confidence interval from 72.0 to 89.6 per 1,000 (CI=± 1.96*SE).The lower limit of 72.0 was still 3.3 times higher than the average of 21.5 per 1,000 reported by the SCS for 1991-2000. The survey neonatal death rate of 38.1 per 1,000 was almost 10 times higher than the average official rate of 4.2 per 1,000 for 1991-2000 (ranging from 5.3 per 1,000 in 1991 to 3.2 in 2000). Similarly, the survey found postneonatal mortality of 42.7, which was 2.5 times higher than the official average of 17.3 per 1,000 (ranging from 24 per 1,000 in 1993 to 9.6 in 2000). Thus, the difference between 104 the survey estimates and the official rates was observed in both neonatal and postneonatal mortality rates, but much more for neonatal deaths. As a result, neonatal deaths contributed to only 19% of the IMR for 1991-2000, according to MOH data, whereas neonatal deaths in respondents' histories accounted for almost half (47%) of the reported deaths during the first year of life. In conclusion, the survey estimates of neonatal deaths were substantially higher than the official data. Presumably, high underreporting of these deaths exists within the vital records reporting system. Part of the gap between the official rates and survey estimates can be explained through differences in definition of live birth. Respondents in AZRHS01 were asked to report pregnancy outcomes (e.g., stillbirths and live births) according to international definitions that were recently adopted by Azerbaijan. Thus, a live birth was defined as any infant born alive, irrespective of the duration of the pregnancy, that breathes or shows any other signs of life after separation from the mother. The time interval for which survey mortality rates were calculated extends, however, over a period when the former Soviet Union definition of a live birth was largely in use. Under that definition, any infant with signs of life present at the time of delivery but whose weight was less than 1,000 g, had gestational age less than 28 weeks, or measured less than 35 cm and died within the first 7 days of life was classified as a miscarriage or stillbirth (Velkoff and Miller, 1995; Anderson and Silver, 1986). Thus, by applying the WHO definition, the survey recorded a certain number of births with a relatively low survival probability as live births, whereas the same births may have been misclassified in the official statistics as late fetal deaths, particularly if they occurred in the early 1990s. However, the difference in definition of what constitutes a live birth should only affect mortality rates for the first 7 days of life (i.e., early neonatal mortality), whereas estimates for postneonatal or child mortality should not be affected. If the mortality rates presented in this chapter excluded very preterm babies (i.e., pregnancies terminating at 28 or more weeks of gestation or less) who died during the first week of life, the resulting IMR would be about 2 percentage points lower than if the WHO definition were applied (78.9 vs.80.8 deaths per 1,000 live births); the entire difference between IMRs calculated with the two definitions is due to the decrease in neonatal mortality, from 38.1 per 1,000 to 36.0 per 1,000 (data not shown). Thus, differences in definition clearly do not account for the gap between survey estimates and official rates. Alternatively, underreporting of births and infant deaths to civil registries (particularly for infants who were not delivered in medical facilities) and uneven quality of reporting from local vital record offices to the central level may play a more important role than differences in definition. 105 106 As shown in Table 6.8.2, both IMR and the under-5 mortality declined in the most recent 5-year period compared with the period 1991-1995, consistent with downward trends documented by the official reports. The decline in neonatal mortality was more substantial than the decline in postneonatal mortality (17% vs. 10%), perhaps indicating better intrapartum and neonatal care in the most recent period. Neither infant nor under-5 mortality rates differed significantly by mother's residence. Mortality differentials by age of the mother at the time of birth showed that the highest infant and under-5 mortality rates were found among births to women aged 30 or older (93.8 per 1,000 and 112.6 per 1,000, respectively). Infant mortality, classified by education level of the mother, was highest among mothers without postsecondary education. The greatest differentials were observed in the levels of postneonatal mortality: Children born to women without postsecondary education were at least 6 times more likely to die between 28-364 days of age than were children born to women with a university education. Better access to preventive and curative health care services and better living standards among women with higher levels of education are probably the likely explanation for the difference. Infant mortality differentials by IDP/R status illustrate that the rates for infants born to IDP/R women and non-IDP/R women living in conflict-affected areas were almost 50% higher than among non- IDP/R infants living in areas not directly affected by war. No significant differences were found in child mortality by IDP/R status. Thus, the difference in child under-5 mortality rates (which were almost 40% higher among infants born to IDP/R women and non-IDP/R women in conflict-affected areas than among non-IDP/R women in non-conflict-affected areas) were entirely caused by differences in IMRs, particularly neonatal mortality rates. The infant and child mortality rate among infants born with birth order of four or higher was higher than among those preceded by two or fewer births. Male infant mortality (83.6 per 1,000) was slightly higher than the rate for females (77.8 per 1,000), reflecting the sex differential in neonatal mortality. After the neonatal period, however, female infants had higher probability of dying than male infants. In conclusion, the levels of infant and child mortality in Azerbaijan are quite high. The relatively high magnitude of this public health problem was not easy to document through official statistics until recently. To improve reporting, the Azerbaijan MOH began applying ICD-10 definitions in 2001 and introduced new birth, death, and perinatal death certificates through Order No. 100, October 2001 (MOH, 2001b). Moreover, the MOH issued specific instructions for medical facilities on how to report such events (Order No. 88, July 2002) and on how to improve medical statistical reporting (Order No. 137, October 2002) (MOH, 2002a and 2002b). 107 108 CHAPTER 7 NUTRITIONAL STATUS OF MOTHERS AND CHILDREN Childhood malnutrition in the population is generally estimated using the indices of height-for-age, weight-for-height, and weight-for-age. A high prevalence of low height-for-age (less than -2.00 height-for-age z score, often referred to as shortness ox stunting) is usually an indication of chronic malnutrition in a population. Although chronic malnutrition can be the result of long-term food shortages and disease, it can also be the result of poor socioeconomic conditions. A high prevalence of low weight-for-height (less than -2.00 weight-for-height z-score, referred to as thinness or wasting) is an indicator of acute malnutrition. Acute malnutrition is usually a reflection of recent food shortages, infections, or diarrhea. General malnutrition is usually indicated by a low weight-for- age z score and can be the result of long-term poor living conditions or the result of a more acute situation. However, a low prevalence of low weight-for-height is most likely a reflection of more chronic conditions (World Health Organization [WHO], 1995). 7.1 Methodology To assess the nutritional status of children in the population, all children ages 3-59 months in sampled households were measured and weighed at their place of residence. Measurements were taken by nutrition health professionals who had received intensive training on anthropometry assessment and hemoglobin measurements prior to the beginning of fieldwork. Height was measured using the Shorr length/height board for children 3-59 months. The height of children age 2 years or older was measured in the standing position (without shoes), and the recumbent length of children younger than 2 years was measured. The results were recorded to the nearest 0.1 cm on the questionnaire. The weight of all children (wearing only undergarments) was measured using the Uniscale, which is designed to measure adults and children of all ages. Infants and toddlers under age 2 years were weighed with the mother holding the child on the Uniscale (the mother's weight was automatically deducted to get the child's weight). The weight was read and recorded to the nearest 0.1 kg. The child's age was calculated from the mother's report of the child's month and year of birth. To assess how the nutritional status of children differed from what would be expected in a reference population, the indices were compared with the international growth reference developed by the U.S. 109 National Center for Health Statistics-Centers for Disease Control and Prevention and WHO (WHO, 1995) to obtain a standardized z score. The NCHS/CDC/WHO reference is based on growth data for healthy children in the United States. This reference has been chosen because studies have shown that well-nourished children from most countries follow a similar pattern of growth to that of the reference population. On the basis of this reference, children were classified as malnourished if they had a z score level less than 2 standard deviations below the mean of the reference population (i.e., -2.00 SD) for any indicator. Children who had a z score of less than -3.00 were classified as severely malnourished. Anthropometry data were available on 2,446 children ages 3-59 months. However, the final sample size for each indicator changed once those who had missing or out-of-range z scores were excluded (height-for-age n=2,426, weight-for-height n=2,435, and weight-for-age n=2,442). Using the standard deviation of the mean z scores as a measure, the data quality was quite good: height-for-age, mean z score = -0.83 (1.1 SD); weight-for-height mean z score = -0.04 (0.96 SD); weight-for-age mean z score = -0.058 (1.03 SD). To assess the prevalence of anemia in children in the population, a capillary blood sample was taken by fingerstick for all children aged 12 to 59 months. Written informed consent was obtained from the mother before samples were taken. Hemoglobin measurements were completed immediately after sample collection using a portable HemoCue® hemoglobinometer in the home. Results for each child tested were read and recorded on the mother's questionnaire. Anemia was defined as a hemoglobin of less than 11.0 g/dL, using the WHO anemia criteria for children younger than 5 years old (WHO, 2001) The mother was told immediately if the child was anemic. To assess the nutritional sta

View the publication