Trinidad and Tobago - Demographic and Health Survey - 1987

Publication date: 1987

Trinidad and Tobago Demographic and Health Survey 1987 @ Family Planning Association of Trinidad and Tobago ~I)HS Demographic and Health Surveys Institute for Resource Development/Westinghouse REPUBLIC OF TRINIDAD AND TOBAGO Trinidad and Tobago Demographic and Health Survey 1987 Kenneth Heath Dona Da Costa-Martinez Amy R. Sheon Family Planning Association of Trinidad and Tobago Port-of-Spain, Trinidad and Institute for Resource Development/Westinghouse Columbia, Maryland USA November 1988 This report presents the findings of the Trinidad and Tobago Demographic and Health Survey, implemented by the Family Planning Association of qq'inidad and Tobago in 1987. The survey is part of the worldwide Demographic and Health Surveys (DHS) Program, which is designed to collect data on fertility, family planning, and maternal and child health. Additional information on this survey can be obtained from the Family Planning Association of Trinidad and Tobago, Corner of Charlotte and Oxford Streets, Port-of-Spain, Trinidad, W.I. (Telephone: 809-623-4764). The Trinidad and Tobago Demographic and Health Survey was carried out with the assistance of the Institute for Resource Development (IRD), a subsidiary of Westinghouse Electric Corporation, with offices in Columbia, Maryland. Funding for the survey was provided under a contract with the U.S. Agency for International Development (Contract No. DPE-3023-C-00-4083-00). Additional information about the DHS program can be obtained by writing to: DHS Program, IRD/Westinghouse, 8850 Stanford Boulevard, Suite 4000, Columbia, MD 21045, USA (Telex: 87775, Fax: 301-290-2999, Telephone: 301-290-2800). TABLE OF CONTENTS Page TABLE OF CONTENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . lii IJ,~T OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi CHAPTER 1 BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1 History, Geography, Economy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.3 Population and Family Planning Policies and Programmes . . . . . . . . . . . . . . . . 3 1.4 Health Priorities and Programmes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.5 Objectives of the Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.6 Organisation of the Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.7 Background Characteristics of Survey Respondents . . . . . . . . . . . . . . . . . . . . 5 CHAPTER 2 NUPTIALITY AND EXPOSURE TO RISK OF PREGNANCY . . . . . . . . . 9 2.1 Current Marital Status . . . . . . . ' . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.2 Median Age at First Union . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.3 Breastfeeding and Postpartum Insusceptibility . . . . . . . . . . . . . . . . . . . . . . . 11 2.4 Mean Duration of Breast-feeding and Postpartum Insusceptibility . . . . . . . . . . . 13 CHAFrER 3 FERTIL ITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 3.1 Current and Cumulative Fertility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 3.2 Fertility Trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 3.3 Children Ever Born . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 3.4 Children Ever Born and Age at First Union . . . . . . . . . . . . . . . . . . . . . . . . 21 3,5 Age at First Birth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 3.6 Median Age at First Birth by Background Characteristics . . . . . . . . . . . . . . . 23 CHAPTER 4 FERTIL ITY REGULAT ION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 4.1 Knowledge of Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 4.2 Problems with Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 4.3 Knowledge of Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 4.4 Ever Use of Contraception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 4.5 Current Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 4.6 Parity at First Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 4.7 Age at Sterilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 4.8 Knowledge of the Reproductive Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 4.9 Knowledge and Use of Pap Smears . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 4.10 Source of Contraceptive Methods and Satisfaction with Services . . . . . . . . . . 38 4.11 Discontinuation of Contraceptive Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 4.12 Attitude Toward Becoming Pregnant and Reasons for Nonuse of Contraception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 4.13 Intention to Use Contraception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 4.14 Exposure to Mass Media and Family Planning Messages . . . . . . . . . . . . . . . 43 4.15 Discussion of Family Planning with Partner . . . . . . . . . . . . . . . . . . . . . . . , 43 Page CHAPTER 5 FERTILITY PREFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 5.1 Fertility Preferences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 5.2 Need for Family Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 5.3 Ideal Family Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 5.4 Fertility Planning Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 CHAPTER 6 INFANT AND CHILD MORTAL ITY , AND HEALTH . . . . . . . . . . . . . . 57 6.1 Infant and Childhood Mortality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 6.2 Infant and Childhood Mortality by Socioeconomic Characteristics . . . . . . . . . . 57 6.3 Infant and Childhood Mortality by Demographic Characteristics . . . . . . . . . . . 59 6.4 Children Ever Bom and Surviving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 6.5 Antenatal Care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 6.6 Assistance at Delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 6.7 Immunization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 6.8 Diarrhoea Prevalence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 6.9 Diarrhoea Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 6.10 Knowledge of ORT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 6.11 Nutritional Status of Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 6.12 Nutritional Status of Children According to Height-for-Age . . . . . . . . . . . . . . 72 6.13 Nutritional Status of Children According to Weight-for-Height . . . . . . . . . . . . 73 6.14 Nutritional Status of Children According to Weight-for-Age . . . . . . . . . . . . . 75 6.15 Summary of the Nutritional Status of Children Aged 3-36 Months . . . . . . . . . 75 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 APPENDIX A SURVEY DESIGN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 A.1 Sample Design and Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 A.2 Questionnaire Design and Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 A.3 Fieldwork . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 A.4 Data Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 APPENDIX B SAMPLING ERRORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 APPENDIX C QUESTIONNAIRE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 ii LIST OF TABLES Table 1.1 Table 1.2 Table 1.3 Table 2.1 Table 2.2 Table 2.3 Table 2.4 Table 2.5 Table 2.6 Table 3.1 Table 3.2 Table 3.3 Table 3.4 Table 3.5 Page Population of Trinidad and Tobago at Census Years . . . . . . . . . . . . . . . . . 2 Distribution of Women 15-49 by Age, Residence, Education, and Ethnicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Percent Distribution of Women by Education, According to Selected Background Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Percent Distribution of Women by Ctm, ent Union Status, According to Background Characteristics, TTDHS 1987 . . . . . . . . . . . . . . . 9 Percent Distribution of Women by Age at First Union and Median Age at First Union, According to Background Characteristics, 'ITDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Median Age at First Union Among Women Age 20-49 Years. by Current Age and Selected Background Characteristics, qTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Percentage of All Births in the Last 5 Years Who Have Ever Been Breastfed, According to Selected Background Characteristics of the Mother, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . 13 Percentage of Births Whose Mothers are Still Breastfeeding, Postpartum Amenorrhoeic, Abstaining, and Insusceptible, by Number of Months Since Birth, TrDHS 1987 . . . . . . . . . . . . . . . . . . . . 14 Mean Number of Months of Breastfeeding, Postpartum Amenorrhoea, Postpartum Abstinence, and Postpartum Insusceptibility by Selected Background Characteristics, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Total Fertility Rates (TFR) for Calendar Year Periods and for Five Years Preceding the Survey, and Mean Number of Children Ever Born (CEB) to Women 40-49 Years of Age, by Selected Background Characteristics, 'ITDHS 1987 . . . . . . . . . . . . . . . 17 Percent of AU Women Who are Currently Pregnant by Age. TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Age-Period Fertility Rates (per 1,000 Women) by Age of Woman at Birth of Child, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . 20 Percent Distribution of All Women and Women in Union by Number of Children Ever Born, According to Age, TFDHS, 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Mean Number of Children Ever Born to Women Ever in Union by Age at First Union and Years Since First Union, TTDHS, 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 iii Table 3.6 Table 3.7 Table 4.1 Table 4.2 Table 4.3 Table 4.4 Table 4.5 Table 4.6 Table 4.7 Table 4.8 Table 4.9 Table 4.10 Table 4.11 Table 4.12 Page Percent Distribution of Women by Age at First Birth According to Current Age, TTDHS, 1987 . . . . . . . . . . . . . . . . . . . . . . . 22 Median Age at First Birth Among Women Aged 25-49 Years, by Current Age and Selected Background Characteristics, TFDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Percentage of All Women and of Women in a Union Knowing Any Method, Any Modem Method, and Specific Contraceptive Methods, by Age, TTDHS, 1987 . . . . . . . . . . . . . . . . . . . 26 The Percentage of Women in Union Who Know at Least One Modem Method, by Number of Living Children and Selected Background Characteristics, "ITDHS 1987 . . . . . . . . . . . . . . . . . 27 Percent Distribution of Women Who Have Ever Heard of a Contraceptive Method by Main Problem Perceived in Using the Method, TFDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Percent Distribution of Women Who Have Ever Heard of a Contraceptive Method by Supply Source Named, According to Specific Method, "ITDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Percentage of All Women and Women in Union Who Have Ever Used Any and Specific Contraceptive Method, by Specific Method and Age, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . 29 Percent Distribution of All Women and Women in Union by Contraceptive Method Currently Used According to Age, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Percent Distribution of Women in Union by Contraceptive Method Currently Used, According to Selected Background Characteristics, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Percent Distribution of Women Ever in Union by Number of Living Children at Time of First Use of Contraception, According to Current Age, TTDHS, 1987 . . . . . . . . . . . . . . . . . . . . . . . 35 Percent Distribution of Sterilized Women by Age at Time of Sterilization, According to the Number of Years Since the Operation, qTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Percent Distribution of All Women and Women Who Have Ever Used the Safe Period by Knowledge of the Fertile Period During the Ovulatory Cycle, TTDHS 1987 . . . . . . . . . . . . . . . . . 36 Percentage of Women Ever in Union Who Know About the Pap Smear, Have Ever Had a Pap Smear, and Have Had a Pap Smear in the Last Year, According to Selected Background Characteristics, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . 37 Percent Distribution of Current Users by Most Recent Source of Supply, According to Specific Method, TTDHS 1987 . . . . . . . . . . . . . 38 iv Table 4.13 Table 4.14 Table 4.15 Table 4.16 Table 4.17 Table 4.18 Table 4.19 Table 4.20 Table 4.21 Table 5.1 Table 5.2 Table 5.3 Page Percent of Current Users of Modem Methods Who Said There Were No Problems With the Service, by Type of Source Last Visited. "I'rDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Percent Distribution of Women who Have Discontinuee a Contraceptive Method in the Last Five Years by Main Reason for Last Discontinuation, According to Specific Method, TrDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Percent Distribution of Nonpreguant Women Who Have Ever Had Sexual Intercourse and Who Are Not Using Contraception by Attitude Toward Becoming Pregnant in the Next Few Weeks, According to Number of Living Children, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Percent Distribution of Nonpregnant Women Who Are Sexually Active and who Are Not Using Contraception, and Who Would be Unhappy ff they Became Pregnant by Main Reason for Nonuse, According to Age, TrDHS 1987 . . . . . . . . . . . . . . . 41 Percent Distribution of Women in Union Who Are Not Currently Using Any Contraceptive Method, by Intention to Use in the Future, According to Number of Living Children, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Percent Distribution of Women in Union who Are Not Using a Contraceptive Method but Who Intend to Use in the Future by Preferred Method, According to Whether They Intend to Use in the Next 12 Months or Later, "I'rDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Percentage of Women Whose Households Have Selected Mass Media, Who Are Exposed to Media Regularly, and who Were Recently Exposed to Family Planning on the Mass Media, by Selected Background Characteristics, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Percentage of All Women Who Believe That it is Acceptable to Have Messages About Family Planning on the Radio or TV, by Age and Selected Background Characteristics, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Percentage of Women in a Union and Who Know Any Method of Contraception who Have Discussed Family Planning With a Partner at Least Once in the Past Year, by Age and Selected Background Characteristics, TTDHS 1987 . . . . . . . . . . 45 Percent Distribution of Women in Union by Desire for More Children, Timing of Next Birth, and Sterilization Regret According to Number of Living Children, TTDHS 1987 . . . . . . . . . . . . . 48 Percent Distribution of Women in Union by Desire for More Children, According to Age, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . 49 Percemage of Women in Union Who Want No More Children by Number of Living Children and Selectee Background Characteristics, TTDHS, 1987 . . . . . . . . . . . . . . . . . . . . . . 50 V Table 5.4 Table 5.5 Table 5.6 Table 5.7 Table 5.8 Table 5.9 Table 6.1 Table 6.2 Table 6.3 Table 6.4 Table 6.5 Table 6.6 Table 6.7 Page Percentage of Women in Union Who Are in Need of Family Planning and the Percentage Who Are in Need and Who Intend to Use Family Planning in the Future by Selected Background Characteristics, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . 51 Percent Distribution of All Women by Ideal Number of Children; Mean Ideal Number of Children for All Women and for Women in Union, Percentage of All Women Whose Current Number of Children Exceeds Ideal Number, According to Number of Living Children, I IDHS 1987 . . . . . . . . . . . . . 52 Mean Ideal Number of Children for All Women by Age and Selected Background Characteristics, "ITDHS 1987 . . . . . . . . . . . . . . . . . 53 Percent Distribution of All Births in the Last Five Years by Contraceptive Practice and Fertility Planning Status, According to Birth Order, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . 54 Percent Distribution of Births in the Year Before the Survey by Fertility Planning Status, According to Birth Order, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Total Wanted Fertility Rate, Total Fertility Rate for the Five Years Preceding the Survey, and Percentage of the Total Fertility Rate Identified as Unwanted Births by Selected Background Characteristics, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . 55 Infant and Childhood Mortality for Five-Year Calendar Periods, T'FDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Infant and Childhood Mortality 1977-1987 by Selected Background Characteristics of Mother, TTDHS 1987 . . . . . . . . . . . . . . . 58 Infant and Childhood Mortality 1977-1987 by Selected Background Characteristics, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . 60 Mean Number of Children Ever Born, Surviving, and Dead, and Proportion of Children Dead Among Children Ever Born, by Age of Mother, 'I'I'DHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . 60 Percent Distribution of Births in the last 5 Years by Type of Prenatal Care for the Mother and Percentage of Births Whose Mother Received a Tetanus Toxoid Injection, According to Selected Background Characteristics of Mother, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Percent Distribution of Bi~hs in the Last 5 Years by Type of Assistance During Delivery, According to Selected Background Characteristics of Mother, TTDHS 1987 . . . . . . . . . . . . . . . 62 Percent Distribution of Births in the Last 5 Years by Place of Delivery, According to Selected Background Characteristics of Mother. TFDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . 63 vi Table 6.8 Table 6.9 Table 6.10 Table 6.11 Table 6.12 Table 6.13A Table 6.13B Table 6.14A Table 6.14B Page Among All Children Under 5 Years of Age, the Percentage With Health Cards, the Percentage Who Are Recorded as Immunized on Health Card; Among Children with Health Cards, the Percentage of Whom Yellow Fever, DPT, Polio and Measles Immunizations are Recorded on the Health Card, According to Selected Background Characteristics, "ITDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 Among All Children Under 5 Years of Age Without Health Cards, the Percentage Who Are Reported by the Mother as Having Been Immunized Against Yellow Fever, DPT, Polio, and Measles, According to Selected Background Characteristics, TTDHS, 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 Among Children Under 5 Years of Age, the Percentage Reported by the Mother to Have Had Diarrhoea in the Past 24 Hours and the Past Two Weeks, According to Selected Background Characteristics, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . 68 Among Children Under 5 Years of Age Who Had Diarrhoea in the Past Two Weeks, the Percentage Consulting a Medical Facility, and the Percentage Receiving Different Treatments as Reported by the Mother, According to Selected Background Characteristics, ITDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . 69 Among Mothers of Children Under 5 Years of Age, the Percentage Who Know About ORT by Education, According to Selected Background Characteristics, TTDHS 1987 . . . . . . . . . . . . . . . 70 Percent Distribution of Children Aged 3-36 Months by Standard Deviation Category of Height-For-Age Using the NCHS/CDC/WHO International Reference Population, According to Selected Background Characteristics, "ITDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Percent Distribution of Children Aged 3-36 Months by Percent of Median Height-For-Age Using the NCHS/CDC/WHO International Reference Population, According to Selected Background Characteristics, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Percent Distribution of Children Aged 3-36 Months by Standard Deviation Category of Weight-For-Height Using the NCHS/CDC/WHO International Reference Population, According to Selected Background Characteristics, TFDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 Percent Distribution of Children Aged 3-36 Months by Percent of Median Weight-For-Height Using the NCHS/CDC/WHO Intematiunal Reference Population, According to Selected Background Characteristics, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 vii Table 6.15A Table 6.15B Table 6.16A Table 6.16B Page Percent Distribution of Children Aged 3-36 Months by Standard Deviation Category of Weight-For-Age Using the NCHS/CDC/WHO International Reference Population, According to Selected Background Characteristics, TFDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 Percent Distribution of Children Aged 3-36 Months by Percent of Median Weight-For-Age Using the NCHS/CDC/WHO International Reference Population, According to Selected Background Characteristics, TrDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Percent Distribution of Children Aged 3-36 Months, the Percent in Each Height-for-Age Standard Deviation Category by Each Weight-for-Height Standard Deviation Category (Waterlow Classification) Using the NCHS/CDC/WHO International Reference Population, TTDHS 1987 . . . . . . . . . . . . . . . . . . 78 Percent Distribution of Children Aged 3-36 Months, the Percent in Each Height-for-Age Percent of Median Category by Each Weight-for-Height Percent of Median Category (Waterlow Classification) Using the NCHS/CDC/WHO International Reference Population, TrDHS 1987 . . . . . . . . . . . . . . . . . . 78 APPENDIX A Table A.1 APPENDIX B Table B.1 Table B.2 Summary of Results of Household and Individual Interviews, by Residence, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 List of Selected Variables with Sampling Errors, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Sampling Errors: Total TrDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . 91 Sampling Errors: Women Aged 15-24, IrDHS 1987 . . . . . . . . . . . . . . . 92 Sampling Errors: Women Aged 25-34, TTDHS 1987 . . . . . . . . . . . . . . . 93 Sampling Errors: Women Aged 35-49, TTDHS 1987 . . . . . . . . . . . . . . . 94 Sampling Errors: Urban, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . 95 Sampling Errors: Rural, TTDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . 96 viii LIST OF FIGURES Figure 1.1 Figure 1.2 Figure 2.1 Figure 2.2 Figure 3.1 Figure 3.2 Figure 3.3 Figure 4.1 Figure 4.2 Figure 4.3 Figure 4.4 Figure 4.5 Figure 5.1 Figure 5.2 Figure 6.1 Figure 6.2 Figure 6.3 Figure 6.4 Figure 6.5 Page Crude Birth and Death Rates 1901-1984 . . . . . . . . . . . . . . . . . . . . . . . . . 2 Percent Distribution of Women in the Survey by Current Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Union Status by Current Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Duration of Breastfeeding, Amenorrhoea and Postpartum Abstinence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Total Fertility Rate, Various Years and Data Sources . . . . . . . . . . . . . . . 18 Total Fertility Rate 0-4 Years Before the Survey, and Children Ever Born to Women 40-49 Years . . . . . . . . . . . . . . . . . . . . . . 19 Age-Specific Fertility Rates, WFS and TTDHS . . . . . . . . . . . . . . . . . . . 20 Current Use of Family Planning by Method, Women in Union 15-49 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Family Planning Knowledge and Use, Women in Union 15- 9 , . . . . , . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . 32 Current Use by Age, TTDHS and WFS Women in Union 15-49 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Current Use of Family Planning by Education and Number of Living Children, Women in Union 15-49 . . . . . . . . . . . . . . . . . . . . . 35 Source of Family Planning Supply Current Users . . . . . . . . . . . . . . . . . . 39 Fertility Preferences, Women in Union 15-49 . . . . . . . . . . . . . . . . . . . . . 48 Fertility Preferences by Parity, Women in Union 15-49 . . . . . . . . . . . . . . 50 Infant Mortality Rates, TTDHS, WFS, Vital Statistics . . . . . . . . . . . . . . . 58 Trends in Infant and Child Mortality . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Immunization Coverage, Children 1-5 . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Age Distribution of Weighed and Measured Children, and All Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 Nutritional Status of Children Aged 3-36 Months . . . . . . . . . . . . . . . . . . 71 ix X PREFACE The Trinidad and Tobago Demographic and Health Survey (TTDHS) was conducted as part of the worldwide Demographic and Health Survey Program (DHS) in which more than thirty countries participated. Trinidad and Tobago was the thirteenth country, and the only participant from the Caribbean Region. The demographic and health characteristics of women in their reproductive years, and those of their young children were studied in order to obtain a better understanding of factors related to population growth and the health of children. The Association feels extremely proud for having had the privilege of undertaking a survey of such importance and magnitude in Trinidad and Tobago for two main reasons. Firs.fly, be.canse it gave us an oppo.rt3mity to update the health data base of the country for the first ume smee the 1977 World Fertility Survey. Secondly, this project has enhanced the capability of the Association to nndertake demographic and health surveys through the transfer of skills in the areas of project management, research methodology and computer literacy. The survey findings will be useful in providing baseline and evaluative information for policy makers and administrators of health and family planning programmes in the country. Any project as complex as this one takes a number of dedicated and professional people to ensure that it is successfully implemented and completed. Many organizations and individuals were involved in the design and execution of this survey, to whom we would like to express our sincerest appreciation. Firstly, we are extremely grateful to the staff of the Institute for Resource Development for their technical assistance. We wish to pay special tribute to Ms. Amy Sheen, Country Monitor for the TTDHS. She was endlessly resourceful and her advice was always practical and wise. We also wish to recognize the invaluable contributions made by Ms. Anne Cross for her overall consultancy, Mr. Alfredo Aliaga for his professionalism in designing the sample, Mrs. Jeanne Cushing, Mr. John Heinrich, Mr. J. Guiilermo Rojas and Mr. Brian Taaffe for providing training and coordinating the data processing activities, Mr. Roger Pearson for providing the anthropometric training and to Ms. Kaye MitcheU and Mr. Robert Wolf for report production support. We extend our thanks to USAID who made it possible to conduct this survey by providing the necessary funding. The Association feels indebted to its Survey Director, Mr. Kenneth Heath, who successfully managed the technical aspects of the survey. The Association was also privileged to have its own Research, Evaluation and Training Officer, Mrs. Dona Da Costa-Martinez, to efficiently coordinate the administrative activities of this project, from its inception to the writing of the final report. We also wish to extend sincerest thanks to the following agencies/institutions which participated very actively in planning the survey and reviewing the questionnaire: The Central Statistical Office (CSO) The Ministry of Health The Government's Population Programme The Institute for Social and Economic Research (ISER) Pan American Health Organization/WHO The United Nations Economic Commission for Latin America and the Caribbean (UNECLAC) UNICEF Caribbean Food & Nutrition Institute (CFND We most profoundly appreciate the very constructive reviews done by Mr. Jack Harewond, Demographer, and Dr. Sunney Alexis of the Food Nutrition Laboratory, on chapters 4, 5 and 6 of the Final Report. Their help in improving the quality of the Final Report cannot go unnoticed. xi We especially wish to acknowledge the significant contributions of the temporary staff of TIDHS. Without their support the process would have been more rocky and indeed less challenging. We owe a significant debt of gratitude to the Fieldwork Coordinator, the Supervisors, Interviewers and the respondents. It goes without saying that without them there would have been no survey. But even more, many of the potential users of the findings provided encouragement and suggestions. We feel no less gratitude to all the other individuals and organisations whose in-kind contributions added to the success of the project. To the pretest and fieldwork gaining resource personnel we say a special thank you. We thank also the Ministry of Energy, Labour, Employment and Manpower Resources for providing us with accommodation to present the Preliminary Findings; the Child Welfare League for making their facilities and children available for anthropometric training; The Joint Services Staff College and All Saints Parish Hall for providing us with the accommodation from which to conduct training. Finally, we extend sincere thanks to all the District Health Visitors of the various counties and to all those who contributed one way or another to the success of the Trinidad and Tobago Demographic and Health Survey. Emile P. Elias President, FPATI" xii xiii REPUBLIC OF TRINIDAD AND TOBAGO County Ward (Trinidad) 1 Parish (Tobago) • Municipalities N 2. St. Andrew 3. St. George 4. St. David 5. St, Mary 6. St. Paul 7. St. John nidad Scale in Miles 5 0 5 10 x Jr CHAPTER 1 BACKGROUND 1.1 History, Geography, Economy The Republic of Trinidad and Tobago consists of twin islands in the southem part of the Caribbean Sea. With an area of 4,828 square km. (1,864 square miles), Trinidad lies seven miles north of the Venezuelan coast and is separated from it by the Gulf of Paria. Tobago, with an area of 300 square kin. (116 square miles) is situated 19 miles northeast of Trinidad. The islands' climate is pleasant throughout the year, varying between 20 degrees and 33 degrees Celsius. Trinidad was discovered by Christopher Columbus in 1498, and occupied by Spain for 300 years. Tobago was acquired by Britain in 1763 by the Treaty of Paris following more than 260 years of Dutch and French control. In 1797 Britain seized Trinidad from Spain and the two islands were unified for administrative convenience in 1889, and became a joint colony in 1899. Because of its history as a plantation economy, Trinidad and Tobago has attracted waves of migrants who came as colonists and slaves. Today, the Islands arc inhabited by persons of Portuguese, Chinese, Syrian, Lebanese, African, and East Indian descent. Africans and East Indians predominate, each comprising about 41 percent of the total population. The importance placed on education over the past thirty years has resulted in a population almost universally literate. Census data show that the proportion of the population 15 years of age and over who reported no education declined from 11 percent in 1960 to 5 percent in 1980. The proportion attaining secondary level rose from 14 percent in 1960 to 32 percent in 1980, while the proportion attaining university education increased from less than 1 percent to 2 percent (Central Statistical Office 1987c). One-third of the country's population is Roman Catholic. One-fourth is Hindu and 15 percent are Anglican. The remainder includes Muslims and other Christian denominations. The economy is dominated by petroleum which constitutes nearly one-fourth of the GDP, 28 percent of government revenue, and 71 percent of exports. The collapse of oil prices this decade has caused the economy to deteriorate. 1.2 Population The population of Trinidad and Tobago was estimated by the Central Statistical Office to be 1.2 million in mid-1986, and is projected to reach approximately 1.6 willlon by the end of the century. The intercensal growth rate has moved in cycles, climbing from 1.8 percent per annum in 1851 to 3.0 percent in 1881 before dropping to 0.9 percent in 1921 (see Table 1.1). During the so-called baby boom years, the growth rat* rose again, to 2.8 percent in 1960 before falling to 1.2 percent in 1970. The rate of growth betw~n the two most recent censuses, 1970 and 1980, was 1.5 lX~rcent, suggesting an upturn once again. Fluctuations in the growth rate arc due to changes in three components--the crude birth rate, the crude death rate (see Figure 1.1) and net migration. The birth rate fell five points, from 35 births per 1,000 at the turn of the century to 30 in 1931. During the baby boom years, the birth rate increased 12 points to reach 42 per thousand in 1954, and then dropped 15 points in the next two decades. The stagnation in the birth rate since the mid 1970s m due to the countervailing influences of a large number of women (born during the baby boom years) entering their peak reproductive years, and a decline in the fertility rates of women at all ages. Table 1.1 Population of Trinidad and Tobago at Census Years Average Annual Census Rate of Year Male Female Total Growth {%) 1844 . . . . . . . . . . 73,023 1851 . . . . . . . . . . 82,978 1.8 1861 . . 99,848 1.9 1871 68,667 58,025 126,692 2.4 1881 92,410 78,769 171,179 3.0 1891 117,060 101,321 218,381 2.4 1901 144,491 129,408 273,899 2.3 1911 174,349 159,203 333,552 2.0 1921 186,802 179,111 365,913 0.9 1931 206,619 206,164 412,783 1.2 1946 282,299 280,923 563,222 2.1 1960 411,580 416,377 827,957 2.8 1970' 459,512 471,559 931,071 1.2 1980 539,640 540,151 1,079,791 1.5 Source: Central Statistical Office 1987a. * Non-instltutional population only. F igure 1.1 Crude Birth and Death Rates 1901-1984 Rate 50- 45- 40- 35 ~ 30- 25- 20- 15- 10- 5- 0 E 1901 6 i i i i i i i ~ i i ~ r i I i i 11 16 21 26 31 36 41 46 51 54 56 61 66 71 76 81 84 Year • 8ouroe : Popu la t ion & V i ta l S ta t i s t i cs Reporta. C80, Rap. o! Trinidad & Tobago Birth Rate Death Rate The crude death rate--deaths per 1,000 population--has declined markedly, paralleling a trend in other developing countries. Between 1901 and 1931, the rate fluctuated between 20 and 25 deaths per 1,000, and then fell to 8 in 1961. In 1984, the crude death rate stood at 7 per 1,000. Migration contributed significantly to the growth of the islands' population until the turn of the century. Since 1970, however, high net out-migration has reduced by one-half the impact of the excess of births over deaths (Central Statistical Office 1987c). 1.3 Population and Family Planning Policies and Programmes The provision of family planning services in Trinidad dates back to 1956 when a group of concerned citizens opened a family planning clinic in Point Fortin. A second clinic was opened in Port of Spain in 1959, heralding the genesis of the Family Planning Association of Trinidad and Tobago (FPATr). The Association became the thirty-second member of the International Planned Parenthood Federation (IPPF) in 1961, and started a clinic in the second major town in Trinidad, San Femando, one year later. In 1967, the government began providing maternal and child health services in health centres throughout the country. Currently, the government offers family planning at 95 health centres, the FPATr operates two facilities, and the Archdiocesan Family Life Commission (AFLC), established in 1968, provides instructions on natural family planning at 10 facilities. Contraceptive information and supplies are thus easily available on both islands of the Republic. In 1967, the govemment convened a population council to coordinate family planning activities throughout the country. The council included representatives of the government's family planning programme, the FPATr, and the Catholic Marriage Advisory Council (since renamed the AFLC). A main objective of the council was to reduce fertility to less than 19 births per 1,000 population by 1983. Has the availability of family planning made a difference? Fertility studies conducted in 1970 and 1977 suggest that the use of contraceptives by women in union increased from 44 percent to 52 percent during this interval (Harewood 1978; Sathar and Chidambaram 1984). On the other hand, the crude birth rate, measured by vital statistics data, did not decline during th'rs period, due in part to the increased number of women entering the peak fertile years. (See Chapter 3 for a more extensive discussion of the effect of contraception on fertility in Trinidad and Tobago). The continued high growth rate, coupled with deteriorating economic conditions, prompted the FPATT to carry out a survey of the factors affecting population growth. In 1987, with assistance from the Institute for Resource Development/Westinghouse (IRD), FPATr collected population and health data that will be useful for making informed policy choices. 1.4 Health Priorities and Programmes The goal of the Ministry of Health (MOH) is to protect, promote, and maintain the mental, social, physical health and well-being of the people of Trinidad and Tobago, and thereby improve the quality of life of the citizens. Primary health care is the main strategy pursued to achieve the goal of health for all. Basic health services are provided through 102 health centres, two general hospitals, three county hospitals, six district hospitals, and extended care units which provide outpatient care as well. Within the framework of primary health care, health persounel rely on interdisciplinary, intersectoral collaboration to achieve an integrated health care delivery service. The MOH utilizes epidemiological surveillance, health education, and environmental monitoring as tools to implement specific preventative health programmes. Maternal and child health care services have been a major focus of activities. One objective of this programme is the promotion and protection of the health of mothers and pre-school aged children, which is achieved through prenatal clinics, postnatal clinics, and child health clinics. At the prenatal clinics, routine checks are conducted, and health education about topics such as family life and family planning is provided. Most postnatal clinics offer family planning services and vaccinations against rubella for unprotected mothers. Child health clinics emphasize the promotion and maintenance of health, prevention of communicable diseases, and early detection of abnormalities. In Trinidad and Tobago, programmes to prevent communicable diseases through immunization have been implemented since the middle of this century. In 1973, the programme was first introduced nationwide, and has continued to be a priority within the Maternal and Child Health Services Division of the MOH. The nation is committed to the World Health Organization's Expanded Programme on Immunization (EPI). The EPI programme strives to ensure that the target population of children under age one, pregnant mothers, and puberty-aged girls are appropriately immunized. Diseases included in the international EPI are diphtheria, tetanus, whooping cough, poliomyelitis, measles, and tuberculosis. In Trinidad and Tobago, yellow fever and rubella are included in the immunization schedule, while the BCG injection against tuberculosis is not routinely given. Gains made in recent years reflect the MOH effort in the area of immunization. There have been no reported cases of poliomyelitis since the 1971-72 epidemic, and the incidence of diphtheria and neonatal tetanus has declined. However, the number of reported cases of measles remains high (2,660 in 1986), and vaccination coverage is low (Central Statistical Office 1987a). In addition, risk of a polio epidemic exists, since immunization coverage is not universal. TTDHS findings regarding health appear in Chapter 6 of this Report. Other priority areas of the MOH include school health and adolescent development, drug abuse, chronic diseases, and Acquired Immune Deficiency Syndrome; these topics were not covered in the survey. 1.5 Objectives of the Survey The short term objective of the Trinidad and Tobagn Demographic and Health Survey (TrDHS) is to collect and analyse data on the demographic characteristics of women in the reproductive years, and the health status of their young children. Policymakers and programme managers in public and private agencies will be able to utilize the data in designing and administering programmes. The long term objective of the project is to enhance the ability of organisations involved in the TTDHS to undertake surveys of excellent technical quality. 1.6 Organisation of the Survey The Trinidad and Tobago DHS survey--a national-level self-weighting random sample survey--was funded by the United States Agency for International Development (US/kID) and executed by the Family Planning Association of Trinidad and Tobago (FPATr). Technical assis- tance was provided by the Demographic and Health Surveys Program at the Institute for Resource Development (IRD), a subsidiary of Westinghouse located in Columbia, Maryland. The timetable for survey activities is as follows: November 1986 Contract Signed February 1987 Pretest Aptil-May 1987 Training May-September 1987 Fieldwork January 1988 Preliminary Report November 1988 Final Report December 1988 National Seminar 4 The sampling frame for the 'ITDHS was the Continuous Sample Survey of Population (CSSP), an ongoing survey conducted by the Central Statistical Office based on the 1980 Population and Housing Census. (For details on sample design, see Appendix A.) The TTDHS used a household schedule to collect information on residents of selected households, and to identify women eligible for the individual questionnaire. The individual ~uestionnaire was based on DHS's Model "A" Questionnaire for High Contraceptive Prevalence ountries, which was modified for use in Trinidad and Tobago. It covered four main areas: (1) background information on the respondent, her partner and marital status, (2) fertility and fertility preferences, (3) contraception, and (4) the health of children. A steering committee was established to provide guidance throughout project development and implementation. While the FPATT was responsible for overall coordination, staff recruitment and training, accountability of funds, and publicity, other committee members provided valuable assistance. During all phases of the survey, the Central Statistical Office shared experience gained during the 1977 Trinidad and Tobago World Fertility Survey. In addition to providing the Survey Director, CSO assisted with survey planning and development; offered consukation on survey methodology, operations, and data processing; prepared the sampling information; assisted with training and supervision of staff; and provided office space and printing services. The Ministry of Health (MOH) assisted in questionnaire design, provided transportation for field workers, and assisted in collection of data on the availability of health services. The Institute of Social and Economic Research (ISER) offered general guidance on survey methodology, training, and field operations, and is conducting further analysis of the TI'DHS data. Representatives of the Ministry of Finance and the Economy; the Ministry of Community Development, Welfare and the Status of Women; and the United Nations Economic Commission for Latin America and the Caribbean (UN/ECLAC) provided general assistance on survey methodology. The Caribbean Food and Nutrition Institute helped design questions regarding breastfeeding and nutrition. Office staff for the j~roject included the Survey Director, who was responsible for technical aspects of the survey, the ~urvey Co-ordinator (FPATI" Research and Training Officer) who was responsible for administrative and financial aspects of the survey, three Data Entry Clerks, the Chief Editor, the Control Clerk/typist, and the Messenger. Thirty-three field personnel were employed, including the Fieldwork Co-ordinator, five Supervisors, four Field Editors and twenty- three Interviewers. (For details of survey implementation activities, see Appendix A.) 1.7 Background Characteristics of Survey Respondents The TTDHS consisted of a sample of 4,799 households, 4,122 of which were successfully interviewed. These households included 4,196 women eligible to be interviewed, from which 3,806 completed questionnaires were obtained. The response rate was 94 percent at the household level, and 92 percent at the individual level, giving an overall response rate of 87 percent. (Details for response rate calculations appear in Appendix A.) The distribution of the sample population by age, residence, education, and ethnicity is shown in Table 1.2, along with corresponding figures from the 1980 Census. The 15-19 age group comprises 18 percent of the sample, compared with 20 percent for each of the next two age groups. The proportion of the sample in the older age groups declines steadily, from 14 percent for the 30-34 group to 7 percent for the 45-49 age group. In the youngest age group, the survey population contains fewer women than the Census population. The drop in the size of the 15-19 cohort in the seven years between the Census and the Survey reflects the sharp decline in the birth rate which occurred between 1960 and 1970. Table 1,2 Distribution of Women 15-49 by Age, Residence, Education, and Ethnicity, 1980 Census and TTDHS 1987 Background 1980 1987 Characteristic Census TTDHS Age 15-19 24.2 17.9 20-24 20.3 19.6 25-29 16.1 19.6 30-34 12.9 14.3 35-39 10.3 11.6 40-44 8.7 9.7 45-49 7.4 7.3 Residence Urban 48.7 44.4 Rural 51.3 55.6 Education <Complste primary 15.6 8.2 Completed primary 40.7 38.0 Secondary+ 43.7 53.9 Ethniclty African 39.6 35.3 Indian 42.9 47.0 Mixed 15.4 17.1 other 2.1 0.7 lO0.O Total 100,0 Central Statistical office, Population and Sousing Census 1980, Vol. 2, 1983 Forty-four percent of women live in urban areas ! and 56 percent in rural areas; this distribution is slightly less urbanized than the 1980 Census population. The sample population is better educated than the comparable population from the Census, reflecting ongoing gains in education. Fifty-four percent of women in the sample have some secondary schooling, a 10-point advance since the Census. The ethnic composition of the sample, as reported by respondents, differs from Census figures. Africans make up a smaller part of the sample than the Census population, 35 versus 40 percent, respectively. East Indians comprise 47 percent of the survey compared to 43 percent in the Census. Seventeen percent of survey respondents are of mixed race; fewer than 1 percent belong to other ethnic groups. The reasons for the difference in the ethnic composition of the sample have not been ascertained, but could be due to unintended oversampling in areas where the East Indian population is heavily concentrated, a higher response rate among this group, or a larger household size. Table 1.3 gives an overview of the sample population according to level of education. Overall 92 percent have completed at least five years of primary education, while 54 percent have at least some secondary education. ~ Urban includes Port of Spain, St. George county, and the boroughs of San Femando, Arima and Pt. Fortin. In this report, respondents with some secondary education were classified into two groups. "Secondary I" includes women with some or full secondary education, but fewer than five "0" Level exams passed. "Secondary II" includes women with some or full secondary education, with five "0" level exams passed, at least one "A" level, or some University education. It was assumed that exam results were a better indicator of academic achievement than years of education, the more customary measure. 6 Table 1.3 Percent Distribution of Women by Education, According to Selected Background Characteristics, TTDHS 1987 Eduoation Background Characaoter- <Complete Completed Secondary Secondary istic Primary P r~ary I ~ II* Total Wumber Age 15-19 0.6 12.7 76.0 10.7 100 683 20-24 4.0 26.2 55.6 14.2 I00 745 25-29 5.2 40.9 40.7 13.2 100 745 30-34 5.3 55.8 26.3 12.5 100 543 35-39 11.8 51.5 27.0 9.6 I00 441 40-44 19.5 51.6 20.3 8.6 100 370 45-49 30.6 49.1 12.9 7.2 100 279 Residence Urban 5.7 32.0 47.0 15.3 100 1,690 Rural 10.2 42.6 38.5 8.6 i00 2,116 Ethnicity African 3.8 39.1 45.8 11.3 i00 1,342 Indian 12.4 40.2 37.5 9.8 i00 1,787 Mixed 6.0 31.0 48.7 14.3 i00 649 Other 0.0 0.0 25.9 74.1 i00 28 Total 8.2 38.0 42.3 11.6 100 3,806 x Some or full secondary education, but fewer than five "0" level exams passed. * Some or full secondary education, with five "O" level ex~s passed, at least one "A" level, or some university education. Educational differences in the age groups in Table 1.3 reflect improvements in recent years. Among women 45-49 who attended schoolthree decades ago, 31 percent did not complete primary school, and only 20 percent received secondary education. By contrast, fewer than 1 percent of women 15-19 failed to complete primary school while 87 percent attained secondary education. Urban respondents are better educated than rural dwellers; 62 percent of the former have at least some secondary schooling, versus 47 percent of those in rural areas. The table also shows the differences in education among the various ethnic groups. East Indian women are the least educated. Twelve percent did not complete primary education, compared with 4 percent of Africans and 6 percent of those of mixed race. The "other" ethnic category (mostly White and Chinese) is the best educated. All have at least some secondary school, and 74 percent have full certifxcation. (This figure should be regarded with caution since the category comprises only 28 women). Since respondents who comprise the "other" ethnic category are so few in number, this category is not shown in subsequent tables where ethnicity is a background characteristic. One important aspect of data quality is the single-year age distribution of the sample, shown in Figure 1.2. The year-to-year fluctuations may be due to the high out-migration in reeunt years. There is little evidence of heaping on ages ending in digit "5" or "0", suggesting that the data are free of gross age estimation bias. (Further analysis of age data exceeds the scope of this report, but is required to make a more definitive assessment of the quality of age data.) 7 6% Figure 1.2 Percent Distribution of Women in the Survey by Current Age 5% 4% 3% 2% 1% 0% ' 15 i i i i i i~ l i l l i i i l l L i i l l i i i i J l l l l l i i 20 25 30 35 40 45 50 Current Age Trinidad & Tobago DHS 1987 8 CHAPTER 2 NUPTIALITY AND EXPOSURE TO RISK OF PREGNANCY In Trinidad and Tobago, as in most Caribbean and Latin American countries, sexual unions occur not only in the context of legal marriage, but also in common-law and visiting unions as well. In this report, formal marriage refers to those persons legally married and living together in the same househnld; common-law refers to those not legally married but living together; and visiting refers to those in a regular sexual relationship but not living together. Unless otherwise specified, "women in union" includes those in all three types of arrangements. Table 2.1 Percent Distribution of Women by current Union Status, According to Background Characteristics, TTDHS 1987 Dnion Status Never widowed/ Background in Mar- Common- Visl- Divorced/ Characteristic Union fled Law tlng Separated Total Number Age 15-19 75.4 5.4 3.2 11.7 4.2 I00 683 20-24 31.8 29.0 12.1 19.3 7.8 I00 745 25-29 9.9 49.3 18.3 16.2 6.3 I00 745 30-34 4.4 54.0 18.0 14.4 9.2 100 543 35-39 3.4 59.9 15.6 12.7 8.4 i00 441 40-44 3.0 57.6 15.4 11.4 12.7 i00 370 45-49 1.4 59.9 15.1 9.0 14.7 i00 279 Residence Urban 20.2 35.7 15.1 18.5 10.4 100 l, 690 Rural 25.4 45.1 12.2 11.0 6.3 100 2,116 Education <complete primary 7.4 49.4 25.3 6.1 11.9 100 312 Completed primary 10.2 50.6 17.6 12.9 8.7 i00 l, 445 SecondaryI* 34.7 31.4 i0.0 16.8 7.0 100 1,609 SecondaryI I* 34.3 37.7 4.5 15.9 7.5 i00 440 Ethnlcity ) African 18.6 27.2 17.4 25.6 11.2 I00 i, 342 Indian 26.5 53.9 0.8 5.0 5.8 I00 1,787 Mixed 23.0 33.3 18.8 16.6 9.3 I00 649 Total 23.1 40.9 13.5 14.3 8.1 100 3,806 * Some or full secondary education, but fewer than five "0" level exams passed. * Some or full secondary education, with five "0" level exams passed, at least one "A" level, or some university education. ) Excludes 27 women of "other" ethnlclty, and one respondent with missing information. 2.1 Cur rent Mar i ta l Status Table 2.1 and Figure 2.1 present the distribution of women in the sample according to their current union status. Overall, 23 percent of the respondents have never been in union, while 41 percent am formally married, 28 percent are living together or visiting, and 8 percent am either separated, widowed or divorced. Nearly all women in Trinidad and Tobago enter some type of union during their reproductive years, since the percentage reporting themselves as "never in a union" drops from 75 percent of women 15-19 to only 1 percent of women 45-49. While the 9 proportion currently in some type of union is quite high for women aged 25-49 (ranging fi'om 84 to 88 percent), younger women are more likely to report being in the less stable visiting unions, while older women tend to be in formal marriages. The relationship between education and union status is striking. More than 34 percent of women with secondary education have never been in unions, compared to 10 percent of women who have completed primary school. (Recall that the better educated women are younger than average.) Table 2.1 also indicates that union status differs considerably by ethnic origin. While the proportion currently in union is nearly alike among the three ethnic groups, East Indian women tend to be formally married (54 percent) rather than living together or visiting (14 percent), while the reverse is true among African women (27 percent in formal unions and 43 percent in less formal arrangements). Figure 2.1 Union Status by Current Age 1 oo 75~ 50~ 25~ O~ 15-19 20-24 25-29 30-34 35-39 Current Age m 40-44 45-49 I. I~ Never In union ~ Formally married F~ Common law | I Vlaltl~g ~ W|dowed/dlv./aap. Trinidad & Tobago DHS 1987 2.2 Median Age at First Union Table 2.2 shows that the median age at first union is just about 20, and has not changed in the last two decades. Only for women 45-49 is the median age lower, by one full year. Entry into a union before age 15 is relatively uncommon and has been so for more than two decades. Eleven percent of women in their forties entered a union before age 15; about one-half as many women 15-39 did so. Thirty percent or more of each cohort of women 20-49 were in a union before age 18. 10 Table 2.2 Percent Distr ibut ion of Women by Age at First Union and Median Age at F irst Union, According to Background Characterist ics, TTDE$ 1987 Age at F irst Union Never Background in Character ist ic Union <15 15-17 18-19 20-21 22-24 25+ Total Number Median 2 Age 15-19 75.4 6.6 14.3 3.7 . . . . . . 100 683 -- 20-24 31.6 6.0 28.3 19.1 11.8 3.0 -- 100 745 19.7 25-29 9.9 5.1 24.6 23.2 16.9 14.5 5.8 i00 745 19.8 30-34 4.4 5.5 26.2 21.0 16.4 15.7 10.9 100 543 19.7 35-39 3.4 5.2 26.1 20.6 16.6 13.6 12.2 i00 441 19.7 40-44 3.0 10.5 21.1 20.5 15.9 13.8 15.1 1O0 370 19.8 45-49 1.4 11.1 30.5 20.8 14.7 11.5 10.0 190 279 18.8 Residence Urban 20.2 7.0 22.9 18.9 13.7 10.4 6.9 I00 1,690 20.1 Rural 25.4 6.3 25.2 17.0 11.6 8.6 5.8 100 2,116 20.2 Educat ion <Complete pr imary 7.4 16.0 35.9 19.9 9.6 5.4 5.8 100 312 17.9 Completed pr imary 10.2 8.4 31.8 21.0 12.7 i0.I 6.5 100 1,445 19.0 Secondaryl 2 34.7 4.6 20.2 15.5 12.4 7.7 5.0 i00 1,609 21.6 SecondaryII* 34.3 I.I 8.2 14.8 14.5 16.1 10.9 I00 440 24.1 Ethnicity 4 Afr ican 18.6 9.2 26.2 18.7 12.2 8.9 6.1 i00 1,342 19.6 Indian 26.5 4.8 22.1 17.2 13.2 9.7 6.4 100 1,787 20.8 Mixed 23.0 6.3 26.3 17.3 11.2 9.6 6.3 i00 649 20.0 Total 23.1 6.6 24.2 17.8 12.5 9.4 6.3 i00 3,806 -- -- Omitted due to censoring. Def ined as the age by which one-half of women have ever married. l Some or full secondary education, but fewer than five "O" level exams passed. ) Some or full secondary education, with five "o" level exams passed, at least one "A" level, or some univers i ty education. Excludes 27 wom~n of "other" ethnicity, and one respondent with miss ing information. As expected, better educated women marry later, on average, than less educated women. The median age at marriage for those with full secondary education is 24 years, versus 18 years for women with less than completed primary education. On average, African women entered their first union at age 20, one year younger than East Indian women. Table 2.3 compares the median age at first union for women 20-49 according to various background characteristics. The median age at first union for women aged 20-49 is 19.6. Women with less than secondary education and African women enter unions slightly earlier than women with more education or mixed and Indian women. While the median age for women with less than a complete primary education is only 17.9 years, this figure rises to 22.8 for those with at least five "O" level passes. (The median ages at first union for various ethnic groups are different in Tables 2.2 and 2.3 because the former refers to all women while the latter is restricted to women aged 20-49.) 2 .3 Breast feed ing and Postpar tum Insuscept ib i l i ty Aside from the age at which women enter into unions, several other factors which affect fertility and birth intervals are measured in the JqDHS and presented in Tables 2.4, 2.5 and 2.6. Susceptibility to pregnancy after a birth can be delayed by breasffeeding, which inhibits the resumption of ovulation and menstruation, and by practicing postpartum sexual abstinence. II Table 2.3 Median Age at First Union among Women Age 20-49 Years, by Current Age and Selected Background characteristics, TTDHS 1987 Current Age Background Ages characteristic 20-24 25-29 30-34 35-39 40-44 45-49 20-49 Residence Urban 19.2 19.8 19.7 20.3 20.3 19.0 19.7 Rural 20.1 19.7 19.7 19.3 19.2 18.5 19.6 Education <complete primary 19.0 18.9 17.1 17.6 17.7 17.7 17.9 Completed primary 18.2 19.1 18.9 19.2 19.3 18.7 19.0 SecondaryI ~ 19.7 20.3 20.4 20.8 21.2 21.0 20.2 SecondaryII t * 22.2 22.5 21.4 24.8 23.0 22.8 Eth~Icity* African 18.5 19.5 19.2 19.4 20.1 19.1 19.2 Indian 21.4 20.2 20.3 19.7 19.3 18.4 20.1 MiNed 18.7 19.4 19.8 20.1 20.5 18.4 19.5 Total 19.7 19.8 19.7 19.7 19.8 18.8 19.6 * Fewer than 25 cases. Some or full secondary education, but fewer than five "o" level eNams passed. • Some or full secondary education, with five "0" level exams passed, at least one "A" level, or some university education. Excludes 25 women of "other" ethnlclty or with missing information. Respondents who gave bLrth in the five years preceding the survey interview were asked ff they breasffed, and the duration of breasffeeding. Table 2.4 shows that breasffeeding is common, though not univexsal. Overall, 89 percent of births in the five years prior to the survey were breasffed. The practice was slightly more common for babies born to women in rural areas, to African women, and to women with the highest level of education, although differences in each case are slight. In addition, women were asked how many months they were amenorrhoeic after each delivery, and how long they abstained from intercourse. Also, women were asked ff they were currently breasffeeding, amenorrhoeic, and/or practicing abstinence. Since it may be difficult for respondents to recall the timing of these events, and since it may be difficult to precisely define when weaning takes place, data in Tables 2.5 and 2.6 are current status estimates which refer to whether or not the woman was breasffeeding and/or amenorrhoeic at the time of the survey interview, rather than her reported durations for these events. In Table 2.5, all births three years before the survey are considered, although twins are counted as a single birth. Durations of breasffeeding are quite short. Table 2.5 shows that while 84 percent of women with births 2-3 months ago were still breasffeeding, fewer than one-half of those who delivered 6-7 months ago continued the practice. In other words, most women who breasffeed at all continue the practice fur at least three months, but many stop shortly thereafter. Menstruation returned very shortly after birth for most women. Only 46 percent of women 2-3 months postpartum were amenorrhoeic; this figure dropped to 19 percent for women 4--5 months postpartum. Sexual abstinence, too, is practiced for only a short time following delivery. More than three-fourths of women resumed having intercourse 2 to 3 months after delivery, and only 3 percent continued to abstain after 7 months. The fourth column in Table 2.5 shows the proportion of women protected from pregnancy due to either amenorrhoea or abstinence. While 98 percent of women who delivered less than 2 months ago are insusceptible to pregnancy, only one-third are 12 still protected 4 to 5 months after a birth. Thus, most woman who want to space a birth will need to take steps to prevent pregnancy shortly after delivery. Table 2.4 Percentage of All Births in the Last 5 Years Who Have Ever Been Breastfed, According to Selected Background Characteristics of the Mother, TTDHS 1987 Background Ever Characteristic Breastfed Number Age 15-19 96.9 91 20-24 89.3 516 25-29 87.7 660 30-34 89.9 388 35-39 99.3 196 40-44 83.3 66 45-49 * 12 Residence Urban 87.7 826 Rural 89.4 i, 103 Education <Complete primary 88. I 134 Completed primary 88.9 855 Secondary I a 87.5 782 Secondary II 2 93.7 158 Ethniclty ) African 91.4 748 Indian 85.7 831 Mixed 89.6 338 Total 88.6 1,929 Note: Includes births 1-59 months before the survey. * Fewer than 25 cases. Some or full secondary education, but fewer than five "0" level exan~s passed. i Some or full secondary education, with five "o" level exams passed, at least one "A" level, or some university education. ) Excludes 12 children of "other" ethnlclty or with missing information. Note that Table 2.5 uses cross-sectional data, representing all women at a single point in time, rather than showing the experience of an actual cohort over time. For this reason, the proportions breasffecding and amenorrhceic at increasing durations since birth do not decline in a steady fashion. For example, more mothers 14-15 months postpartum were breastfeeding at the time of the survey than were mothers of 12-13 month old children. To minimize such fluctuations, the births arc grouped in 2-month intervals. 2.4 Mean Duration of Breastfeeding and Postpartum Insusceptibility Table 2.6 and Figure 2.2 present the mean number of months of breasffe~ing, postpartum an~norrhoca, postpartum abstinence, and postpartum insusceptibility by background characteristics of the mother. These mean durations were calculated by dividing the total number of women breasffceding, amcnorrhceic or abstaining by the average number of births per month in the past 36 months. This technique is based on an epidemiological method of estimating the mean duration of a disease, calculated by dividing its prevalence by its incidence. 13 Table 2.5 Percentage of Births Whose Mothers are Still Breastfeedlng, Postpartum ~d~enorrhoelc, Abstaining, and Insusceptible, by Nuraber of Months since Birth, TTDBS 1987 Percentage Still Months Breast - ~anenor- Absta in - Insus- Number Since Birth feeding rhoelc Ing ceptlble ~ of Births Less than 2 81.4 88.4 86.0 97.7 43 2-3 83.6 46.3 22.4 56.7 67 4-5 55.6 19.4 13.9 33.3 36 6-7 49.4 16.1 13.8 26.4 87 8-9 36.1 8.2 3.3 11.5 61 10-11 29.7 12.5 3.1 15.6 64 12-13 26.9 3.0 3.0 6.0 67 14-15 32.8 6.9 3.4 10.3 58 16-17 16.7 O.O 1.9 1.9 54 18-19 17.1 2.9 4.3 7.1 70 20-21 17.1 0.0 1.3 1.3 76 22-23 12.7 0.0 0.0 0.0 63 24-25 18.3 0.0 0.0 0.0 60 26-27 5.9 2.0 0.0 2.0 51 28-29 6.1 0.0 0.0 0.0 66 30-31 7.0 0.0 1.4 1.4 71 32-33 7.5 0.0 0.O 0.0 80 34-35 4.7 1.6 1.6 3.1 64 Total 26.9 9.9 7.4 13.4 1,138 Median 6.3 2.3 i. 7 3.1 -- Note: Includes births 0-35 months before the survey. * Either amenorrhoeic or abstaining at the time of the survey. On average, women breastfeed their children for 10 months. (Note that the means in Table 2.6 appear elongated relative to the medians in Table 2.5. The small proportions of women who continue to breastfeed for 24 months or longer after delivery lengthen the mean but not the median. In addition, the time periods covered by Tables 2.5 and 2.6 are slightly different. Recall, too, that because the tables are calculated with quite different procedures, the means and medians are not strictly comparable.) The information is useful, however, for comparing breasffeeding practices among different groups of women. Younger and more educated women, who tend to be the forerunners of behavioral change, breastfeed for shorter durations than older and less educated women. This suggests that the practice may be declining. A decline in the already short period of breastfeeding has serious implications for the nutritional status of infants. One possible reason for the decline is a 31 percent increase in labour force participation by women 25-34, which occurred between 1970 and 1980 (Central Statistical Office 1987c). On average, menstruation resumed 3.5 months after the most recent birth, which is more than 6 months prior to the cessation of breastfeeding. This suggests that the intensity with which women breastfeod their children may be diminished due to the introduction of supplemental foods long before breasffeeding ceases. The relationship between duration of breastfeeding and menstruation is not consistent among different subgroups of the population. For example, the least educated women breastfeed longer than women in all other education groups, but experience the shortest durations of amenorrhoea. The small number of births to uneducated women may be responsible for the unexpected results. 14 Table 2.6 Mean Number of Months of Breastfeedlng, Postpartum Amenorrhoea, Postpartum Abstinence t and Postpartum Insusceptibility by Selected Background Character- istics, TTDHS 1987 Post- Post- Post- partum partum partum Number Background Breast- Am.nor- Abstl- Insuscep- of Character i s t i c feeding rhoea nence tiblllty ~ Births Age <30 9.6 3.5 2.6 4.6 801 30+ 11.2 3.5 2.8 5.1 362 Residence Urban 9.9 3.0 2.8 4.8 500 Rural i0.3 3.9 2.5 4.7 663 Education <complete primary 12.0 1.0 2.1 2.6 69 Completed primary 10.8 4.3 3.1 5.8 494 Secondary I* 9.8 3.2 2.5 4.3 501 Secondary II 3 6.9 2.5 1.5 3.3 99 Ethnlclty' African i0.0 3.6 3.2 5.3 452 Indian 10.6 3.2 1.7 3.8 496 Mixed 9.5 4.2 3.9 6.1 205 Total 10.1 3.5 2.6 4.8 1,163 Note: Includes births 1-36 months before the survey. * Either amenorrhoelc or abstaining at the time of the survey. * Some or full secondary education, but fewer than five "o" level exams passed. * Some or full secondary education, with five "O" level exams passed, at least one "A" level, or some university education. 4 Exc ludes 8 children of "other" ethniclty or with missing information. Figure 2.2 Duration of Breastfeeding, Amenorrhoea and Post-Partum Abstinence AGE ~30 30* RESIDENCE Urban Rural EDUCATION c Complete Prim. Complete Prim. Secondary I Secondary II ., . ,_,, . In . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 2 4 6 8 10 Mean Duration in Months I 12 [--7 Breastfeedlng k\\\\\~ Amenorrhoea ~B Abstinence i Trinidad & Tobago DH8 1987 15 16 CHAPTER 3 FERTILITY 3.1 Cur rent and Cumulat ive Fert i l i ty Information about the past and present fertility of women is among the most important information collected in the I IDHS. A full birth history was collected from each woman, including the name, sex, and month and year of each llve birth; the age at death for births that died; and whether or not living children reside with their mother. Table 3.1 presents the total fertility rates (TFR) for recent periods prior to the survey, and the mean number of children ever born (CEB) to women 40-49. The former figure is a measure of current fertility--the number of children that a woman would bear during her lifetime if she were to experience the age-specific fertility rates prevailing during a given period. Children ever born, on the other hand, represents cumulative fertility, and is a measure of past reproductive behaviour. Table 3.1 Total Fertility Rates (TFR] for Calendar Year Periods and for Five Years Preceding the Survey, a~d Mean Number of Children Ever Born (CEB) to Women 40-49 Years of Age, by Selected Background Characteristics, TTDHS 1987 Total Fertility Rate Mean Number o£ Children Calendar Calendar 0-4 Years Ever Born Background Period I Period II Preceding to Women Age Characteristic (1984-1987) I (1981-1983) the Survey 40-49 Years Residence Urban 3.0 3.0 3.0 3.8 Rural 3.1 3.6 3.2 4.8 Education <complete primary 3.6 4.2 4.0 5.3 Completed prir0ary 3.5 3.9 3.6 4.5 Secondary ~i 3.2 2.9 3.i 3.2 Secondary II ) 2.3 2.3 2.3 2.4 EthnIcity' African 3.4 3.6 3.5 4.3 Indian 2.7 3.1 2.0 4.4 Mixed 3.4 3.5 3.4 4.3 Total 3.1 3.3 3.1 4.3 Note: TFRs are calculated on women 15-49. Includes exposure up to the month prior to the interview in 1987. a some or full secondary education, but fewer than five "o" level exams passed. ) Some or full sgcondary education, with five "0" level exams passed, at least one "A" level, or some university education. ( Excludes 27 women of "other" ethnlclty, and one respondent with missing information. 17 In the five years prior to the survey, the "1'1~1~ was about 3.1. In the more recent period, 1984-1987) the rr,l,t is slightly lower than in the 1981-1983 period--3.1 versus 3.3 children. The I-ruHS figures are generally consistent with vital statistics data, which measured the TFR as 3.4 in 1981 and 3.0 in 1984 (see Figure 3.1). F igure 3.1 Total Fertility Rate, Various Years and Data Sources 375 3,50 3,25 3.00 2.75 1972-76 1981" 1981-83 1984. 1984-87 WFS TTDH8 TTDHS • Vital Statistics Trinidad & Tobago DHS 1987 As shown in Table 3.1, fertility rates differ markedly among subgroups of the population. In all periods for which TFRs were calculated, fertility is slightly higher in rural than urban areas. The largest fertility differentials are seen when educational background is considered. Whereas women with less than completed primary education have 4 children on average, those with completed secondary certification or university education are having only 2.3 children each. In the five years before the survey, the TFR for African women was 3.5, compared with 2.8 and 3.4 for East Indian and mixed women, respectively. Comparing the last two columns of Table 3.1 also indicates that fertility has declined in recent years. This information appears graphically in Figure 3.2. Women aged 40-49 had 4.3 children during their reproductive years, on average, which is more than one child greater than the level of current fertility. One trend worth noting is that while past levels of fertility were slightly higher among East Indian than African women (4.4 children ever born versus 4.3), currently, fertility is lower among East Indian than African women (2.8 versus 3.5). Another indicator of current fertility is the percentage of women who are pregnant. Of all women in the sample, 5 percent reported being pregnant. Of these 47 percent were under 25 years (see Table 3.2). i Includes exposure up to month prior to the month of interview in 1987. 18 Figure 3.2 Total Fertility Rate 0-4 Years Before the Survey, and Children Ever Born to Women 40-49 Years RESIDENCE Urban Rural EDUCATION ( Complete Prim. Complete Prim. Secondary I Secondary II ETHNICITY African Indian Mixed I I I I I I I 0 1 2 3 4 5 6 Number of Children m Total Fertility Rate Children Ever Born i Trinidad and Tobago DHS 1987 Table 3.2 Percent of All Women Who are Currently Pregnant by Age, TTDHS 1987 Percent Age Pregnant Number 15-19 4.1 683 20-24 8.6 745 25-29 7.5 745 30-34 5.3 543 35-39 3.4 441 40-44 0.3 370 45-49 0.4 279 3.2 Fertility Trends The age-period fertility rates presented in Table 3.3 show that, in the 0-4 years before the survey, teen fertility is rather low, 84 births per 1,000 women, more than doubles for women in their twenties, and then falls sharply for women age 35 and above. This table also confirms the fertility decline experienced in recent years. One can construct total fertility rates from TYDHS data for women 15-34 for up to 20 years in the past, and see a decline from 3.4 children born to women in this age group 15-19 years before the survey, compared to 2.7 children born to women of the same ages in the 5 years prior to the survey. The TFR, which summarizes the age-specific fertility rates for women 15-49, has declined from 3.4 in 1972-1976 (Hume 1983) to 3.1 a decade later (see Figure 3.1). 19 Table 3.3 Age-Perlod Fertility Rates (per 1,000 Women) by Age of Woman at Birth of Child, TTDHS 1987 Years Prior to Survey Maternal Age at Birth 0-4 5-9 10-14 15-19 20-24 25-29 30-34 15-19 84 92 94 98 115 20-24 181 204 199 226 306 25-29 164 187 173 218 (312) 30-34 114 109 139 (147) 35-39 67 69 (75) 40-44 17 (25) 45-49 (2) Cumulated Fertility Ages 15-34 2.7 3,0 3.0 3.4 Note: Numbers in parentheses are partially truncated rates. 140 (106) {328) Figure 3.3 permits a comparison between WFS and TTDHS data, and cortfirms the modest decline in fertility in recent years. The period 10-14 years prior to the TFDHS corresponds with the 1972-1976 time period. Fertility, as measured by the TI'DHS for that period was slightly higher for women 15-29, and lower for women 30-39 than that measured by the WFS (Hunte, 1983). Fertility in the 5 years preceding the TTDHS, however, is lower than that measured by either the TI'DHS or the WFS a decade earlier. 250 Figure 3.3 Age-Specific Fertility Rates, WFS and TTDHS 200 150 100 50 15-19 ~-~-. . i I I 20-24 25-29 30-34 35-39 - - WFS, t972-1976" 0-4 Yrs. Prior DH8 • Souroe: Hunte, 1983 --+ 10-14 Yrs. Prior DHS i Trinidad & Tobago DHS 1987 20 3.3 Children Ever Born Table 3.4 shows the distribution of all women, and those women currently in a union, by age and the number of children ever born (CEB) according to age. Childbearing begins relatively late in Trinidad and Tobago. Only 11 percent of all women under 20 have had a child, as have 46 percent of women aged 20-24 years. Not surprisingly, fertility is higher among women in union than among all women in each age group. On average, respondents have had 2.1 births, while women in union have had 2.7. However, twenty percent of the women 15-19 years are currently in a union, and of these 47 percent had at least one child. Also, 60 percent of women 20-24 years ate currently in a union, and of those, 70 percent had at least one child. Only about 4 percent of women 45-49 never have a birth, and this figure is reduced to 2 percent among currently married women. On average, women in union age 20-24 have had 1.4 children, those 30-34 have had 2.9, and those nearing the end of their reproductive lives have had 5 children. Table 3.4 Percent Distribution of All Women and Women in Union by Number of Children Ever Born, According to Age, TTDHS, 1987 children Ever Born Total Mean Per- NO. Age 0 1 2 3 4 5 6 7 8 9 I0+ cent Number CEB All Women 15-19 89.0 8.3 2.5 0.0 0.1 0.0 0.0 O.O 0.0 0.0 0.0 100 683 0.1 20-24 53.6 20.7 15.7 5.8 2.8 1,1 0.3 0.1 0.0 0.0 0.0 100 745 0.9 25-29 25.4 19.6 25.1 15.0 8.2 3.9 1.9 0.3 0.5 0.0 0.i i00 745 1.9 30-34 14.0 10.5 27.6 18.0 14.5 7.9 4.1 1.7 0.6 0.6 0.6 i00 543 2.7 35-39 ii.I 10.2 18.4 20.6 14.7 11.6 5.0 4,1 2.0 i.i i.i i00 441 3.2 40-44 9.2 8.6 16.5 15.9 13.5 14.1 7.6 3.2 4.1 4.3 3.0 i00 370 3.9 45-49 3.9 6.1 10.8 11.8 15.8 9.0 12.2 13.6 6.1 4.7 6.1 i00 279 4.9 Total 35.9 13.3 16.9 11.5 8.4 5.5 3.2 2.1 1.3 1.0 1.0 100 3,806 2.1 Women Currently In Union 15-19 53.2 34.5 12.2 0.0 0.0 O.O 0.0 0.0 0.0 0.0 O.O 100 139 0.6 20-24 30.0 30.7 23.8 9.7 4.4 1.8 0.4 0.2 0.0 0.0 0.0 100 450 1.4 25-29 15.7 21.3 28.8 17.0 9.5 4.6 2.1 0.3 0.6 0.0 0.0 100 624 2.1 30-34 8.5 10.4 29.2 20.0 15.8 8.7 4.3 1.3 0.6 0.4 0.6 100 469 2.9 35-39 6.9 9.5 19.5 22.4 16.5 11.6 5.1 4.1 2.1 1.0 1.3 i00 389 3.4 40-44 5.1 7.1 17.3 17.3 13.1 15.7 7.7 3.8 4.5 4.8 3.5 100 312 4.2 45-49 2.1 6.4 10.3 12.8 16.2 8.5 12.4 14.5 5.6 5.1 6.0 100 234 5.0 Total 15.1 16.9 22.7 15.7 11.3 7.3 4.1 2.7 1.6 1.3 1.3 100 2,617 2.7 3.4 Children Ever Born and Age At First Union Women who enter a union at a young age can be expected to have more children than those entering at an older age, since they are likely to have a longer time period of exposure to the risk of pregnancy. As seen in Table 3.5, fertility generally declines as age at first union increases. For example, among women married for 10-14 years, those who married before age 22 had between 2.8 and 3.2 children each, compared with 2.1 children born to women marrying at age 25 or above. However, for women in union less than 10 years, there seems to be little relationship between age at fast union and fertility. 21 Table 3.5 Mean Number of Children Ever Born to Women Ever in Union by Age at First Union and Years Since First Union, TTDHS, 1987 Years Age at First Union Since First All Union <15 15-17 18-19 20-21 22-24 25+ Ages 0-4 * 0.8 0.6 0.6 0.8 0.6 0.7 5-9 2.8 2.2 1.9 2.0 1.8 1.6 2.0 10-14 2.9 3.2 2.9 2.8 2.4 2.1 2.8 15-19 4.1 3.7 3.6 2.9 2.8 2.7 3.4 20-24 4.3 4.5 4.0 3.4 4.0 3.7 4.1 25-29 6.0 5.5 5.2 4.0 4.2 -- 5.2 30+ 6.0 6.2 5.0 . . . . . . 5.9 Total 3.7 3.1 2.7 2.3 2.1 1.6 2.7 • Fewer than 25 cases. -- Not applicable. 3.5 Age at F i rs t B i r th Table 3.6 presents data on the age at first birth by the woman's current age. While only 1 percent of the women had a birth before age 15, 30 percent had at least one child while still in the teen years. One can compare this table with Table 2.2 and assess the gap between age at first union and first birth. While the median age at first union for women aged 25-29 was 19.8, the median age at first birth for these women was 22.2. While the median age at union has not changed in recent years, the age at first birth has risen steadily from 20.5 for the oldest women, to 22.2 years for women 25-29. This gap--nearly two years--raises questions about the delay in onset of fertility after entry into union. Use of contraception at this point is addressed in Chapter 4. Table 3.6 Percent Distribution of Women by Age at First Birth According to Current Age, TTDHS, 1987 Age At First Birth Total Current No Per- Median Age Bir th <15 15-17 18-19 20-21 22-24 25+ cent Number Age 15-19 89.0 0.3 7.3 3.4 0.0 0.0 O.0 I00 683 -- 20-24 53.6 0.5 12.2 17.0 12.1 4.6 0.0 100 745 -- 25-29 25.4 1.2 13.6 16.4 17.3 18.7 7.5 I00 745 22.2 30-34 14.0 1.3 14.4 19.2 16.0 17.1 18.0 100 543 21.9 35-39 ii.i 0.9 14.1 20.6 18.1 16.6 18.6 100 441 21.6 40-44 9.2 3.0 15.7 21.4 16.5 14.6 19.7 i00 370 21.1 45-49 3.9 3.6 19.7 22.2 17.6 16.8 16.1 100 279 20.5 Total 35.9 1.2 13.0 16.0 13.0 11.6 9.3 I00 3,806 -- -- omitted due to censoring. 22 3.6 Med ian Age at F irst Birth by Background Character is t ics Table 3.7 shows the median age at fast birth according to selected background characteristics of the respondent. Women residing in urban areas delay the first birth by almost one year compared to their rural-dwelling counterparts. Variations by education are more dramatic. Whereas women with less than primary education have their first child before age twenty, those with some secondary education delay the first birth by an additional 3.4 years; those with full secondary certification delay by a further 2.5 years. Differences by ethnicity are slight. Particularly interesting is the increase in the age at first birth for all ethnic groups in recent years. The sharpest increase occurred among East Indian women, from 19.9 for women 45-49 to 22.0 for those 25-29. This rise in the age at first birth could be partly responsible for the lower fertility now experienced by East Indian women (see Table 3.1), and could also be a result of contraceptive practice, a subject to be addressed in the following chapter. Table 3.7 Median Age at First Birth Among Women Aged 25-49 Years, by Current Age and Selected Background Characteristics, TTDHS 1987 Current Age Background Ages Characteristic 25-29 30-34 35-39 40-44 45-49 25-49 Residence Urban 23.1 21.9 22.4 22.0 21.1 22.1 Rural 21.8 21.9 21.0 20.6 19.9 21.3 Education <Complete primary 20.4 19.7 19.4 19.4 19.1 19.5 Completed primary 21.1 20.7 20.7 20.4 20.3 20.7 Secondary I' 22.9 22.9 23.2 22.8 23.0 22.9 Secondary II* 25.9 25.4 25.0 25.5 24.0 25.4 Ethniclty* African 22.5 21.3 21.3 21.1 20.7 21.6 Indian 22.0 22.2 21.5 20.9 19.9 21.5 Mixed 22.2 22.0 22.3 22.6 21.4 22.2 Total 22.2 21.9 21.6 21.1 20.5 21.7 * Some or full secondary education, but fewer than five "o" level exams passed. • Some or full secondary education, with five "o" level exams passed, at least one "A" level¢ or some university education. • Excludes 21 women of "other" ethnlclty or with missing information. 23 24 CHAPTER 4 FERTILITY REGULATION In the past 30 years, the use of contraception has played a key role in fertility decline. However, a sign that the decline was leveling off was an important reason for implementing the TTDHS. Levels and trends of contraceptive knowledge and use were particular areas of investigation in the survey. 4.1 Knowledge of Methods Respondents were asked to name all methods of family planning they had ever heard of. Interviewers then probed to see if respondents had heard of any methods not named spontaneously. Then, for each method known, women were asked if they had ever used it, if they knew of a place to obtain it, and if there were any problems they had heard about with using it. Following that, a detailed contraceptive history was recorded, including the use of methods since the last birth, and between births in the last 5 years. Overall, knowledge of contraceptive methods is quite high--97 percent of all respondents, and 99 percent of those currently in union know at least one modem method of contraception (Table 4.1). The pill, condom, female sterilization, and IUD are the most widely known methods, and were familiar to more than 90 percent of women in union. Injection and vaginal methods are known by about 80 percent, while withdrawal, male sterilization, and the safe period are less well known. Knowledge of specific methods is only slightly lower for all women, suggesting that women who are not in union are quite likely to be knowledgeable about contraception. The level of knowledge of methods according to the age of respondents follows the usual pattern where knowledge is higher among women in the intermediate age groups, and lower among the youngest and oldest women. While knowledge of at least one modem method is only slightly lower for women 15-19 than for older women, larger disparities emerge when looking at knowledge of specific methods. This pattern is particularly pronounced in the table showing "all women," indicating lack of knowledge among young women not in unions. For example, vaginal methods are known by only 55 percent of women 15-19, compared to 80 percent of women 25-29. While several methods are relatively unknown among the youngest women, educational efforts directed at them might focus on temporary methods, such as the pill, condom, and vaginal methods. Table 4.2 shows the percentage of women in union who know at least one modem method according to the number of living children the woman has and certain background characteristics. Knowledge is nearly universal--more than 92 percent of all subgroups of women know of a modem method. 4.2 Problems with Methods As Table 4.3 shows, respondents who know of methods cite few problems that they have heard of, with the exception of the pill, for which more than 60 percent of respondents named problems. Health concerns are the leading problems mentioned for the pill (57 percent), the IUD (34 percent), injection (27 percent), and female sterilization (17 percent). Ineffectiveness is most often mentioned for the condom (23 percent), the safe period (27 percent), withdrawal (24 percent), and vaginal methods (9 percent). Very few women mentioned problems obtaining methods, that religious beliefs rendered certain methods unacceptable, or that partners disapproved of methods. It is somewhat surprising that disapproval by partners was not mentioned more often as a problem with specific methods, particularly with the condom, withdrawal, safe period and male sterilization. One possible reason for this is that respondents may be reluctant to give the impression that their partners influence their thinking about contraception. 25 0 4 ~ ~ 4 0 0 ,,-i u~ 4 • O @ 0 ~ J U ~ 0 iJ 0 @ 0 0 m ~ U I ~ 0 0 0 @ g ~ o. N o @ m p I m @ m I @ ~ g J m l @ O @ . . . . . . ~ o ~ 2 6 Table 4.2 The Percentage of WC~n in Union Who Know at Least One Modern Contraceptive Method, by Number of Living Children and Selected Background Characteristics, TTDHS 1987 Number of Living Children Background Women Cha~acteristlc 0 1 2 3 4 5 6+ In Union Residence Urban 98.6 98.6 99.3 100.0 98.3 i00.0 97.5 99.0 Rural 99.0 99.6 97.5 99.6 99.4 i00.0 97.0 98.8 Education <Complete primary I00.0 1O0.0 97.4 i00.0 97.4 1OO.0 92.4 96.8 Completed primary 99.1 98.1 97.5 99.6 98.9 i00.0 99.3 98.8 secondary I* 98.1 99.6 99.8 100.0 I08.0 I00.0 100.0 99.1 Secondary II: 100.0 i00.8 I00.0 100.0 100.0 I00.0 100.0 100.0 Ethnicity' African I08.0 98.4 98.5 100.0 97.8 100.0 98.9 99.0 Indian 98.1 108.0 98.0 99.5 99.3 i00.0 95.3 98.7 Other 97.6 98.7 99.0 I08.0 i00.0 10O.0 98.0 98.9 TOTAL 98.8 99.1 98.4 99.8 99.0 i00.0 97.2 98.9 Some or full secondary educatlon, but fewer than five "o" level eMams passed. : Some or full secondary education, with five "o" level exams passed, at least one "A" level, or some university education. * Excludes 19 women of "other" ethniclty or with missing information. Table 4.3 Percent Distribution of Women Who Have Ever Heard of a Contraceptive Method by Main Problem Perceived in Using the Method, TTDHS 1987 Vag- Female Male Inal Sterl- Steri- InJec- Meth- llza- liza- Safe With- Main Problem Pill IUD tlon ods" Condom tlon tion Period drawal No problem 39.0 52.7 68.4 78.5 65.5 67.9 84.5 71.4 68.1 Partner disapproves 0.0 0.I 0.0 0.4 4.4 0.i 0.4 0.0 1.9 Health concerns 57.0 34,1 26,7 6.8 b i,I b 16,5 I,i 0,2 b 0,9 b Not effective 3.6 11.7 4.0 8.8 23.3 4.2 1.7 26.5 23.9 Method permanent 0.0 0.0 0.2 0.0 0.0 10.8 10.2 0.0 0.0 Inconvenient to use 0.2 0.9 0.i 4.1 2.5 0.0 0.0 1.2 3.4 Other 0.0 0.I 0.0 0.8 2.5' 0.i 0.2 0.i 0.i Don't know/Not stated 0.2 0.5 0.4 0.6 0.6 0.5 1.7 0.6 1.6 Total Percent i00.0 100.0 100.0 100.0 100.0 100.0 i00.0 100.0 108.0 Number 3,532 3,090 2,678 2,813 3,523 3,396 2,225 1,760 2,652 Includes diaphra~, fo~,~, Jelly, and foaming tablets. May include women who said that the method was "not safe" i.e., not effective. 4.3 Knowledge of Source Survey respondents who have heard of a method were asked if they knew of any place where it could be obtained. As Table 4.4 shows, nearly all women could name at least one source where methods could be obtained, reflecting the widespread availability of contraceptives discussed in Chapter 1. The four types of sources most commonly identifiod are the government's health centres, FPATT clinics, private outlets (doctors, hospitals and nursing homes) and pharmacies. 27 One should interpret information on individual supply sources with caution. While interviewers were trained to elicit specific information, it seems that respondents could not always distinguish government outlets from FPATT clinics. Government outlets were the most often mentioned source of the pill, IUD, vaginal methods, and male and female sterilizations. Injection was associated with private sources by most respondents, while condoms were most commonly identified with pharmacies. The FPATT was the second most commonly cited source for the pill, IUD, vaginal methods, and instructions for using the safe period. Table 4.4 Percent Distribution of Women Who Have Ever Heard of a Contraceptive Method by Supply source Named, According to Specific Method, TTDHS 1987 METHOD Vag- Female Male inal Steri- Steri- InJec- Meth- llza- liza- Safe source Pill IUD tlon ode I Condom tion tlon Period Government Health Centre 38.2 39.9 26.1 37.3 27.4 61.1 49.0 26.0 FPATT 23.4 26.5 19.9 25.1 15.3 14.7 15.8 26.9 Private Sources = 16.5 24.7 47.1 14.6 3.2 20.6 25.0 27.0 Pharmacy 19.2 1.4 0.2 17.6 50.6 0.1 0.0 0.3 Archdiocesan Family Life Commission 0.0 0.0 0.0 0.0 0.0 0.0 0.0 6.7 Other 0.4 0.5 0.4 0.2 0.5 0.4 0.6 6.8 Don't Know 2.3 7.1 6.3 5.0 3.0 3.0 9.5 6.3 Total Percent 10O.0 100.0 I00.0 I00.0 I00.0 i00.0 i00.0 I00.0 Number 3,532 3,090 2,678 2,813 3,523 3,396 2,225 1,760 * Includes diaphragm, foam, Jelly, and foaming tablets. = Includes private doctors, private hospitals, and private nursing homes. 4.4 Ever Use o f Cont racept ion As Table 4.5 shows, 63 percent of all women have used a method of contraception at some time, and 60 percent have used a modem method. As expected, ever use of contraception is higher among women who are cun~nfly in union than among the entire sample population. That 79 percent of women in union have used a modem method at some time demonstrates widespread acceptance of family planning. Among women in union, the pill has been most widely used (56 percent), followed by the condom (49 percent), withdrawal (30 percent) and vaginal methods (23 percent). Among all women, those in their thirties are the most likely to report ever use of a method (81 percent), while the oldest and youngest women are the least likely. 28 U 0 ¢ J O - , - t k - I . , - I U ¢ 0 n l "O W .-i O U l O ~ m 'O ~ O ,li -,-4 0 ) O 1 ¢ 1 ~ O ,--t 0 E~ 0 , - I ; ! 0 W 0 ~ E N O E 0 C 0 0 I > ~ 0 I .n O H ~ -,-t W O O ~ ~ 4 J " 0 0 l @ @ t @ I @ 0 0 0 0 0 0 0 0 E < 0 ~ 0 ~ • @ I @ I @ I I 0 ~ 0 ~ @ l l i l @ l I I I I I I I I 0 @ I ~ O I @ I O O O O O O O O O O * . . o . * = • @ l @ @ l @ l I 1 1 1 1 1 1 0 ~ d # J N ~ d 0 4J E o ~ J N -,-I 4 J ,-4 O • O 0 ) O .--I O ~ O C ,--t O ~ ~ O d ~ H ~ O .'O .--t ,'-I I--II--4 29 Some interesting patterns of method use by age appear in Table 4.5. The youngest users are more likely to have tried less effective and temporary methods such as withdrawal and the condom, while older women are more likely to have tried the pill/ These data point to a pattern of women trying several methods over the course of their reproductive lives, but with limited use of the more effective and permanent methods such as sterilization and the IUD. Figure 4.1 Current Use of Family Planning by Method Women in Union 15-49 Female Sterilization 8% Vaginale IUD 4 q Other Modern 1% Traditional 8% Condom 12% Pill 14"/, ~ol unin9 47% Trinidad & Tobago DHS 1987 4.5 Current Use Table 4.6 presents data on current use of contraception among all women and among those in union. The distribution of methods used by women in union is shown in Figure 4.1. The subsequent discussion is limited to data on women in union, the population of greatest interest to service providers. Fifty-three percent of women in union report using a method at the time of the survey. The pill and the condom are the most widely used methods, reported by 14 and 12 percent of respondents respectively, followed by female sterilization (8 percent), and withdrawal and vaginal methods (5 percent each). In view of the relatively high level of acceptance of male temporary methods, the almost complete absence of male sterilization is worth noting. According to a comprehensive review of contraceptive failure rates in the United States, the percentage of typical couples who would experience .an accidental pregnancy during the first year of use of various methods is: male sterilization 0.2, female sterilization 0.4, pill 3, IUD 6, condom 12, vaginal methods 18-28, withdrawal 18, safe period 21, chance 89 (TrnsseU and Kost 1987). 30 -0 0 U *4 ~ J -0 0 @ 0 U . 0 o -o o ~ J I1 U W g ¢ 1 1 0 0 ~ k ,.l~ Iii I0 0 ~ I ~ .~ ~ 0 m ~ o o o o o o ~ o ° ° ° ° ° ° ° o o o o o o o O 0 0 0 0 0 0 0 I J g U U l l U o o ~ o @ l l l @ l l @ , , , , , , , . l @ ~ I g J @ g ~ @ @ I U U @ U o - - - o - - I I I I I I I O Q O O Q O 0 0 ° ° ° ° ° ° ° o o o o o o o l l @ b ~ g • O 0 0 0 0 0 0 O O 0 0 0 0 0 Q 0 @ l l @ I I @ # e l u e l m 4 @ I I @ B @ 4 1 I l l l l l l l " 0 .o 4 4 0 O) ~ J ~ 0 , - * nl I ~ . " 0 ' I ~ " o 4J ~ I 31 The youngest and oldest women are less likely to use methods than women in the intermediate years. Forty-two percent of women in the youngest age group use contraception, compared to 57 percent of those 30-34, and 36 percent among the oldest women. This Ipattem may reflect the desire for children on the part of the younger women, and a combinauon of traditional values and infecundity on the part of older women. While the pill is the most common method among women under 35, and female sterilization among women 35 and over, it should be noted that the next most popular methods generally are temporary and less effective--the condom and withdrawal. The thirty point difference between ever use and current use, together with the heavy reliance on temporary methods suggests a quite high drop-out rate (see Figure 4.2). Temporary methods require a continuing effort on the part of the family planning programme and supply network to make methods available, and to maintain users' motivation. In view of the limited use of permanent methods, particular attention should be paid to subsequent discussions of reasons given for discontinuation and non-use of contraception. 100 90 80 70 60 50 40 30 20 10 0 Figure 4.2 Family Planning Knowledge and Use Women in Union 15-49 Percent 99 Knows method Ever Used Currently Using Trinidad and Tobago DHS 1987 Figure 4.3 shows current use of any method by age for women in the TTDHS and the WFS (Sathar and Chidambaram 1984). Overall, there has been virtually no change in current use in the decade between the two surveys. FkPty-two percent of women in union were using methods in 1977, versus 53 percent in 1987. Currently, women 25-34 are slightly less likely, and women 40-49 more likely to be using methods than was the case a decade ago. Table 4.7 permits the comparison of contraceptive use and the method mix among sub- groups of the population. Women in urban areas are slightly more likely to report current use of methods than women in rural areas. Use of contraception varies dramatically with education, as seen in Figure 4.4; while only 41 percent of women with less than primary education are current users, nearly 68 percent of those with full secondary certification are currently using a method. Female sterilization is preferred among women with less than full primary education (reflecting their older age distribution), while the pill is preferred among women in all other education categories. The second most popular method at all levels of education is the condom. 32 70 50 50 40 30 20 15-19 Figure 4 .3 Current Use by Age, TTDHS and WFS Women in Union 15-49 I I I I I 20-24 25-29 30-34 35-39 40-44 45-49 TTDHS + WFS Trinidad & Tobago DHS 1987 East Indian women arc slightly more likely to be current users than are African women (56 versus 49 percent), with the "mixed" category falling in between. Higher prevalence in the East Indian population is consistent with their lower current fertility as discussed in Chapter 2. Also, higher contraceptive prevalence among East Indian women represents a reversal from the situation a decade ago when the 1977 WFS reported that East Indian women had higher fertility and lower contraceptive prevalence than African women. (Note: women of mixed ethnicity are included with non-Indian women in the WFS data. However, since they represent a small proportion of the population, their inclusion does not substantially affect the comparison.) East Indian women are twice as likely to use condoms as Roman Catholic women, and are more likely to use withdrawal, as well. As expected, contraceptive use is least common among women with no living children (32 percent), rises smadily to 61 percent of women with three children, and declines slightly among women with 4 or more children. The table also shows that higher parity women prefer female sterilization while women with fewer children am morn likely to choose the pill or condom. Hindu women are most likely to be current users. Women who report their religion as Roman Catholic are slightly more likely than all women to be current users of contraception. Differences by method am slight. 4.6 Parity at First Use The timing of introduction of contraception has implications for the choice of method. First use of contraception early in the family-building process implies a postponement of the first birth and the need for temporary methods of contraception; first use at later stages implies the need for more permanent methods to limit births. As shown in Table 4.8, 28 percent of women who have ever been in a union had no living children when they first used a method of contraception, while 25 percent had one child, 12 percent had two children, and 16 percent had three or more. 33 '0 ¢ : 0 W m @ 0 @ @ 0 0 0 @ @ W W (.3 ,,-t @ M 0 U 121 G ~ I . C ,~o o ° ° o ° o o o . ° ° ° ,~ooo°°° ooooOooooo oooOOO ,~ ~,. o. ~. ~. o. ~. ~. o. o o o o o o o o o ~ - . . . . . ~ . ~ . ~ . ~0 o o o o o o o o o ~ . o ~ ,~ . ~. ~. ~. ~ o ~ ~ ~ ~ = ,~ ~ ~ o o . ~ o ~ , ' " ' ' ' ~ - ~ , ® ' ' ' ~ , ~ ~ ~. ~. ~. ~. ~. ~. ~ . ~ . ~ . . . . . ~ = ~ . . . o. ~ .o. o ~. o, =. ~. ~. ~. o~ o~ o. ~. ~. =. ~. ~ . ~ . ~. o. ~. ~. ~. ~. ~. ~ ~ ~ , ~ ~- ~ , ~ , ~. ~ ~. ~. ~. ~. ~. ~. ~. ~ o ~ o. , ~ . ~ ,~ ~ ~ . ~ . ~ ~ ~ ~ - = = ~ - ' ' ' ' ' . . . . ~ ~{ ~ . ~ .~ ,~ C t1 ~ @ I o o 0 0 • 4J @ @ 0 O X ,,~ 0 I~ .,-4 O ~ ~ 0 0 ~ 3 4 80 70 60 50 40 30 20 10 0 Figure 4.4 Current Use of Family Planning by Education and Number of Living Children Women in Union 15-49 Percent *Prim. Prim. Sec. I Sec.n 0 1 2 3 4* EDUCATION NO. OF CHILDREN Trinidad & Tobago DHS 1987 Table 4.8 shows that there has been an increase in the proportion of women who first used contraception bcforc having any children. Over 40 percent of women 15-24 used a method of contraception before they had any children, compared to 25 percent of women 30-34, and 10 percent of women aged 45 and above. Table 4.8 Percent Distribution of Women Ever in Union by Number of Living Children at Time of First Use of Contraception, According to Current Age, TTDHS, 1987 Number of Living Children At Time of First Use Current Never Total Age Used 0 1 2 3 4+ Percent Nu~er 15-19 34.1 46.1 18.6 1.2 0.0 0.0 100 20-24 18.5 41.9 27.2 8.3 3.0 1.2 100 25-29 14.3 34.6 30.7 12.1 5.1 3.3 100 30-34 15.0 25.4 29.5 18.1 6.6 5.4 100 35-39 16.4 21.4 24.9 15.5 9.6 12.2 i00 40-44 20.3 16.7 16.7 12.8 Ii.I 22.3 I00 45-49 27.3 9.8 9.5 11.6 12.0 29.8 I00 Total 18.6 28.4 24.6 12.4 6.7 9.2 100 167 508 671 519 426 359 275 2,925 35 4.7 Age at Sterilization On average, respondents who wew sterilized or whose husbands were sterilized were 32 years old at the time of the operation (see Table 4.9). In many countries, the age at sterilization drops as women complete their families at lower parity leveis, hut this is not the case in Trinidad and Tobago, where there has been no consistent trend in recent years. Table 4.9 Percent Distribution of sterilized Women by Age at Time of Sterilization, According to the Number of Years Since the Operation, TTDHS 1987 Age at Time of Operation Years Since Total Median Operation <25 25-29 30-34 35-39 Percent Age Number <2.0 7.7 25.0 32.7 34.6 I00 32.2 52 2.0 - 3.9 7.9 28.9 31.6 31.6 I00 31.0 38 4.0 - 9.9 5.3 22.8 35.1 36.8 I00 33.1 57 i0+ 8.3 33.3 36.7 21.7 1O0 31.4 60 All Women 7.2 27.5 34.3 30.9 100 32.2 297 4.8 Knowledge of the Reproductive Cycle Knowledge of the female reproductive cycle provides a useful background for successful practice of coital-related n~thods, and is essential for preventing pregnancy while using the safe period. Respondents were asked when during the monthly cycle they thought a woman was the most likely to become pregnant. As Table 4.10 shows, 50 percent had no idea, and only 18 percent correctly responded that the middle of the cycle is the fertile period. Table 4,10 Percent Distribution of All Women and women Who Have Ever Used the Safe Period by Knowledge of the Fertile Period During the Ovulatory Cycle, TTDHS 1987 Ever Users of the Safe Fertile Period All Women Period During menstrual period 1.8 Just after period has ended* 19.6 In the middle of the cycle 17.9 Just before period begins* 10.8 At any time 49.6 Other/Not stated 0.3 0.9 23.0 47.2 11.3 17.6 0.0 Total percent i00.0 I00.0 Number 3,806 335 * Includes numerous respondents who said "Just before and Just after the period." 36 While women who have ever used the safe period are more than twice as likely to know when the fertile period occurs (47 percent), more than one-half of this group did not know the correct answer, either. In view of the popularity of coital-related methods among couples in Trinidad and Tobago, these data imply a need for better education about the reproductive system. 4.9 Knowledge and Use of Pap Smears Cancer of the cervix is the second leading cause of cancer deaths to women in Trinidad and Tobago (Central Statistical Office 1987b). Because the pap smear is an invaluable tool for the early detection of this disease ff performed regularly, all respondents were asked if they had heard of the test, and had had one performed. Women were asked where they had the smear taken the last time, and if the test was done in the past year. As Table 4.11 shows, 57 percent of respondents have heard of the pap test, 31 percent have ever had a test, and only 11 percent have had a test in the last year. Knowledge of the procedure increases from 31 percent of women 15-19 to 64 percent of women in their thirties. Despite the increased risk of cervical cancer with age, knowledge drops to 56 percent among women 45-49. Urban women are much more likely to know of the procedure than rural women. Education is even more strongly related--89 percent of women with full secondary certification know about the test, compared with only 32 percent of those lacking complete primary education. Similar pattems are seen for whether women have ever had a pap smear, and whether they have had one in the last year. Table 4.11 Percentage of Women Ever in Union Who Know About the Pap smear, Have Ever Had a Pap smear, and Have Bad a Pap Smear In the Last Year, According to Selected Background Characteristics, TTDHS 1987 Know Ever Had Smear Background About Had In Last Characteristic Smear smear Year Age 15-19 31.0 6.5 3.6 20-24 48.0 15.2 7.1 25-29 58.7 24.9 10.9 30-34 63.6 38.5 13.3 35-39 63.6 42.5 15.0 40-44 60,2 39.8 11.4 45-49 56.4 41.8 9.8 Residence Urban 67.4 40.9 14.7 Rural 47.8 21.7 7.5 Education < Complete primary 31.5 22.8 4.6 Completed primary 49.7 27.4 S.4 Secondary I* 63.8 29.1 11.5 Secondary III 88.6 57.4 24.9 Number of Living Children 0 58.5 26.5 12.3 1 56.1 24.3 10.7 2 62.2 34.1 13.3 3 58.4 35.1 12.1 4 55.7 34.6 10.2 H 50.3 33.1 7.2 6+ 45.9 26.4 3.8 Total 56.8 30.6 10.8 Some or full secondary education, but fewer than five "O" level exams passed. * Some or full secondary education, with five "0" level exams passed, at least one "A" level, or some university education. 37 4.10 Source of Contraceptive Methods and Satisfaction with Services Respondents were asked where they obtained their current method of contraception. These responses are presented in Table 4.12 and grouped according to whether the method requires regular resupply or infrequent clinic visits. The pharmacy is the leading source for current users of each of the supply methods, followed by Government health centres, and the FPATT. Overall, government health eentres are the leading suppliers of clinical methods, providing 67 percent of female sterilizations and 44 percent of IUDs. The FPATT is the second leading provider of those two methods. Private clinical sources are the third leading supplier of methods requiring a visit to a medical facility. The supply sources for all methods appear in Figure 4.5. Table 4.12 Percent Distribution of Current Osers by most Recent Source of Supply, According to Specific Method, TTDHS 1987 Supply Methods Clinic Methods Female Vaginal Sterl- Source of Supply Pill Methods x Condom Total IUD llzatlon Total* Government Bealth Centre 30.S 34.6 28.7 30.5 43.8 67.1 55.7 FPATT 7.2 15.8 11.5 10.2 32.2 19.3 24.3 PriVate sources' 8.5 0.8 1.0 4.4 23.1 11.8 18.1 Pharmacy 53.2 48.i 56.7 53.7 0.0 0.0 0.D Other/Not Stated 0.6 0.8 2.2 1.2 0.8 1.7 1.9 Total Percent i00.0 i00.0 I00.0 i00.0 I00.0 i00.0 i00.0 Number 376 133 314 823 121 228 375 * Includes diaphragm, foam, Jelly, and foaming tablets. • Includes 21 users of injection, and 5 users of male sterilization. * Includes private doctors, private hospltals and private nursing homes. Current users of any method who have visited a contraceptive supply source in the past 12 months were asked if there was anything they disliked about the services they received there. As shown in Table 4.13, 90 percent of these women reported no problems with the services. The leading problem given was that the wait was too long, but even that was cited by fewer than 3 percent of those visiting contraceptive supply outlets. 4.11 Discontinuation of Contraceptive Use As mentioned earlier in this Chapter, the large gap between current and ever use of various contraceptive methods implies method switching and discontinuation. Respondents were asked if they had ever discontinued using a method in the five years before the survey, and were probed for reasons for the discontinuation. Table 4.14 presents reasons for the most recent discontinuation according to the method discontinued. (Note that the table does not indicate whether or not the woman began using again after the stoppage.) 38 Figure 45 Source of Family Planning Supply Current Users Governmenl ~er 1% FPATT 15¢ armacy 37% Private Sources* 9% • Includes private doctors, hospitals and nursing homes Trinidad & Tobago DHS 1987 Table 4.13 Percent of Current Users of Modern Methods Who Sald There Were No Problems Wlth the Service, by Type of Source Last Visited, TTDHS 1987 Source of Supply NO Problems Number Government Health Centre 89.3 460 FPATT 89.1 175 Private Sources* 94.2 104 Pharmacy 93.7 442 Other 0.0 7 Don't Know/Not Stated 0.0 1O Total Percent 90.1 1,198 * Includes private doctors, private hospitals, and private nursing homes. Nearly one-third of women in the sample have discontinued using a method of contraception at least once in the five years preceding the survey. The three main reasons women stopped using methods were health concerns, the desire to become pregnant, and method failure. In fact, health concerns were cited as a reason for cessation by more than half of discontinuers of the IUD and injection, and 43 percent of former pill users. A common reason given for discontinuing use of less effective rr~thods of contraception was method failure, including 42 percent of those who stopped using withdrawal, 32 percent who discontinued using the safe period, 24 percent who stopped using vaginal methods, and 20 percent of those who stopped using the condom. In view of the small proportion of women who understand the ovulatory cycle, as presented in Table 4.10, the large number of women who had failures with coitus-related methods is not surprising. 39 h should be pointed out that the information collected in the TTDHS does not permit an assessment of whether method failure resulted in pregnancy or was merely a concern of the respondent. Also, it is not known whether method failures resulted from the improper or inconsistent use of methods, or from failures of the methods despite correct usage. However, particularly in view of the high drop-out rates among contraceptive users, the area of discontinuation of contraception warrants further investigation. Table 4.14 Percent Distribution of Women Who Have Discontinued a Contraceptive Method in the Last Five Years by Main Reason for Last Discontinuation, According to Specific Method, TTDHS, 1987 Method Discontinued Vag- Reason for InJec- inal Safe With- All Discontinuation Pill IUD tlon Methods" Condom Period drawal Other Methods To become pregnant 21.9 21.7 i0.0 22.2 27.1 29.8 20.3 23.1 23.1 Method failed i0.i 17.4 i0.0 24.3 20.2 31.6 42.3 23.1 19.2 Partner disapproves 0.7 0.0 2.0 2.1 11.6 3.5 3.3 0.0 4.2 Health concerns 43.2 50.7 56.0 13.2 b 7.5 ~ 3.5 ~ 0.8 3.8 24.2 Access/availabillty/ expensive 4.3 0.0 8.0 4.2 6.0 0.0 0.0 0.0 3.9 Inconvenient to use 1.1 1.4 2.0 13.2 6.5 12.3 7.3 19.2 5.5 Infrequent sex 6.7 1.4 2.0 4.9 5.7 7.0 10.6 7.7 6.2 Change to permanent method 0.4 1.4 6.0 0.0 2.7 1.8 3.3 3.8 1.7 Other 8.0 2.9 0.0 13.9 i0.i 7.1 7.3 11.5 8.6 Don't know 2.2 2.9 2.0 2.1 1.8 1.8 3.3 3.8 2.2 Not stated 1.3 0.0 2.0 0.0 0.9 1.8 1.6 3.8 1.1 Total i00.0 100.0 100.0 100.0 100.0 i00.0 I00.0 100.0 i00.0 Number 447 69 50 144 336 57 123 26 1,252 " Includes diaphragm, foam, Jolly, and foaming tablets. b May include women who said that the method is "not safe" i.e., not effective. 4.12 Attitude Toward Becoming Pregnant and Reasons for Nonuse of Contraception In order to investigate reasons for nonuse, all nonpregnant nonusers who had ever had intercourse were asked whether they would be happy or unhappy if they became pregnant in the next few weeks. According to Table 4.15, 52 percent said that they would be unhappy if they became pregnant in the next few weeks, yet they were not using contraception at the time. Thirty- one percent of women who were asked said they would be happy if they became pregnant, while 15 percent said it would not matter. Not surprisingly, the more children women have, the more likely they are to express unhappiness about becoming pregnant. While 28 percent of childless women would be unhappy to become pregnant in the next few weeks, this figure escalates to 75 percent for women with four or more children. Table 4.16 presents the reasons given for not using family planning among women who have ever had sex, are not using contraception, but who said they would be unhappy about becoming pregnant in the next few weeks. While only a small proportion of the sample fits into this category, these women are likely to have an unmet need for family planning, and are thus of great importance to family planning service providers. 40 Table 4.15 Percent Distribution of Nonpregnant Worsen Who Save Ever Had Sexual Intercourse and Who Are Not Using Contraception by Attitude Toward Becoming Pregnant in the Next Few Weeks, According to Number of Living Children, TTDHS 1987 Attitude Toward Becoming Pregnant Number Does of Living not Not Total Children Happy Unhappy Matter Stated Percent Number 0 55.2 28.2 15.2 1.5 100 330 1 44.2 40.7 13.0 2.2 tOO 231 2 23.9 56.1 18.3 1.7 100 230 3 22.6 63.2 ii.0 3.2 100 155 4+ 7.4 75.2 16.2 1.2 10O 339 Total 31.1 52.1 15.1 1.8 1O0 1,285 Note: Excludes women who have never had sexual intercourse and those women who have not resumed sexual relations since the last birth. Table 4.16 Percent Distribution of Nonpregnant Women Who Are Sexually Active and Who Are Not Using Contraception, and Who Would be Dnha~y if they Became Pregnant by Main Reason for Nonuse, According to Age, TTDHS 1987 ~e All Reason for Nonuse <30 30+ Ages Opposed to family planning 7.7 7.6 7.6 Partner disapproves 3.2 3.3 3.3 Others disapprove 3.2 O.0 1.2 Health concerns 13.8 25.6 21.2 No partner 30.0 23.2 25.7 Postpartum/breastfeeding 6.5 3.8 4.8 Menopausal/subfecund 0.8 14.2 9.3 Other 23.9 16.6 19.3 Don't know 10.9 5.5 7.5 Not stated 8.0 0.2 0.i Total i00.0 i00.0 10O.0 Number 247 422 669 Note: Excludes women who have never had sexual inter- course and those women who have not resumed sexual relations since the last birth. Health concerns are the leading reasons given for nonnse of contraeeptinn by women age 30 and above, followed by lack of partner and menopanse/subfecundity. The leading reason for nonuse given by younger women is that they do not have partners. 4.13 Intention to Use Contraception As shown in Table 4.17, 42 percent of women currently in a union and not currently using any contraceptive intend to use a method in the future, including 28 percent who intend to use in the next year, while 11 percent are unsure, and 46 percent do not intend to use at all. Women with 1 to 3 children are slightly more likely to intend to use a method soon, but the differences are slight. 41 Table 4.17 Percent Distribution of Women in Union Who Are Not Currently Using Any contraceptive Method, by Intention to Use in the Future, According to Number of Living Children, TTDHS 1987 Number of Living Children* All Intention 0 1 2 3 4+ Women Intends to use: 43.3 In next 12 months 18.8 Later 14.4 Doesn't know when 1O.l Unsure about using 15.5 Does not intend to use 41.2 Not stated 0.0 50.5 44.2 43.1 31.5 41,7 34.9 32.4 35.3 23.5 27,9 7.8 5.3 3.0 1.0 6,4 7.8 6.5 4.8 7.0 7.4 12.1 13.0 10.9 6.7 11,5 37.5 42.5 44.9 61.6 46,4 0.8 0.4 1.2 0.3 0,3 Total Percent 100.0 100.0 i00.0 1O0.0 I00.0 1O0,O Number 277 232 247 167 315 1,238 * Current pregnancy is counted as a living child. Table 4.18 shows that among women in union who intend to use methods in the future, the pill is preferred by 35 percent, followed by female sterilization (13 percent) and the IUD (11 percent). Timing of contraceptive use does not appear to play a large part in choosing a method. Table 4.18 Percent Distribution of Women in Union Who Are Not Using a Contraceptive Method but Who Intend to Use in the Future by Preferred Method, According to Whether They Intend to Use in the Next 12 Months or Later, TTDHS 1987 Timing Preferred Next 12 Use All Method Months Later Women Pill 35.0 32.9 34.6 IUD 9.8 17.7 10.6 Injections 9.8 8.9 9.6 Vaginal methods 5.5 1.3 4.7 Condom 6.9 6.3 6.8 Female sterilization 13.3 13.9 13.4 Safe period 1.2 1.3 1.2 Withdrawal 2.3 1.3 2.1 Uther 0.6 0.0 0.5 Doesntt know which method 16.5 16.5 16.5 Total Percent 100.0 100.0 100.0 Number 346 79 425 42 4.14 Exposure to Mass Media and Family Planning Messages The purpose of Table 4.19 is to assess the extent to which respondents are exposed to mass media in general, and to family planning messages in particular. Overall, 94 percent of the sample llve in houses with a radio, while 90 percent have televisions, and 36 percent have videocassette players. Ninety-eight percent of respondents have at least one media source in their households. Nearly all women (97 percent) either read the newspaper at least one a week or watch tv or listen to the radio each day. Exposure to mass media is slightly higher in urban than rural areas, and much higher for more educated women. Despite nearly universal exposure to the mass media, only 55 percent of the respondents were exposed to a family planning message on radio, television, in the newspaper, or on a poster in the month before the survey. Exposure to family planning messages follows the trends for exposure to media in general. Only 37 percent of women with less than complete primary education were exposed to a family planning message compared with 68 percent of women with the highest level of education. Support for running family planning messages on the radio and television is nearly universal among women in Trinidad and Tobago as indicated by Table 4.20. Overall, 94 percent of women believe that the practice is acceptable. Not surprisingly, those least likely to approve of family planning messages on the radio or 'IV are women with less than primary education, and the oldest women. Even among these groups, however, at least 74 percent of respondents approve of the practice. Differences according to other background characteristics are slight. 4.15 Discussion of Family Planning with Partner While discussion of contraception between couples is not necessary for adoption of certain methods, the absence of such conversations may be an impediment to increasing contraceptive prevalence. In addition, the effectiveness of coital-related methods, in particular, can no doubt be improved with increased communication between couples. Overall, only 48 percent of women who know any method have discussed contraception with a partner in the past year (see Table 4.21). Given that knowledge of contraception is nearly universal, and that 83 percent of women in union have ever used methods, this finding is somewhat surprising. Recent discussions about contraception were most common among women under 25 (about 62 percent). The older the woman is, the less likely she is to have discussed the subject with her partner. Less than 30 percent of women in their forties have discussed contraception with a partner in the last year. In general, the lack of discussion may reflect a general reluctance to talk about matters related to sex. While the oldest women might continue to regard family planning as a "taboo" subject not suitable for discussion, they might also have less need to discuss the subject, either because childbearing has ceased due to menopause or sterilization, or because contraceptive practice has become routine. More educated women axe more likely to have discussed the subject recently. Differences according to ethnicity are minimal. 43 ,-4 m o ~ M 'IJ I~ O ~ o J -.4 ~ m ,~ :l ~ -,-4 ,-4 .,4 O ,1 :~ O m O ~ N O ) M m .-4 m -,4 ~ J 0 E ~ O ~ O ~ U l ~ 4 m O 13a > O .,-i m E -t % 0 > > 0 1 3 ~ 0 I 0 0 ~ 0 ~ 0 ~ 0 ~ ~ .* . . . . . . . -- Q ~ ~ Q Q I m ~ l O ~ ~ ~ ~ 0 ~ ~ Q I Q ~ ~ i I e ~ m e J ~ m e ~ .° . . . . . . . . O Q • # • J e • O e l O ~ Q # i ~ Q ~ = ~ ~ = ~ 0 E-~ ~J E 0 0 ~J O -d .~ o m 3 1 -,d m O O ~ O 0 N O M M " -~ -0 0 0 O 0 4 J~ e ~ O E ~ .~ o ,-.i ,~ -I ~ ,.-t ,.~ ~ ~ ~ O O M ~ ~ 0 0 ~ 44 Table 4.20 Percentage of All Women Who Believe That it is Acceptable to Have Messages About Family Planning on the Radio or TV, by Age and Selected Background Characteristics, TTDHS 1987 Current Age Background Charac- All terlstlc 15-19 20-24 25-29 30-34 35-39 40-44 45-49 Ages Residence Urban 92.9 95.2 94.2 95.7 93.8 98.1 91.2 94.5 Rural 92.8 94.7 97.5 94.4 93.9 90.4 83.8 93.6 Education <Complete primary 75.0 90.0 87.2 93.1 88.5 94.4 74.4 86.2 Completed primary 82.8 95.4 97.0 94.4 93.4 92.1 92.7 93.8 Secondary 11 93.8 95.9 96.0 95.8 95.8 96.0 97.2 95.2 Secondary II 2 98.6 91.5 95.9 97.1 10O.0 96.9 90.0 95.7 Ethnicity* African 93.1 97.1 96.6 96.6 94.7 94.4 91.2 95.2 Indian 92.6 94.2 96.2 94.3 94.4 92.9 85.0 93.7 Mixed 92.8 93.4 94.6 93.1 90.3 94.7 84.7 92.4 Total 92.8 94.9 96.0 95.0 93.9 93.9 87.5 94.0 L Some or full secondary education, but fewer than five "o" level exams passed. • Some or full secondary education, with five "o" level exams passed, at least one "A" level, or some university education. * Excludes 27 women of "other" ethnlcity, and one respondent with missing information. Table 4.21 Percentage of Women in a Union and Who Know ~my Method of Contraception Who Have Discussed Family Planning With a Partner at Least Once in the Past Year, by Age and Selected Background Characteristics, TTDHS 1987 Current Age Background Charac- All teristlc 15-19 20-24 25-29 30-34 35-39 40-44 45-49 Ages Residence Urban 61.0 60.4 51.4 46.0 38.5 28.9 16.4 46,4 Rural 62.3 64.5 65.9 54.5 43.3 30.1 15.5 51,5 Education <Complete primary * * 36.7 * 40.0 37.3 17.1 35,6 Completed primary 62.1 58.8 59.2 51.0 38.5 25.2 15.7 44,9 Secondary I* 60.8 63.6 62.1 44.4 44.0 38.1 16.1 54.9 Secondary II z * 66.7 60.0 64.9 48.7 19.2 * 53,1 Ethnicity* African 62.3 65.3 59.0 46.2 45.5 26.9 13.6 49.4 Indian 63.3 62.1 59.3 55.3 40.4 32.1 16.7 49,4 Mixed 58.8 57.0 61.9 46.2 34.4 27.1 16.0 45,7 Total 61.8 62.7 59.4 50.6 41.1 29.6 15.9 48,4 Number 136 445 621 468 384 311 226 2,591 * Fewer than 25 cases. i Home or full secondary education, but fewer than five "O" level exams passed, • Home or full secondary education, with five "O ~ level exams passed, at least one "A" level, or some university education. * Excludes 19 women of "other" ethniclty or with missing information. 45 46 CHAPTER 5 FERTILITY PREFERENCES 5.1 Fertility Preferences One important rationale for the development of the National Family Planning Programme in Trinidad and Tobago was to enable couples to bear the number of children they desire, with the births spaced according to their preferences. The TrDHS collected information on three aspects of fertility preferences which are of importance to family planning policy makers trying to gauge the family planning programme to meet the needs of the population. First, respondents were asked whether births in the five years preceding the survey were planned and timed according to their preferences at the time of the pregnancies. Second, women in unions were asked if they wanted to have another child, and if so, when. Finally, all women were asked to state how many children they would prefer to have if they could live their lives again. Data on fertility preferences are generally subject to more measurement error than objective phenomena such as actual fertility or contraceptive use. For example, a woman may rationalize the birth of a child which was unplanned, and be unwilling to state that a birth was not wanted. And, her ability to implement her preferences might be curtailed if her partner objects to her using contraception to achieve her fertility desires. Such phenomenon may not be captured in a standardized questionnaire. DHS surveys included several innovations to try to overcome these measurement difficulties. For example, respondents were asked the certainty of their stated fertility preferences. While the validity of these follow up questions is unknown, they can serve to remind the analyst of the degree to which answers may or may not actually reflect the views of respondents. Table 5.1 shows the desire for more children among women in union according to the number of living children they have. (Note that pregnant women were asked about their desire for another child after the one that they were expecting; in tabulations, the expected child has been counted as an additional child.) Overall, 47 percent of women in union want no more children, while 38 percent want to have another child, and 5 percent are undecided. In addition, 10 percent have had contraceptive sterilizations or are infecund. Fertility preferences for all women in union are shown in Figure 5.1. Among the 38 percent of women in union who want another child, more than one-half wish to delay the birth for at least two years. Thus, 20 percent of women in union want to delay a birth for at least two years, and can be considered as potentially in need of temporary methods of contraception. The desire to space births is strongest among women wanting to postpone the first or second birth. Thirty percent of childless women want to postpone the first birth, while 46 percent of women with one child want to delay a second birth. Sterilized women (and two women whose partners are sterilized) were asked if they regretted the operation, and if so, ff they would like more children. As shown in Table 5.1, the level of regret is minimal; in subsequent tables in this chapter, sterilized women are counted along with women wanting no more children. The highest proportion of women who want to have a child are those who have no children (90 percent) followed by those who have only one child (76 percent). The desire for additional children drops sharply for women with two or more children. Only 18 percent of women with three children want another child. Conversely, as parity increases, women are more likely to want no more children, or to have been sterilized. More than 54 percent of women with two or more children want to cease childbearing. These women are candidates for permanent methods of contraception such as sterilization. 47 Table 5.1 Percent Distribution of Women in Union by Desire for More Children, Timing of Next Birth, and Sterilization Regret According to Number of Living Children, TTDHS 1987 Nu~er of Living Children ~ Desire For More Women in Children 0 1 2 3 4 5 6+ Union Wants no more 4.8 17.8 54.7 65.9 69.1 70.2 63.7 47.0 Have another: 89.5 75.9 32.0 17.9 10.4 6.4 3.5 38.4 Soon a 53.0 27.1 10.3 6.4 2.6 4.2 1.9 16.3 Later I 30.3 46.2 20.1 10.2 7.5 4.2 1.2 20.1 Undecided when 6.2 2.6 1.6 1.3 0.3 0.0 0.4 2.0 Undecided' 4.0 4.9 7.5 3.8 2.6 4.8 5.1 4.9 Sterilized:' 0.3 0.6 4.8 10.4 15.6 15.5 25.4 8.3 Regret-have another 0.3 0.0 0.3 i.I 1.0 0.6 0.4 0.5 Regret-no more, undecided 0.0 0.0 0.3 0.0 0.0 0.0 0.4 0.1 No regret 0.0 0.0 3.9 8.9 13.3 13.7 23.0 7.1 Infecund 1.4 0.9 1.0 2.0 2.3 1.2 2.3 1.5 Total Percent i00.0 100.0 100.0 I00.0 i00.0 100.0 i00.0 100.0 Nut, bet 353 468 612 451 308 168 257 2,617 * Current pregnancy is counted as a live child. ' Wants next birth within 2 years. * Wants to delay next birth for 2+ years. ' Includes cases missing information on deslre for more children. s Includes cases missing information on sterillzatlo~ regret. Figure 5.1 Fertility Preferences Women in Union 15-49 Want to Space (2 or more yre) 20% : NO More 47% Want Soon (within 2 yre) 16% Sterilized or Infecund 10% Undecided If Want More 5% Vant Another, Undecided When 2% Trinidad & Tobago DHS 1987 48 Respondents appear fairly certain about their fertility preferences. Only 12 percent of women in union responded with uncertainty when asked if they were sure about their decision either to have another or have no more children (no table). Among women whose initial preference for children was uncertain, few expressed a preference in either direction when asked a follow-up question. In tables in this chapter, respondents whose first answer was either to have more or no more children are classified according to the fertility preference initially stated. On the other hand, those few women whose first response was "undecided" are reclassified ff the second response indicated a preference for having more or no more children. Table 5.2 shows that the pattern of fertility preferences by age of the woman closely follows that of the parity-specific pattern seen in Table 5.1. This is consistent with expectations, since age and parity are so closely linked. The proportion of women in union wanting no more children increases from 17 percent of women aged 15-19 to more than 82 percent of women age 40 and above. ALso, the fact that no more than 20 percent of women in any age group want children soon indicates a clear need for reliable methods of contraception. Table 5.2 Percent Distribution of Women in Union by Desire for More Children, According to Age, TTDHS 1987 AGE Desire For More Women in children 15-19 20-24 25-29 30-34 35-39 40-44 45-49 Union Wants no more* 17.3 31.3 44.7 58.6 69.9 84.0 82.5 55.3 Have another: 78.3 64.8 49.0 35.8 21.1 10.6 5.6 38.4 Soon 2 12.9 19.1 20.7 19.4 16.5 6.7 4.7 16.3 Later* 64.7 42.4 26.1 14.5 3.1 0.3 0.0 20.1 Undecided 0.7 3.3 2.2 1.9 1.5 1.6 0.9 2.0 Undecide~ 4.3 3.8 6.1 4.9 6.9 2.6 4.3 4.9 Infecund 0.0 0.0 0.2 0.6 2.1 2.9 7.7 1.5 Total Percent 100.0 i00.0 100.0 I00.0 i00.0 100.0 100.0 100.0 Number 139 450 624 469 389 312 234 2,617 * Includes sterilized women. t Wants next birth within 2 years. * Wants to delay next birth fo~ 2+ years. ( Includes cases missing information on desire for more children. In Table 5.3, the percentage of women in union who want no more children is shown for each parity by selected background characteristics. Overall, rural women are more likely to want to cease childbearing than urban women. However, since differences at each parity level are slight, the overall difference is due to the higher parity of women in rural areas. The same is tree of race, where East Indian women are slightly more likely than African women to want no more children, but the differences at specific parity levels are slight. Fertility preferences by parity are summarized for women in union in Figure 5.2 With regard to education, it is not surprising that the least educated women are the most likely to want no more children, since they have more children than more educated worsen. However, at low parity levels, the diff~'ential in the desire to have no more children persists. Among women with one child, 44 percent of those with less than primary education want no more children, compared with fewer than 19 percent of women with more education. 49 100% Figure 5.2 Fertility Preferences by Parity Women in Union 15-49 80% 80% 40% 20% 0% 0 1 2 3 4 Number of Living Children 5 6* m Want No More * Want Soon Want to Space "* [---] U ndec lded • Inc ludea s ter i l i zed & In facund women ** Inc ludea women who want a ch i ld , unsure when i Trinidad & Tobago DHS 1987 Table 5.3 Percentage of Women in Union Who Want NO More Children by Nu.~er of Living Children and selected Background Characteristics, TTDHS, 1987 Number of Living Children ~ Background Women In Characteristic 0 1 2 3 4+ Union Residence Urban 5.8 19.7 58.7 75.0 84.1 51.4 Rural 4.3 17.1 60.2 77.4 87.8 58.4 Education <Complete primary i0.0 43.8 65.7 80.5 88.0 77.8 completed primary 3.1 17.8 61.2 74.2 86.1 62.4 Secondary I* 4.6 18.3 53.5 78.0 B6.4 43.0 Secondary III 8.6 12.2 71.4 79.4 84.6 45.3 Ethnlclty 4 African 4.0 18.0 56.7 75.2 81.8 50.6 Indian 6.4 17.2 60.3 77.8 90.0 59.2 Mixed 4.1 22.5 61.2 72.5 85.7 53.6 Total 5.1 18.4 59.5 76.3 86.5 55.3 z Current pregnancy counted as a living child. : Some or full secondary education, but fewer than five "0" level exams passed. * SOme or full secondary education, with five "0" level exams passed, at least one "A" level, or some university education. 4 Excludes 19 women of "other" ethniclty or with missing information. 50 5.2 Need for Family Planning Table 5.4 permits the examination of women's need for family planning in order to space or limit future births according to their intention to use contraception. These data can enable the family planning programme to identify population subgroups which have an unmet need for contraception in order to achieve their fertility desires. Also, the right hand side of the table further restricts the numerator to those women who intend to use contraception to achieve their preferences for further children. The differences between the first and second panels for any subgroup of women represents women who will need to be motivated to use family planning to achieve their preferences. Table 5.4 Percentage of Wemen in Union Who Are in Need of Family Planning and the Percentage Who Are in Need and Who Intend to Use Family Planning in the Future by Selected Background Characteristics, TTDHS 1987 Not Contracepting and In- Not Contraceptlng tends to use Contraception Want Want to Want Want to Women Background NO Postpone/ NO Postpone/ In Characteristic More Undecided I Total More Ondecided I Total Union Residence Urban 19.5 10.4 29.9 7.3 4.5 11.8 1,172 Rural 23.2 11.5 34.7 9.7 5.7 14.3 1,445 Education <Complete primary 42.5 8.7 51.2 7.5 2.0 9.5 252 Completed primary 24.1 9.4 33.5 8.5 3.5 12.0 1,171 Secondary I* 16.4 14.4 30.9 %.7 9.0 16.7 939 Secondary II ~ 7.8 8.2 16.0 3.9 5.5 9.4 256 Ethniclty 4 African 20.7 13.1 33.8 8.8 6.4 15.2 942 Indian 22.3 9.3 31.7 6.9 3.8 10.7 1,210 Mixed 20.9 11.4 32.3 9.2 6.3 15.5 446 Total 21.5 11.0 32.5 8.0 5.2 13.2 2,617 * Includes women who are undecided about whether to have another birth or about the timing for the next birth. Some or full secondary education, but fewer than five "0" level exams passed. Some or full secondary education, with five "o" level exams passed, at least one "A" level, or some university education. 4 Excludes 19 women of "other" ethniclty or with missing information. From Table 5.1, recall that 47 percent of women in union want no more children. Nearly one-haLf of these women, or 22 percent of women in union are not using contraception. Similarly, among the 22 percent of women in onion who want to postpone the next birth or are uncertain about having another child, one-half are not using methods, suggesting that 11 percent of women in union are at risk of having a mistimed pregnancy. Again, only about one-half of these women intend to use a method in the future. In sum, Table 5.4 shows that 33 percent of women in union have an unmet need for contraception. Fewer than one-haif of these women intend to use a method in the future. 5.3 Ideal Family Size Respondents were asked to consider an abstract situation independent of their current family size and state the number of children they would choose to have if they could start their reproductive years again. Table 5.5 shows that many respondents (42 percent) would prefer to have 2 children. The mean ideal number of children is higher, 2.9, reflecting the fact that the 51 second most popular number of children is 4. Preferred fertility is slightly lower than actual fertility, 3.1, indicating that on average, women are having more children than they want. Table 5.5 Percent Distribution of All Women by Ideal Number of Children; Mean Ideal Number of Children for All Women and for Women in Union, Percentage of All Women Whose Current Number of Children Exceeds Ideal Number, According to Number of Living Children, TTDHS 1987 Number of Living Children* Ideal Number of All Children O 1 2 3 4 5 6+ Women 0 1 2 3 4 5 6+ Non-numerlc responses Total Percent Number Mean Ideal Number All Women Women in Union Percentage Whose Current Children Exceed Ideal Number (All Women) 3.4 1.5 0.9 1.3 2.1 I.I 2.1 2.1 7.1 7.0 3.6 4.4 3.3 3.3 3.1 5.3 52,2 50,6 43,2 27.8 29.8 27.1 21.6 41.8 19.2 18.7 21.6 26.6 6.9 13.8 10.6 18.5 13.1 18.0 25.2 30.5 39.8 26.0 30.8 22.4 1.7 i.i 2.4 4.0 6.3 18.2 2.4 3.3 1.7 2.4 2.1 4.4 8.7 9.9 24.0 4.9 1.4 0.7 0.9 1.0 3.0 0.6 5.5 1.6 100.0 100.0 100.0 100.0 100.0 I00.0 100.0 100,0 1,321 540 662 478 332 181 292 3,806 2.5 2.6 2.9 3.1 3.5 3.7 4.1 2.9 2.6 2.7 2.9 3.2 3.4 3.6 4.0 3.1 1.5 4.5 33.5 42.1 71.3 70.6 17.7 * Current pregnancy counted as a living child. Among all women in the survey, the mean ideal family size increases from 2.5 children for childless women to 4.1 children desired by women with 6 or more children. The last row of Table 5.5 shows the percemage of all women who have more children than they consider ideal. As expected, this figure increases from 2 percent of women with one child to 34 percent of women with 3 children, and to more than 70 percent of women with more than 4 children. Of the total sample, 18 percent have exceeded their ideal family size. The fertility preferences of women in union are similar. There are several possible masons why women with larger families express a higher ideal family size. First, women with large families may genuinely desire more children than women with smaller families. Secondly, women with more children are likely to be older than women with fewer children. Their ideal family sizes may therefore reflect more traditional views. (This explanation is supported by the ideal family size of 3.8 expressed by women in union in the TTWFS a decade ago). Finally, women may tend to rafionallze the births that they had, and thus express a preference for a larger family size than they otherwise might indicate. It is difficult to distinguish among such factors; however, it is clear that women with four or more children have exceeded their ideal family sizes. As indicated earlier, ideal family size for high parity women may be influenced by a number of factors. Table 5.6 confirms that older women do indeed prefer larger families than younger women. Ideal number of children increases from 2.5 for women 15-19 to 4.0 for women 45 -49. 52 Table 5.6 Mean Ideal Number of Children For All Women by Age and Selected Background Characteristics, TTDBS 1987 current Age Background All Characteristic 15-19 20-24 25-29 30-34 35-39 40-44 45-49 Ages Residence Urban 2.6 2.7 2.8 2.8 3.0 3.3 3.8 2.9 Rural 2.4 2.6 2.8 2.9 3.3 3.6 4.1 2.9 Education <Complete primary * 2.9 2.9 3.2 3.8 4.0 4.0 3.6 completed primary 2.4 2.9 2.8 2.9 3.2 3.5 4.1 3.I Secondary I a 2.5 2.6 2.8 2.8 2.8 3.3 3.8 2.7 Secondary If* 2.5 2.6 2.6 2.7 3.4 2.6 3.5 2.7 Ethnlclty ~ African 2.6 2.8 2.8 2.9 3.0 3.7 4.0 3.0 Indian 2.3 2.5 2.8 2.9 3.3 3.5 3.9 2.9 Mixed 2.6 2.7 2.8 2.7 3.0 3.0 4.1 2.9 Total 2.5 2.7 2.8 2.9 3.2 3.5 4.0 2.9 * Fewer than 25 cases. Some or full secondary education, but fewer than five "o" level exams passed. ' Some or full secondary education, with five "o'; level exams passed, at least one "A" level, or some university education. * Excludes 27 women of "other" ethnlclty, and one respondent with missing information. Table 5.6 also shows differentials in ideal family size according to certain background characteristics. There is little difference according to residence or ethnicity, but educational background is quite important. Ideal family size fails from 3.6 among women with less than complete primary education to 2.7 for women who have attended secondary school. Some of the differences in ideal family size by education level are due to the fact that women with low levels of education tend to be older and of higher parity. At any level of education ideal family size increases with age. 5.4 Fertility Planning Status Table 5.7 presents information on whether births in the last five years were planned, wanted later, or not wanted at all. While women may have a tendency to rationalize unplanned births, results from previous fertility surveys demonstrate that women are indeed willing to admit unwanted births. Mistimed or unplanned pregnancies, however, should still be considered approximations. A/together, 63 percent of births in the past five years were wanted then, while 20 percent were wanted later, and 16 percent were unwanted. The distribution by birth status is similar among women who did and did not use contraception. While one might expect women who used contraception in a given interval to have had only planned births, Table 5.7 shows that this was not the case. About one-third of births to nonusers were unplanned, compared with 39 percent of births to contraceptive users. Table 5.8 presents a summary of the information in Table 5.7, restricted to births in the past 12 months. Overall, 42 percent of recent births were either mistimed or unwanted. First and second order births were much more likely to have been planned (70 percent) than third order or higher (41 percent). Conversely, only 3 percent of lower order births were unwanted, compared with 40 percent of higher order births. 53 Table 5.7 Percent Distribution of All Births in the Last Five Years by Contraceptive Practice and Fertility Planning Status, According to Birth Order, TTDHS 1987 Contraceptive Birth Order* Practice and Planning All Status 1 2 3 4+ Births Non-Contraceptlve Interval 59.7 35.8 38.5 46.0 46.1 Wanted then 47.9 25,8 23.2 23.4 31.1 Wanted later 10.B 8.0 10.9 7.6 9.2 Not wanted 1.0 2.0 4.4 15.0 5.8 Contraceptive Interval 39.2 62.0 61.4 53.3 52.7 Wanted then 30.1 43.6 34.9 21.9 32.0 Wanted later 7.8 14.1 14.2 8.3 10.6 Not wanted 1.3 4.3 12.3 23.1 10.1 Not Stated 1.1 2.2 0.3 0.7 1.1 Total Percent 100.0 i00.0 10O.0 I00.0 100.0 Number of Births 618 539 367 606 2,130 Note: Includes births in the period 0-59 months prior to the survey. * Current pregnancy counted as a living child. The information collected on whether or not births in the last five years were wanted permits calculation of a total wanted fertility rate. The calculation is identical to that used for calculating the TFR for the five years before the survey (see Chapter 2), except that births which were unwanted are excluded from the calculation. Table 5.9 shows that if all unwanted births were prevented, the TI,'R would decline from 3.1 to 2.6. In other words, if current trends continue, each 100 women will have 50 unwanted births during their lives. Differentials in wanted fertility according to the woman's background characteristics are similar to those seen for the TFR as a whole. The third column of Table 5.9 shows the percentage of the 'IFR comprised of unwanted births. It is interesting to note that this figure drops from 24 percent of women with less than full primary education to I0 percent of women with full secondary certification. 54 Table 5.8 Percent Distribution of Births in the Year Before the Survey by Fertility Planning Status, According to Birth Order, TTDHS 1987 Birth order* All Planning Status I-2 3+ Births Wanted child then 70.0 41.2 57.6 Wanted child later 26.5 19.0 23.3 Wanted no more children 2.9 39.9 18.8 Not classified 0.5 0.0 0.3 Total Percent i00.0 10O.0 100.8 NunLber 407 306 713 Note: The number of women with a birth in the past 12 months is roughly equivalent to the nun~ber of births in the past 12 months. Thus, the per- centage who want no more children is equivalent to the percentage of unwanted births. * current pregnancy counted as a living child. Table 5.9 Total Wanted Fertility Rate, Total Fertility Rate for the Five Years Preceding the Survey, and Percentage of the Total Fertility Bate Identified as Unwanted Births by Selected Background Characteristics, TTDHS 1987 Total Percentage of Total Wanted Total Fertility Rate Background Fertility Fertility Identified as Characteristic Rate Rate Unwanted Births Residence Urban 2.5 3.0 15.9 Rural 2.6 3.2 20.0 Education <Complete primary 3.0 4.0 23.7 Completed primary 2.9 3.6 20.1 Secondary I ~ 2.6 3.1 16.3 Secondary II a 2.1 2.3 9.9 Ethnicity ~ African 2.8 3.5 19.8 Indian 2.3 2.8 15.8 Mixed 2.7 3.4 21.6 Total 2.6 3.1 18.5 Some or full secondary education, but fewer than five "o" level exams passed. • Some or full secondary education, with five "o" level exams passed, at least one "A" level, or some university education. i Excludes 27 women of "other" ethnicity, and one respondent with missing information. 55 56 CHAPTER 6 INFANT AND CHILD MORTALITY, AND HEALTH The incidence of mortality during the first year of life reflects the socioeconomic status of the population, and is particularly sensitive to changes in environmental and social conditions. This chapter discusses infant and childhood mortality, and other indicators of child health, including prenatal care, immunization coverage, diarrhoea, and nutritional status. This section summarizes the main health f'mdings in the survey. Further analysis may elucidate some of the unexpected findings presented below. 6.1 Infant and Childhood Mortality Table 6.1 presents infant and childhood mortality rates for three recent time periods, 1972-1976, 1977-1981 and 1982-1987. The most recent figure includes exposure for the few months in 1987 prior to the month of interview. In the last five years, 26 out of every 1,000 babies died before reaching the first birthday, while 3 per 1,000 died between the first and fifth birthdays. This represents a very low level of mortality, approaching that of developed countries. By comparison, the infant mortality rate (IMR) in the Dominican Republic and Colombia were 68 and 33 respectively for the period 1981-1986 (CONAPOFA and IRD/Westinghouse 1987; CCRP et. al. 1987). Table 6.1 Infant and Childhood Mortality for Five-Year Calendar Periods, TTDHS 1987 Infant childhood Both Period (lq0) (4ql) (5q0) 1982-1987" 26.2 3.4 29.5 1977-1981 37.3 3.5 40.6 1972-1976 46.7 7.1 53.4 * Includes exposure up to the month prior to month of interview. The decline of the IMR from 47 to 26 deaths per 1,000 births between the early 1970s and the mid-1980s represents a 44 percent drop. An even greater decline of 52 percent is seen for childhood mortality, which decreased from 7.1 to 3.4. Mortality figures calculated from TIDHS data exceed those published from vital statistics data, as shown in Figure 6.1. Vital statistics data show a decline in the IMR in the 1970s from 28 to 19 deaths per 1,000 births, and a further decline of 5 points by 1984 (Central Statistical Office 1987c, 1987a). Babies dying in the first few days of life may not be registered with vital statistics, but axe likely to have been picked up by TTDHS interviewers who were trained to probe for such events. The IMR computed from WFS data for the 1970-1974 period was 45 per 1,000 births (Ebanks 1985). This is quite comparable to the DHS figure for a similar period. 6.2 Infant and Childhood Mortality by Socioeconomic Characteristics Infant and childhood mortality rates for the 10-year period prior to the survey are presented in Table 6.2 and Figure 6.2. Ten-year rates are used to permit comparisons by background characteristics and to reduce sampling errors. Nevertheless, because the number of children dying is low, caution is advised when viewing the figures. The data show that both infant and child mortality are lower in rural than urban areas. This somewhat unexpected finding may reflect the homogeneity of the society, and the difficuky in distinguishing urban from rural areas. Surprisingly, infant mortality appears highest among the best educated women; note, however, that rates for the highest and lowest education groups are based on a small number of births. As expected, mortality for children aged 1-4 drops as the mother's education increases. 57 Figure 6.1 Infant Mortality Rates, TTDHS, WFS, Vital Statistics Rate per 1,000 births 80 70 60 50 40 30 20 10 0 1872-1977-1982- 1970-1975- 76 81 87 74 78 1972 1974 1978 1878 1980 1882 1984 Trinidad &Tobago DHS 1987 Table 6.2 Infant and Chi ldhood Morta l i ty 1977-1987 by Selected Background Character is t ics of Mother, TTDHS 1987 1977-1987 Background Infant ch i ldhood Both character is t ic (lq0) (4ql) (Sq0) Res idence Urban 36.3 4.9 40.9 Rural 27.5 2.4 29.9 Educat ion <Complete pr imary (27.5) (5.3) (32.6) Completed pr imary 24.7 4.3 26.9 secondary I* 34.8 2.0 36.8 Secondary II ~ (61.0) (0.0) (61.0) Total 31.1 3.4 34.4 Note: Numbers in parentheses indicate a rate based on fewer than 500 exposed persons. * Some or full secondary education, but fewer than five "o" level exams passed. 2 Some or full secondary education, with f ive "o" level exams passed, at least one "A" level, or some univers i ty education. 58 Figure 6.2 Trends in Infant and Child Mortality 60- 4o ! 20- lO- II 7 4 3 0 Infant Mortality Child Mortality Rate per 1,000 mE 1972-76 k~ 1977-81 ~'~ 1982-87 / Under 5 Mortality Trinidad & Tobago DHS 1987 6.3 Infant and Childhood Mortality by Demographic Characteristics Contrary to expectations, survey data show that male infant mortality is slightly lower than female infant mortality (see Table 6.3). However, the usual pattern is seen when the mother's age is considered. Babies born to women under 20 years of age and over 34 years of age are more likely to die before they attain their first birthday than those born to women between the ages of 20 and 34. Birth order also affects the chances of survival of the infant. A typical J-shaped pattern is seen whereby a first birth is slightly more likely to die in childhood or infancy than is a second or third order birth, but mortality increases for higher order births. For seventh and higher order babies, the IMR is almost three times as high as for the second and third order baby. Short birth intervals are strongly linked with increased mortality. Of 1,000 babies born within 2 years of the previous birth, 41 died, as opposed to 15 deaths per thousand births with 2-3 year birth intervals. These findings suggest that further child survival gains could be made if women delayed the first birth until age 20 or above, spaced births for at least 2 years, and ceased childbearing at lower parity levels. 6.4 Children Ever Born and Surviving A further indication of the generally low level of infant and child mortality is seen in Table 6.4 The difference between the mean number of children ever born (2.06), and those surviving at the time of the interview (1.96) is 0.10 child per woman, which represents 5 percent of all ch'fldren born to respondents. The probability of having a dead child increases ten-fold, from 3 per 100 for women aged 20-24 to 32 per 100 for women aged 45-49. This reflects several factors. Older women have more children to begin with, and bore children during periods of higher mortality. In addition, their children are older, and have had more exposure to the risk of death during the course of their lives. Keeping the number of births constant, about twice as many children whose mothers were 35 and over have died, as whose mothers were under 35. 59 Table 6.3 Infant and childhood Mortality 1977-1987 by Selected Background Characteristics, TTDHS 1987 1977-1987 Background Infant Childhood Both Characteristic (lq0) (4ql) (5q0) Sex Male 28.8 3.4 32.1 Female 33.5 3.4 36.8 Age of Mother <20 42.9 6.0 48.7 20-29 28.4 2.3 30.6 30-34 (24.4} (2.1) (26.5) 35+ (37.3) (8.2) (45.3) Birth order i 28.9 3.7 32.5 2-3 25.6 2.1 27.6 4-6 33.0 4.0 36.9 7+ (72.6) (8.7) (00.7) Previous Birth Interval <2 years 41.4 3.0 44.3 2-3 years 14.6 3.6 18.2 4 years or more 25.9 (1.7) (27.6) Note: Numbers i n parentheses indicate a rate based on fewer than 500 exposed persons. Table 6.4 Mean Number of Children Ever Born, Surviving, and Dead, and Proportion of Children Dead Among Children Ever Born, by Age of Mother, TTDHS 1987 Mean Number of Children Proportion Dead Among Children Ever Born Surviving Dead Ever Born Age 15-19 0,14 0,13 O.OO 0.03 20-24 0.89 0.85 0.03 0.04 25-29 1.86 1.80 0.06 0.03 30-34 2.69 2.60 0.08 0.03 35-39 3.24 3.05 0.18 0.06 40-44 3.87 3.64 0.23 0.06 45-49 4.95 4.63 0.32 0.08 Total 2.06 1.96 0.10 0.05 6.5 Antenatal Care The importance of receiving prenatal care during pregnancy is well known, and has been emphasized by the government's health programme in an effort to reduce infant mortality. The timing of this care is also important, but is not dealt with in the survey. Table 6.5 shows that only 1.4 percent of births in the five years before the survey did not receive antenatal ease. In most eases, care was given by doctors (84 percent), while trained nurses or midwives provided care in the remainder of cases. Differences by background characteristics of the mother are slight, except 60 that older women, and those with some secondary education are more likely to receive care from doctors, while younger women, and those with less education are more likely to see trained nurses. It is probable that education is correlated with income, which is likely to determine the type of health care pmvider the woman sees. A completed series of tetanus injecdons offers protection against neonatal tetanus for many years. In keeping with international methods for assessing compliance with child survival goals, and to provide comparability with other DHS surveys, women were asked whether they received at least one tetanus injection while pregnant. Table 6.5 shows that, overall, 31 percent of births in the last 5 years were immunized with at least one dose of tetanus toxoid during pregnancy. Children of the least educated women showed the highest tetanus toxoid coverage rate (43 percent). Table 6.5 Percent Distribution of Births in the Last 5 Years by Type of Prenatal Care for the Mother and Percentage of Births Whose Mother Received a Tetanus Toxoid Injection, According to Selected Background Characteristics of Mother, TTDHS 1987 Type of Prenatal Care Percent Receiving Tetanus Trained Total Toxoid Number No Nurse/ Mis- Per- InJec- of One Doctor Midwife sing cent tion Births Age 15-49 4.4 78.0 17.6 0.0 i00 34.1 91 20-24 1.6 80.4 17.1 1.0 100 26.4 516 25-29 1.1 85.0 13.3 0.6 I00 35.5 660 30-34 0.8 86.3 11.6 1.3 100 28.6 388 35-39 1.5 85.2 11.7 1.5 I00 32.1 196 40-44 1.5 92.4 3.0 3.0 180 25.8 66 45-49 * * * * * * 12 Residence Urban 2.3 85.5 ii,i i.i i00 28.9 826 Rural 0.7 82.9 15.5 0.9 i00 32.3 1,103 Education <Complete primary 0.0 80.6 17.9 1.5 I00 42.5 134 Completed primary 1.8 80.8 17.0 0.6 i00 30.8 855 Secondary I* 1.5 85.7 11.4 1.4 I00 30.4 782 Secondary II: 0.6 95.6 3.2 0.6 I00 23.4 158 Ethnicity a African 1.9 84.2 12.7 1.2 100 30.5 748 Indian 0.8 82.9 15.5 0.7 I00 28.4 831 Mixed 1.8 85.5 11.5 1.2 I00 38.5 338 Total 1.4 84.0 13.6 1.0 100 30.8 1,929 Note: Includes births in the period 1-59 months prior to the survey. * Fewer than 25 births in the age group. * some or full secondary education, but fewer than five "o" level exams passed. * Some Or full secondary education, with five "o" level exams passed, at least one "A" level, or some university education. : Excludes 12 children of "other" ethnlcity or with missing information. 6.6 Assistance at Delivery Assistance at delivery by either a doctor or trained nurse is almost universal--30 percent of births in the 5 years preceding the survey were delivered by doctors, while .68 percent were delivered by trained nurses (see Table 6.6), Women in urban areas, and those wzth full secondary certification were more likely than others to use doctors. 61 Table 6.6 Percent Distribution of Births in the Last 5 Years by Type of Assistance During Delivery, According to Selected Background characteristics of Mother, TTDES 1987 Type of Assistance at Delivery Trained Number Background No Nuree/ Total of Characteristic One Doctor Midwife Other Missing Percent Births Age 15-19 0.0 25.3 73.6 I.i 0.0 I00 91 20-24 0.0 23.8 74.2 1.0 1.0 100 516 25-29 0.2 29.7 68.0 1.4 0.8 I00 660 30-34 0.3 33.5 63.4 1.6 1.3 100 388 35-39 0.5 38.8 58.7 0.5 1.5 i00 196 40-44 0.0 24.2 72.7 1.5 1.5 100 66 45-49 * * * * * i00 12 Residence Urban 0.0 35.7 62.6 0.6 1.1 100 826 Rural 0.3 25.0 72.2 1.6 0.9 100 1,103 Education <complete primary 0.0 27.6 67.9 2.9 1.5 100 134 Completed primary 0.2 25.8 71.7 1.6 0.6 100 855 Secondary I* 0.i 28.1 69.7 0.7 1.4 I00 782 Secondary II ~ 0.0 58.9 40.5 0.0 0.6 I00 158 Ethniclty* African 0.4 27.1 70.2 1.2 1.1 100 748 Indian 0.0 29.0 68.7 1.6 0.7 I00 831 Mixed 0.0 34.9 63.3 0.3 1.5 i00 338 Total 0.2 29.6 6g.1 1.3 1.0 100 1,929 Note: Includes births in the period 1-59 monthe prior to the survey. * Fewer than 25 births in the age group. * Some or full secondary education, but fewer than five rot level exams passed. • Some or full secondary education, with flwe "O r level exams passed, at least one "A" level, or some university education. ' Excludes 12 children of tother r ethnlclty or wlth mlsslng Informatlon. As Table 6.7 shows, nearly 90 percent of births in the last 5 years occorred in government hospitals, while 7 percent took place in private hospitals or nursing homes. Older women, urban dwellers, and those most highly educated were more likely than average to have delivered in private facilities. 6.7 Immunization Immunization of young children is essential for maintaining health and improving their chances of survival. One goal of the Ministry of Health is that by 1990, 85 percent of children one year of age will be completely immunized against diphtheria, permssis, tetanus (DPT), and polio, and 80 percent of children 1-2 years will be immunized against measles/rubella. ~ Women with children born within five years of the survey interview were asked whether they had health records showing immunizations given to their children. If the woman could show the card, interviewers recorded the dates when polio, measles, yellow fever, and the combined DPT shots were given. Tuberculosis is managed on a case by case basis, so the BCG vaccine, requix~ in many countries, is not routinely administered. 62 Table 6.7 Percent Distribution of Births in the Last 5 Years by Place of Delivery, According to Selected Background Characteristics of Mother, TTDHS 1987 Place of Delivery PriVate Govern- Hospital/ Background ment Nursing characteristic Hospital Home Other Total Number Age 15-19 96.7 2.2 i.i i00 91 20-24 94.0 2.7 3.4 10O 516 25-29 87.7 8.8 4.3 100 660 30-34 86.3 8.0 5.7 I00 388 35-39 83.7 12.2 4.0 100 196 40-44 84.8 10.6 4.5 i00 66 45-49 * * * i0~ 12 Residence Urban 86.9 9.3 3.8 100 826 Rural 90.4 5.3 4.4 I00 1,103 Education <Complete primary 94.8 0.7 4.5 I00 134 Completed primary 92.0 4.0 4.0 100 855 S~conda~y I* 89.0 7.3 3.7 i00 782 Secondary II' 66.5 27.2 6.3 i08 158 Ethnlclty* African 92.5 3.7 3.8 i00 748 Indian 87.1 8.4 4.4 1O0 831 Mixed 87.3 8.6 4.2 1O0 338 Total 88.9 7.0 4.0 i00 1,929 Note: Includes births in the period 1-59 months prior to the survey. * Fewer than 25 births in the age group. i Some or full secondary education, but fewer than five "0" level exams passed. • Some or full secondary education, with five "o" level exams passed, at least one "A" level, or some university education. * Excludes 12 children of "other" ethnl01ty or with missing information. If the woman could not produce the card or did not have one, the interviewer asked the respondent whether the child received each of the vaccinations. The survey f'mdings on immunization are presented in Tables 6.8 and 6.9. Note that the top section of both tables includes all children born less than 60 months before the survey. Since some injections are not scheduled to be given until late in the first year of life, subsequent panels in Tables 6.8 and 6.9 refer to children 12-59 months of age only. Overall, health cards were seen by the interviewer for 75 percent of the children. Of the remainder, more than two-thirds had health cards but the interviewers were not able to see them (no table). (It should be noted that women who take their children to private doctors may not have been given health cards to take home.) While cards were seen for only 28 percent of children under 6 months, children 6-11 months were the most likely to have cards--81 percent. The proportion with cards is 69 percent for the oldest children. The variation may be due to recent efforts by the MOH to promote immunizations, or to the greater likelihood that cards of older children have become lost. The chance of a child having a card increases with the mother's education, from 69 percent for children aged 12-59 months with less than primary education to 78 percent for women with less than full secondary certification. The proportion of children with cards then declines for women with full secondary certification, possibly because these women visit private physicians. 63 African women, and women in rural areas were slightly more likely to have produced cards for their children. Three doses of diphtheria-tetanus and polio vaccines are required for children entering public school. Pertussis is not required, but is commonly given as part of a combined DPT shot. Until recently, a yellow fever immunization was required for travel to many countries, and was recorded on health cards, if given. Table 6.8 shows the proportion of children with health cards who received various vaccines. Nearly all children 12-59 months with cards received the first dose of DPT and polio. Coverage decreases for subsequent doses in the series, such that only 87 percent of children received the third dose of DPT or polio. Older children, and those whose mothers were better educated, were more likely to have been immunized. Coverage for other diseases is lower. Seventy-two percent of children 12-59 months were immunized for yeUow fever, and 44 percent for measles. Only 36 percent of children in this age group were fully immunized against yellow fever, DPT, polio, and measles. Table 6.9 shows the proportion of children without health cards whose mothers reported that they received specific immunizations. Levels of immunization coverage are lower for children without cards than for children with cards, but differences by injection and background characteristic are in the expected directions. For example, while more than 80 percent of children 12-59 months received each of the fwst two DPT and polio vaccines, only 69 percent received the third DPT dose, and 68 percent the third polio dose. Fifty-four percent received yellow fever, while 44 percent had a measles vaccine. Only 29 percent received all of these shots, according to the mother's report. While the reliability of information reported by the mother is not known, it is interesting how closely the mothers' reports correspond with information on the cards for the population as a whole. Figure 6.3 shows coverage of selected immunizations according to either the card or the mother's report. 6.8 Diarrhoea Prevalence Diarrhoea is a leading cause of infant and child morbidity and mortality. It is particularly likely to occur during the rainy season, between May and July, most of which coincided with fieldwork. Diarrhoea is also commonly seen in children of weaning age. Mothers were asked whether their children under age 5 had diarrhoea, defined as three or more loose or runny stools per day, in the 24 hours and 2 weeks prior to the survey. Table 6.10 shows that diarrhoea prevalence is quite low, which is consistent with the generally favorable infant mortality situation in Trinidad and Tobago. Only 6 percent of the children had an episode of diarrhoea during the 2 weeks prior to interview, and 2 percent had an episode 24 hours prior to interview. Diarrhoea was most common among children 6-23 months, which is the time weaning occurs, as discussed in Chapter 2. It is interesting to note that mothers whose educational level is highest had a higher proportion of their children with diarrhoea than mothers whose educational level was lower, which is consistent with the breastfeeding durations discussed in Chapter 2. 64 U E-~ -,-4 ~ - o ~ M 4 ~ ,--i ~ .1 O * m 0 ~ M o~ Z o ~ e 4 ~ r ~ ~ 0 .-i o ~ ° i ~ M ~ J ,.i ,.i C,J o ~ B t~ o ~ oB ~ ~ o o .g 9 ~ , ~ ~o~ . . . . . r.-~0~ ,-.- t.~ ~ r~ o t~ tN 4 g H d ~ o , ~ N • , • • • , , , ~ o .,~ ~ '" " ~ ~ o .~ = ,.o o o ~ o ~ o ~ o ~ o ~ ~ ~ ~ ~ ~ ~ ~ '~ t ~ 0 ~ -~ ,-~ C O ~ U ~ (~ ~ ~ ,~ O r-. C ) ~ ~ ~ ~ I " • ~*~ ~ ~ .~ 9 ~ ~ . ~ . L ~ 0 ~ ¢~ ~ F~ " ' ~ ~ ~ F~ " t" U ~ ¢ ~ 0 ~ ~ .~ t~ . ¢ ~ ~ ~ - t~ ~ x~ ,~ ~ t ~ - o ~ c ~ k 9 o Q ¢ ~ ~ N 9 ~ , ~ , ~ '9. ~' 9 '9. o ' = = , ~ ~ " " ' ' '' ~ . ® ~ ~ . ~ ~.o r ~ - r" - u~ ~ r ~ ° ~ ' " ' ' ' ' i ~ - r ~ - ~i~ " ~ ~ ~'~ ~ ,99 9~ . ~ , 0 m = ~u E ~ r~ o ,~ : ? 0 0 X 6 5 U ~ 0 W ~ O O U U ,~ O ~ O M • O ~ CI N I g = O O • U ~ N O ~ U O ~ J ,-4 @ 0 . = : : 4 ~ 3= t o ,*4 O 1 3 0 O ~ ~ ° I I e l .-t I21 .*4 ~ O m ~ 0 ~ 0 O ~ ~ U U ~ 0 0 ~ ~ ~ 0 ~ 0 ~ @ I m o m l ~ 0 0 l @ ~ I m l @ @ @ I ~ l 0 0 ~ 0 I Q @ I J I I @ @ I • l • Q J t m • I . . . . . % 0 ~ 1 1 1 1 o Q • I Q ~ ~ ~ O 0 m I Q I ~ 0 o ~ o ~ ~ r ~ D r ~ 0 E U ~ 0 0 0 ~ . % H l - - I 0 ® ~ ~ 0 E3 o 0 W 0 kD :*4 ~ IJ O ~ ~ ' ~ O ~ o co u o~ g m ~ ' 0 0 w o e ~ q D ' o o o ~ u ' ~ O ~ 66 100 90 80 70 60 50 40 30 20 10 0 Figure 6.3 Immunization Coverage Children 1-5 Percent Cov~ag e 96 Yellow OPT DPT Polio Polio Meaeles All Fever 1 3 1 3 ~1 No Health Cards* ~ Health Cards** i I • 383 Children without Health Cards, 24.9% of sample • - 1153 Children with Health Cards, 75.1% of sample Trinidad & Tobago DHS 1987 6.9 Diarrhoea Treatment Oral rehydration therapy (ORT), the recommended treatment for diarrhoea, is commonly used in Trinidad and Tobago. The solution can be made at home using sugar, salt, and water, or by mixing water with commercially prepared packets of oral rehydration salts (ORS). Table 6.11 shows how recent episodes of diarrhoea were treated. Since the number of children who contracted diarrhoea is small, little analysis by background characteristic is attempted; however, the overall picture is instructive. One-half of the children who contracted diarrhoea consulted a medical facility for treatment. Sixty-six percent of those who contracted diarrhoea were treated with ORT, including 53 percent who were given solution prepared from ORS packets, and 13 percent who were given a homemade solution. Forty-five percent received some other treatment such as tablets, syrups, or a change in feeding, while 20 percent had no treatment at all. 6.10 Knowledge of ORT Knowledge and use of ORT may have had a major influence in the reduction of deaths due to gastroenteritis in recent years. It is simple and inexpensive to prepare and use. Knowledge of ORT among mothers of children 1-59 months of age is quite high, as shown in Table 6.12. Eighty-nine percent were aware of the method. Better educated women were somewhat more likely to know about the method than less educated women. 6.11 Nutritional Status of Children Amhropometry is a widely-used tool for assessing the nutritional status of children. Because young children grow rapidly, inadequate nutrition can be detected in a short period of time. For the present survey, children 3-36 months (born to women interviewed for the TTDHS survey) had their weight and length measured. The children were weighed on 25 kg hanging scales, and measured with portable measuring boards. Trained personnel accompanying each team of interviewers conducted the measuring. 67 Table 6.10 2unong Children Under 5 Years of Age, the Percentage Reported by the Mother to Have Had Diarrhoea in the Past 24 Hours and the Past Two Weeks, Accord ing to selected Background Characteristics, TTDHS 1987 Percentage of Children Under 5 Reported by the Mother as Having Diarrhoea in Background Past 24 Past 2 Number of Characteristic Hours Weeks* Children Age Under 6 months 1.6 5.5 127 6-11 months 3.9 7.2 207 12-23 months 3.2 11.3 380 24-35 months 1.6 5.5 379 36-47 months 1.8 4.5 397 48-59 months I.i 2.4 380 Residence Urban 1.5 6.4 799 Rural 2.5 5.8 1,071 Education <complete pr imary 0.9 4.6 131 Completed primary 1.8 6.1 830 Secondary It 1.7 5.4 757 Secondary If' 6.6 9.9 152 Ethnlc i ty 4 African 1.7 6.0 719 Indian 2.2 6.0 806 Mixed 2.4 6.3 333 Total 2.1 6.0 1,870 Note: Includes children aged 1-59 months. * Includes 24 hour period. * Some or full secondary education, but fewer than five "o" level exams passed. 3 Some or full secondary education, with five "o" level exams passed, at least one "A" level, or some university education. 4 Excludes 12 children of "other" ethnlcity or with missing information. Collection of accurate anthropometric data depends on several factors. Proper training of measurers is of primary importance. "ITDHS measurers were taught to measure children to within the degree of accuracy recommended by the United Nations in the guide "How to Weigh and Measure Children" (United Nations 1986). Details of the training of measurers appears in the Appendix. A second factor affecting the quality of data is the coverage. The 843 children measured represent only 79 percent of the eligible children, so some caution is warranted in interpretation of the anthropomeuic data. Children were missed because they were away from the household, ill, sleeping, or the mother refused. One case was excluded from the tabulations because the measurements were so far out of range for the child's age that they were most likely taken or recorded incorrectly. Thus, tabulations are presented for 842 children. The small sample size limits the disaggregation of data to a few broad categories. Since previous nutrition surveys suggested that localized pockets of undemutrition exist, it is unfortunate that more detailed analysis cannot be carded out. 68 Table 6.11 Among Children Under 5 Years of Age Who Had Diarrhoea in the Past TWO Weeks, the Percentage Consulting a Medical Facility, and the Percentage Receiving Different Treatments as Reported by the Mother, According to Selected Background Characteristics, TTDHS 1987 Percentage of Children with Diarrhoea Treated by I Percentage of Children some with Diarrhoea Solution Number Consulting a of Sugar, Other No of Child- Background Medical ORS Salt, Treat- Treat- ten with Characteristic Facility Packets Water ment ~ ment Diarrhoea Age <23 months 41.5 50.8 15.4 41.5 24.6 65 24-59 months 60.4 56.3 10.4 50.0 14.6 48 Sex Boy 41.7 55.0 15.0 45.0 25.0 60 Girl 58.5 50.9 11.3 45.3 15.1 53 Residence urban 54.9 54.9 15.7 45.1 19.6 51 Rural 45.2 51.6 11.3 45.2 21.0 62 Education <secondary 52.6 49.1 14.0 52.6 17.5 57 Any secondary 46.4 57.1 12.5 37.5 23.2 56 Ethniclty' African 55.8 55.8 18.6 48.8 16.3 43 Indian 47.9 56.3 12.5 37.5 25.0 48 Mixed * * * * * (21) Total 49.6 53.1 13.3 45.1 20.4 113 Note: Includes children aged 1-59 months. * Fewer than 25 cases. Multiple responses were accepted so that percentages may not add to i00. l Includes tablets, injections, syrups, and change in diet (increasing or decreasing food or fluids). * Excludes I child of "other" ethnlcity or with missing information. The validity of anthropometric data also depends on the accuracy of children's reported ages. If a child's true age is just a few months younger than the age reported by the mother, he or she could be reported erroneously as being severely malnourished. TTDHS interviewers were thoroughly instructed in the collection of accurate age data, and performed several checks in the field to verify age information. In fact, no children measured were missing information on the month or year of birth, suggesting that mothers have good recall of their childrens' ages. Figure 6.4 shows the distribution of all children, and of children measured, by age in months. If mothers were estimating ages, one would see heaping at months 12, 18, 24, 30, and 36 for the line corresponding to "all living children". The presence of only minimal heaping suggests that TTDHS anthropometric data are not biased by misreported ages. In addition, the figure shows that children younger than 12 months were slightly more likely to have been measured than were children 13-36 months. In order to facilitate comparisons with DHS surveys done in other countries, and with other nutrition surveys done in Trinidad and Tobago, the nutritional status data in the TTDHS was analysed using the National Center for Health Statistics/Centers for Disease Control (NCHS/CDC) International Reference Population, as recommended by the World Health Organization (U.S. Department of Health, Education and Welfare 1976). Use of the reference data for comparative purposes is based on the finding that ethnic differences are far less important than environmental conditions in determining the growth of pre-school aged children (Martorell and Habicht 1986). 69 Table 6.12 Among Mothers of Children Under 5 Years of Age, the Percentage Who know ~bout ORT by Education, According to Selected Background Characteristics, TTDHS 1987 <Complete Completed Background Primary Primary Secondary I* Secondary II * Total Residence urban 70.0 88.6 91.5 90.9 89.3 Rural 89.3 87.6 87.3 95.3 88.1 Ethnlclty' African * 89.1 91.0 97.8 90.7 Indian 82.1 87.5 85.9 92.7 86.7 Mlxed * 86.8 93.3 * 89.1 Total 82.6 88.0 89.4 92.7 88.6 Note: Includes children aged 1-59 months. * Fewer than 25 cases. x Some or full secondary education, but fewer than flve "0" level exams passed. = Some or full secondary education, with five "o" level exams passed, at least one "A" level, or some university education. ' Excludes 8 women of "other" ethniclty or wlth mlsslng information. Figure 6.4 Age Distribution of Weighed and Measured Children, and All Children Frequency 5O 40 30 20 10 0 0 6 12 18 24 30 Age in Months 36 Measured children ' All living children Trinidad & Tobago DHS 1987 Four standard indices are presented below to describe the measured children: o Height-for-age o Weight-for-height o Weight-for-age o Height-for-age by weight-for-height 70 Tables 6.13A-6.16A show the percentage of children falling into various standard deviation (or z-score) categories from the reference population medians on the above four indices. Since nutritional status is often expressed as percent of the median reference scores in Trinidad and Tobago, Tables 6.13B-6.16B present the four indices using this indicator. The variation in height and weight among children at any given age approximates a normal distribution around the median. Thus, 68.2 percent of well-nourished children fall within one standard deviation (SD) above or below the median height or weight for their ages, while 27.2 percent fall between 1 and 2 SDs above or below the median, and 4.6 percent fall 2 or more SDs from the median. Thus, one would expect to find 2.3 percent of well-nourished children to be quite short or thin for their age. The degree of moderate to severe malnutrition in a population is the proportion above the 2.3 percent normally expected which fall~ below 2 SDs from the median height or weight of the reference population. The proportion over 2.3 percent which is more than 2 SDs above the population median indicates the true proportion of children which is ovemourished. Figure 6.5 summarizes the height-for age, weight-for-height, and weight-for-age findings according to the age of children. At three months of age, children exceed the median measures of the reference population, indicating adequate nutritional status. However, slight stunting (as measured by height-for-age) appears in children age 4 months and above. Wasting, indicated by weight-for height measures falling below the reference population median, begins at 8 months of age. These findings are consistent with the timing of cessation of breastfeeding, which, for most women, occurs at 6 months. Tables 6.13 to 6.15 present each index according to selected background characteristics. Figure 6.5 Nutritional Status of Children Aged 3-36 Months SD's from International Reference 0 -1 i I l l l l l l l l l l l l l l l l l l l = l l l t l l l r l 0 6 12 18 24 30 36 Age of Children in Months Height-for-age - Weight-for-height -- Weight-for-age I Trinidad & Tobago DHS 1987 71 6.12 Nutritional Status of Children According to Height-For-Age Children's height, in reference to age in months shows the degree to which the population suffers from chronic rim]nutrition. Inadequate nourishment over a long period of drne (found typically in impoverishod ~s) results in stunted growth. Table 6.13A shows the proportion of children aged 3-36 months who fall into various standard deviation categories from the reference population median in terms of height-for-age. Four percent of the children arc moderately stunted, and less than 1 percent severely stunted, which slightly exceeds the levels expected in the reference population of well-fed children. Twenty-one percent fall between -2 and -1 SDs, 9 percent more children than expected. There is little difference in stunting according to the sex, age, and residence of children. Consistent with expectations, moderate stunting is more common among children age 6 months and above, and to those born less than two years after another birth. In addition, children of East Indian women are more likely to be stunted, possibly reflecting the educational (and economic) disadvantages of their mothers, as discussed previously. Table 6.13A Percent Distribution of Children Aged 3-36 Months by Standard Deviation Category of Helght-For-Age Using the NCBS/CDC/WHO International Reference Population, According to Selected Background Characteristics, TTDHS 1987 Standard Deviation from the Median of the NCHS/CDC/WHO Reference Population -3,00 -2.00 -1.08 -0.99 +1.00 +2.00 Background or to to to to or N~er of Characteristic more -2.99 -1.99 +0.99 +1.99 More Total Children EXPECTED IN REFERENCE POPULATION 0.I 2.2 13.6 68.2 13.6 2.3 I00 -- Sex M~le 8.2 4,9 21,1 63,1 6,9 3.7 180 407 Female 0.9 3.9 21.6 59.3 12.0 2.3 100 435 Age 3-5 Months 0.0 2.1 8.5 68.1 10.6 10.6 100 47 6-11 Months I.i 5.7 24.1 60.3 5.7 2.9 100 174 12-23 Months 0.3 5.0 26.9 58.1 7.0 2.7 180 301 24-36 Months 8.6 3.4 16.6 63.4 13.7 2.2 1O0 320 Prevlous Birth Interval First Birth 0.9 4.4 21.9 61.8 8.3 2.6 100 228 <24 Months 0.5 5.9 18.1 66.5 5.9 3.2 I00 221 24-47 Months 0.5 3.8 23.5 58.2 18.8 3.3 100 213 48 + Months 0.6 3.3 22.2 57.2 13.9 2.8 100 180 Residence Urban 0.6 4.4 18.6 58.1 13.9 4.4 100 339 Rural 0.6 4.4 23.3 63.2 6.6 2.8 I00 503 Motherts Educatlon <Complete primary 0.0 5.9 37.3 45.1 11.8 0.0 100 51 Completed primary 0.3 5.2 17.4 66.0 9.0 2.2 I00 368 Secondary I l I.i 4.2 25.0 58.4 8.4 2.8 I00 356 secondary II ~ 0.0 0.0 11.9 61.2 16.4 10.4 100 67 Ethnicity' African 0.0 3.5 17.6 61.0 13.1 4.8 100 313 Indian 0.8 5.9 25.9 60.0 5.3 2.1 100 375 Mixed 0.7 2.7 18.2 64.9 12.2 1.4 i00 148 Total 0.6 4.4 21.4 61.2 9.5 3.0 100 842 l some or full secondary education, but fewer * Some or full secondary education, with five level, or some university education. * Excludes 6 children of "other" ethnicity or than five "o" level exams passed. "0" level exams passed, at least one "A" with missing information. 72 Table 6,13B shows the height-for-age as a percent of the reference population median scores, with stunting categories as they are customarily presented in nutrition surveys in qYinidad and Tobago. One percent of children fall below 90 percent of the reference population median, suggesting near absence of stunted growth. Table 6.13B Percent Distribution of Children Aged 3-36 Months by Percent of Median Height-For-Age Using the NCHS/CDC/WHO International Reference Population, According to selected Background Characteristics, TTDHS 1987 Percent of Median of the NCHS/CDC/WHO Reference Population Background Number of Characteristic <90 90-110 >ll0 Total Children Sex Male 0.2 98,8 1.0 I00 407 Female 2.1 97,2 0.7 100 435 Age 3-5 Months 0.0 95,7 4.3 100 47 6-11 Months 2.3 96,6 I.i I00 174 12-23 Months 1.0 98.7 0.3 i00 301 24-36 Months 0.9 98,4 0.6 i00 320 Previous Birth Interval First Birth 1.8 97,8 0.4 i00 228 <24 Months 1.4 97,7 0.9 i00 221 24-47 Months 0.9 97,7 1.4 i00 213 48+ Months 0.6 9H,9 0.6 lO0 180 Residence Urban 1.8 97.1 1.2 100 339 Rural 0.8 98.6 0.6 i00 503 Mother's Education <Complete primary 0.0 i00,0 0.0 I00 51 Completed primary 0.5 98.9 0.5 i00 368 Secondary I ~ 2.2 96,9 0.8 190 356 Secondary II* 0.0 97,0 3.0 10S 67 Ethnicity 3 African 1,3 97,4 1.3 100 313 Indian 1.1 98.1 0.8 i00 375 Mixed 0.7 99.3 0.0 i00 148 Total 1.2 98.0 0.8 I00 842 * Some or full secondary education, but fewer than five "0" level exams passed. • Some or full secondary education, with five "o" level exams passed, at least one "A" level, or some university education. ' Excludes 6 children of "other" ethnlcity or with missing information. 6.13 Nutr i t ional Status of Chi ldren Accord ing to Weight - for -He ight In Table 6.14A, children who are between -2 and -3 SDs below the median of the refer- ence population are considered thin for their height or moderately wasted, while those below -3 SDs are severely wasted. In contrast to stunting, wasting is the result of inadequate nourishment in the months immediately preceding the survey, and can develop quite rapidly. Infection, diarrhoeal diseases, and seasonal reduction in the food supply can result in wasting. Since the child's age is not included in this measure, weight-for-beight is free of bias introduced by age misreporting. 73 Overall, 3 percent of children are moderately wasted, which is just slightly above the level found in the reference population. Severe wasting, although rare in the survey, exceeds the level found in the reference population, as well. Nearly 9 percent of children 3-5 months of age are moderately or severely wasted; note, however, that very few children fall into this age category. Wasting is most common among children of East Indian women (7 percent), and is nearly absent in the other ethnic groups. Differences according to other background characteristics are slight. Table 6.14A Percent Distribution of Children Aged 3-36 Months by Standard Deviation Category of Weight-For-Height Using the RCHS/CDC/WHO International Reference Population, According to Selected Background Characteristics, TTDHS 1987 standard Deviation from the Median of the NCHS/DDC/WHO Reference Population -3.00 -2.00 -i.00 -0.99 +i.00 +2.00 Background or to to to to o~ Characteristic more -2.99 -1.99 +0.99 +1.99 More Number of Total Children EXPECTED IN REFERENCE POPULATION 0.1 2.2 13.6 68.2 13.6 2.3 100 -- Sex Male 1.0 2.7 19.2 62.4 11.5 3.2 100 407 Female 0.5 3.4 21.4 65.5 5.7 3.4 100 435 Age 3-5 Months 4.3 4.3 8.5 55.3 17.0 10.6 100 47 6-11 Months 0.6 2.3 15.5 63.8 12.1 5.7 100 174 12-23 Months 0.3 4.3 23.9 61.5 7.3 2.7 i00 301 24-36 Months 0.6 2.2 21.2 67.8 6.6 1.6 100 320 Previous Birth Interval First Birth 0.4 3.1 16.0 63.6 8.8 6.1 100 228 <24 Months 0.0 3.2 22.2 63.6 8.6 2.3 100 221 24-47 Months 1.9 2.6 21.1 64.6 7.0 2.3 i00 213 48 + Months 0.6 3.3 20.0 63.9 I0.0 2.2 i00 180 R~sldenoe Urban 0.3 2.4 15.0 67.6 10.3 4.4 i00 339 Rural 1.0 3.6 23.9 61.6 7.4 2.6 100 503 Mother's Education <Complete primary 0.0 3.9 31.4 59.6 5.9 0.0 100 51 Completed primary 0.8 3.3 20.9 63.9 6.2 3.0 1O0 368 Secondary I ~ 0.6 2.8 20.2 63.5 9.3 3.7 I00 356 Secondary II* 1.5 3.0 9.0 71.6 9.0 6.0 I00 67 Ethnlclty' African 0.0 1.6 12.1 70.9 10.5 4.8 i00 313 Indian 1.6 5.3 29.1 56.8 5.1 2.1 I00 375 Mixed 0.O 0.0 15.5 67.6 13.5 3.4 1O0 148 Total 0.7 3.1 20.3 64.0 6.6 3.3 100 842 * Some or full secondary education, but fewer than five "o" level exams passed. a Some or full secondary education, with five "o" level exams passed, at least one "A" level, or some university education. ' Excludes 6 children of "other" ethniclty or with missing information. Table 6.14B shows weight-for-height as a percent of the reference population median according to the categories used in other local nutrition surveys. Two percent of children are wasted, while nearly 3 percent are ovemourished. Wasting is most common among children 3-5 months of age, and among children of the most highly educated mothers. (Note that few children fall into either category, however, so caution is necessary in interpreting the results.) 74 Table 6.14B Percent Distribution of Children Aged 3-36 Months by Percent of Median Weight-For-Helght Using the NCHS/CDC/WHO International Reference Population, According to Selected Background Characteristics, TTDHS 1987 Percent of Median of the NCHS/CDC/WHO Reference Population Background NUmber of Characteristic <80 80-120 >120 Total Children Sex Male 1.7 95.6 2.7 I00 407 Female 2.3 95.2 2.5 I00 435 Age 3-5 Months 8.5 83.0 8.5 i00 47 6-11 Months 1.7 93.1 5.2 100 174 12-23 Months 1.7 97.0 1.3 100 301 24-36 Months 1.6 96.9 1.6 i00 320 Previous Birth Interval First Birth 1.8 93.4 4.8 i00 228 <24 Months 2.3 95.9 1.8 i00 221 24-47 Months 2.8 95.3 1.9 100 213 48+ Months i.i 97.2 1.7 100 180 Residence urban 2.1 94.4 3.5 180 339 Rural 2.0 96.0 2.0 i00 503 Mother's Education <Complete primary 2.0 98.0 0.0 i00 51 Completed primary 1.9 96.2 1.9 I00 368 Secondary I ~ 1.7 95.2 3.1 i00 356 Secondary II z 4.5 89.6 6.0 100 67 Ethnlclty* African 0.6 95.5 3.8 i00 313 Indian 3.7 94.4 1.9 i00 375 Mixed O.0 98.0 2.0 i00 148 Total 2.0 95.4 2.6 i00 842 t Some or full secondary education, but fewer than five "0" level exams passed. * Some or full secondary education, with five "0" level exams passed, at least one "A" level, or some university education. * Excludes 6 children of "other" ethnlclty or with missing information. 6.14 Nutritional Status of Children According to Weight-for-Age Tables 6.15A and 6.15B present weight-for-age data for comparison with other surveys which might contain information on the weight of children but not their height. In Table 6.15A, nearly 7 percent of the children weigh less than -2 SDs below the reference population median weight at given ages. Children whose mothers did not complete primary school, or are East Indian, are most likely to be moderately or severely undernourished. The results measured in terms of percent of the reference population median, as shown in Table 6.15B are similar. 6.15 Summary of the Nutritional Status of Children Aged 3-36 Months Tables 6.16A and 6.16B cross-tabulate the weight-for-height and height-for-age indices, showing the relationship between stunting and wasting. Table 6.16A shows the SD categories for comparison with other DHS surveys. While nearly 5 percent of children are stunted and nearly 4 percent wasted, less than 1 percent are both stunted and wasted. Table 6.16B shows the same cross-tabulation in terms of percent of median. Only 1 percent of children are stunted, only 2 percent are wasted, and less than one percent are both stunted and wasted. 75 Table 6.15A Percent Distribution of Children Aged 3-36 Months by Standard Deviation category of Welght-For-Age Using the NCBS/CDC/WHO International Reference Population, According to Selected Background Characteristics, TTDHS 1987 Standard Deviation from the Median of the NCHS/CDC/WHO Reference Population -3.00 -2.00 -i.00 -8.99 +i.00 +2.00 Background or to to to to or Number of Characteristic More -2.99 -1.99 +0.99 +1.99 More Total Children EXPECTED IN REFERENCE POPULATION 0.i 2.2 13.6 68.2 13.6 2.3 i00 -- Sex Male 0.5 5.7 27.0 55.3 7.6 3.9 100 407 Female 0.2 7.4 29.0 53.3 6.4 3.7 100 435 Age 3-5 Months 0.0 0.0 10.6 55.3 23.4 10.6 I00 47 6-11 Months 0.0 6.9 29.3 51.7 6.9 5.2 i00 174 12-23 Months 0.7 8.6 28.2 54.8 4.0 3.7 I00 301 24-36 Months 0.3 5.3 29.7 55.0 7.5 2.2 i00 320 Previous Birth Interval First Birth 0.0 5.7 25.9 54.8 8.6 4.8 i00 228 <24 Months 0.5 7.2 29.9 55.7 2.7 4.1 I00 221 24-47 Months 0.5 7.5 29.1 52.1 8.0 2.8 i00 213 48+ Months 0.6 5.6 27.2 54.4 8.9 3.3 100 180 Residence Urban 0.3 4.7 22.1 58.7 8.3 5.9 I00 339 Rural 0.4 7.8 32.0 51.3 6.2 2.4 100 503 Mother's Education <Complete primary 0.0 11.8 45.1 41.2 2.0 0.0 I00 51 Completed primary 0.3 7.3 26.9 56.3 6.0 3.3 100 368 Secondary I* 0.6 5.9 28.4 53.1 8.1 3.9 i00 356 Secondary II ~ 0.0 1.5 19.4 59.7 10.4 9.0 I00 67 Ethnlcity* African 0.0 2.2 21.1 61.0 8.9 6.7 100 313 Indian 0.5 11.2 36.3 46.1 4.0 1.9 100 375 Mixed 0.0 4.1 22.3 60.1 10.8 2.7 i00 148 Total 0.4 6.5 28.0 54.3 7.0 3.8 i00 842 * Some or full secondary education, but fewer than flve "0" level exams passed. i Some or full secondary education, with five "O" level exams passed, at least one "A" level, or some university education. * Excludes 6 children of "other" ethnicity or with missing information. 76 Table 6.15B Percent Distribution of Children Aged 3-36 Months by Percent of Median Welght-For-Age Using the NCBS/CDC/WBO International Reference Population, According to Selected Background Characteristics, TTDBS 1987 Percent of Median of the NCBS/CDC/WHO Reference Population Background Number of Characteristic <88 80-120 >120 Total Children Sex Male 7.6 87.0 5.4 10O 407 Female 9.4 85.5 5.1 100 435 Age 3-5 Months 4.3 74.5 21.3 100 47 6-11 Months 11.5 82.8 5.7 i00 174 12-23 Months 10.3 86.0 3.7 100 301 24-36 Months 5.9 90.0 4.1 i00 320 Previous Birth Interval First Birth 7.9 85.5 6.6 i00 228 <24 Months 8.1 87.8 4.1 100 221 24-47 Months 9.8 85.4 4.7 100 213 48+ Months 8.3 86.1 5.6 i00 180 Residence urban 5.9 86.7 7.4 I00 339 Rural 10.3 85.9 3.8 100 503 Mother's Education <complete primary 17.6 82.4 0.0 100 51 Completed primary 9.0 86.7 4.3 i00 368 Secondary 11 8.1 86.2 5.6 100 356 Secondary II 2 1.5 86.6 11.9 i00 67 Ethniclty* African 3.5 87.2 9.3 i00 313 Indian 14.1 83.7 2.1 i00 375 Mixed 4.7 90.5 4.7 i00 148 Total 8.6 86.2 5.2 i00 842 i Some or full secondary education, but fewer than five "o" level exams passed. a Some or full secondary education, with five "O" level exams passed, at least one "A" level, or some university education. * Excludes 6 children of "other" ethnlcity or with missing information. 77 Table 6.16A Percent Distribution of Children Aged 3-36 Months, the Percent in Each Height-for-Age standard Deviation category by Each Welght-for-Helght Standard Deviation Category (Waterlow Classification) Using the NCHS/CDC/WHO International Reference Population, TTDHS 1987 Welght-for-Helght Standard Deviation from NCHS/CDC/WHO Reference Population Helght-for-Age Standard Deviation -2.00 -1.00 -0.99 +i.00 from NCHS/CDC/WHO or to to or Total Nu,~0er of Reference Population More -1.99 +0.99 More Percent Children -2.00 or more 0.3 1.0 2.7 1.0 5.0 42 -i.00 to ~1.99 0.7 6.3 13.2 1.2 21.4 180 -0.99 to +0.99 2.1 11.9 40.1 7.0 61.2 515 ÷i.00 or more 0.6 1.2 8.0 2.7 12.5 105 Total 3.0 20.3 64.0 11.9 100 -- Number 32 171 539 100 -- 842 Table 6.16B Percent Distribution of Children Aged 3-36 Months, the Percent in Each Helght-for-Age Percent of Median Category by Each Welght-for-Helght Percent of Median Category (waterlow Classification) Using the NCHS/CDC/WHO International Reference Population, TTDHS 1987 Welght-for-Helght Percent of Median NCHS/CDC/WH0 Reference Population Helght-for-Age Percent of Median NCHS/CDC/WHO Total Number of Reference Population <80 80-120 >120 Percent Children <90 0.I 0.6 0.5 1.2 i0 90-110 1.9 94.3 1.0 98.0 825 >ii0 0.0 0.5 0.4 0.6 7 Total 2.0 95.4 2.6 i00 -- Number 17 803 22 -- 842 78 REFERENCES Central Statistics Office. 1987a. 1986 Annual Statistical Digest, no.3. Republic of Trinidad and Tobago. Central Statistics Office. 1987b. 1984 Pooulation and Vital Statistics Reoort. Republic of Trinidad and Tobago. Central Statistics Office. 1987c. Social Indicators Reoort. Republic of Trinidad and Tobago. Central Statistics Office. 1983. 1980 Pooulation and Housing Census. vol. 2. Republic of Trinidad and Tobago. Central Statistics Office. 1981. Trinidad and Toba~o Fertility Survey 1977. 2 vols. Republic of Trinidad and Tobago. Consejo Nacional de Poblaci6n y Familia (CONAPOFA) and Institute for Resource Development/ Westinghouse. 1987. Reoublica Dominicana Encuesta Demot, rafic;i v de Salud. DHS-1986. Santo Domingo, Dominican Rel~ublic: CONAPOFA. Corporaci6n Centro Regional de Poblaci6n (CCRP) and Ministerio de Salud de Colombia and Institute for Resource Development/Westinghouse. 1988. Tercera Encuesta Nacional de Prevalencia del Uso de Anticonceotivos v Pdmera ~1¢ DCm0gafia v Salud. 1986. Bogota, Colombia: CCRP. Ebanks, G.E. 1985. Infant and Child Mortality and Fertilitv: Trinidad and Toba~o. Guvana. and Jamaica. Voorburg, Netherlands: International Statistical lhstitute. (WFS Scientific Repurts no. 75). Harewood, J. 1978. Female Fertility and Family Plartqiag in Trinidad and Toba¢,o. Kingston, Jamaica: University of the West Indies, Institute of Social and Economic Research. Hunte, D. 1983. Evaluation of the Trinidad and Toba~o Fertiiltv Survey 1977. Voorburg, Netherlands: International Statistical Institute. (WFS Scientific Reports no. 44). Martorell, R. and Habicht, J.-P. 1986. "Growth in Early Childhood in Developing Countries." In Human Growth: A Comorehensive Treatise, edited by F. Falkner and J.M. Tanner, vol. 3. New York and London: Plenum Press. Sather, Z.A. and Chidambaram, V.C. 1984. Netherlands: Intemational Statistical Institute. national Summaries no. 36). Differentials in Contraceotive Use. Voorburg, (World Fertility Survey Cornparative Studies Cross- Tmssell, J. and Kost, K. 1987. "Contraceptive Failure in the United States: A Critical Review of the Literature." Studies in Family Planning 18(5): 237-283. United Nations. Department of Technical Co-operation for Development and Statistical Office. 1986. How to Wei~,h and Measure Children: Assessing, the Nutritional Status of Youn~ Children in Household Surve-vs. New York: United Nations.- (National Household Survey Capability Programme). United States. Department of Health, Education, and Welfare. National Center for Health Statistics. 1976. "NCHS Growth Charts, 1976." Monthlv Vital Statistics Reeort 25 (June 22; Suppl.), no. 3. 79 80 APPENDIX A SURVEY DESIGN 81 82 APPENDIX A SURVEY DESIGN A.1 Sample Design and Implementation The sample for the TTDHS was based on the Continuous Sample Survey of Population (CSSP), used by the Central Statistical Office since 1968, and redesigned on the basis of the 1980 Population and Housing Census. The country is divided into 14 domains of study, comprising a total of 1,638 enumeration districts (EDs). Results from the 1980 Census indicated that some EDs were too large (more than 300 households) and some too small (fewer than 30 households) to be appropriate primary sampling units (PSUs) for the TFDHS. Therefore, the largest units were further subdivided, and the smaller units combined with contiguous ones for the CSSP sample. The CSSP sample is selected in two stages. In the first, PSUs are systematically selected, with probability proportional to size (size equals the number of households in the PSU). Following an operation to list all households in each selected PSU, individual households are selected, with probability of selection inversely proportional to the PSU's size. The CSSP grand sample, which provides an overall sampling fraction of one household in forty (1/40) has been divided into 9 sub-samples, each with an overall sampling fraction of one in three-hundred sixty (1/360). Each CSSP survey round, conducted quarterly, uses three of the nine sub-samples, with an overall sampling fraction of one in one-hundred twenty (1/120). The DHS sample was taken from the CSSP sample selected for the January-March 1987 quarter. The main objectives of the DHS sample were: o a self-weighting sample of households, o a sample take in each selected PSU of about 25 women aged 15-49, and o a total of 4,000 completed interviews with women aged 15-49. To achieve this sample size, 5,000 households were selected. This figure assumes an average of one eligible woman per household, and 294,400 eligible women nationwide, giving an overall sampling fraction of one in sixty (1/60). It also allows for 10 percent non-response at both the household and the individual interview level, commensurate with CSO experience in similar recent surveys. In total, 178 PSUs were selected throughout Trinidad and Tobago. The CSO provided each team with maps of the areas in which they were working. According to the CSSP sample design, f* = P*l P*2 = m" Mi . . . . . . m*2 M where f* P*I overall selection probability (all stages) for households in the CSSP sample, overall selection probability for PSUs in the CSSP sample 83 m*2 m* selection probability for households within PSUs in the CSSP sample number of PSUs selected for the CSSP Mt = measure of size of the i - th PSU, and M = E M~ the sum over all PSUs in Trinidad and Tobago To achieve the TI'DHS sample, the following design was used. (Notation defined above but without an asterisk refers to the corresponding information in the TTDHS sample.) f = Pl P2 = m M I i M 12 m m* M i 1 m' M 12 = m 1 ~ p*l m" 12 where I 2 is the household selection interval in the selected PSU. Households in selected PSUs were selected with the corresponding sampling interval 12. The sampling interval was applied in each PSU beginning with a household selected at random using a table of random numbers. A.2 Questionnaire Design and Training The DHS model "A" questionnaire was adapted for use in Trinidad and Tobago, and pretested during February 1987. Thirteen pretest interviewers were trained for two weeks by FPATI', CSO, and IRD staff, and carded out two days of interviews. The questionnaire was further modified based on pretest results and interviewer comments. Female interviewers were trained for the main survey for four weeks during April and May, 1987. Training consisted of two weeks of classroom lectures, discussions, and practice interviews, followed by a written exam. Trainees then worked in teams conducting practice interviews. To ensure proper supervision of interviewers, field personnel were divided into 5 teams, 4 for Trinidad and 1 for Tobago. Teams consisted of 1 supervisor, 1 field editor/anthropometric measurer, and 4 or 5 interviewers. Supervisors and field editors received special training in their respective duties. The former were taught to read maps and to use the household listings provided by the CSO for each PSU selected in the sample. The latter were trained to scrutinize questionnaires for accuracy, completeness, and consistency. In addition, supervisors and field editors were gained by IRD's anthropometrist to weigh and measure young children. Trainees were taught to measure children to within 100 grams of their true weight, and 0.5 centimetres of their true length, with the true measure defined by the mean of two measurements performed by the gainer. The precision and accuracy of the measurements were assessed during standardization tests administered at the conclusion of the 84 anthropometric training and again midway through data collection. The former test was administered by IRD's anthropometrist, the latter by a nutritional biochemist from the Ministry of Health. A.3 Fieldwork Thirty-three field personnel, including 1 fieldwork coordinator, 5 supervisors, 4 field editors, and 23 interviewers commenced data collection on May 14. Fieldwork required nearly six weeks longer than scheduled. The two main problems were inaccurate maps and lack of adequate transport. Maps were in many cases outdated, and more time was spent locating the selected households than was anticipated. Secondly, lack of adequate transport made it difficult for interviewers to reach their assigned areas. One result of the tr

View the publication