Sri Lanka - Demographic and Health Survey - 1988
Publication date: 1988
~ ~ ~r ~ ¸ ~ Sri Lanka Demographic and Health Survey 1987 Deparlment of Census and Statistics M in~ of Plan Implementation Colombo, Sri Lanka Institute for Resource Development/Westinghouse Columbia, Maryland, USA May 1988 I SBN 955 - 577 - 009 - 3 Publ ished in 1988. This report presents the findings of the Sri Lanka Demographic and Health Survey, implemented by the Department of Census and Statistics, Ministry of Plan Implementation in 1987. The survey is part of the worldwide Demographic and Health Surveys (DHS) Programme which is designed to collect data on fertility, family planning, and maternal and child health. Additional information on this survey can be obtained from: Director, Department of Census and Statistics, No. 6, Albert Crescent Road, P.O. Box 563, Colombo 7, Sri Lanka. The Sri Lanka Demographic and Health Survey was carried out with assistance from the Institute for Resource Development (IRD), a subsidiary of Westinghouse Electrlc Corporation, with offices in Col-mbia, Maryland, U.S.A. Funding for the survey was provided under a contract with the U.S. Agency for Internatlonal Development (Contract No. DPE-3023-C-00-4083-00). Additional information about the DHS Programme can be obtained by writing to: Director, DHS Programme, IRD/Westinghouse, P.O. Box 866, Columbia, Maryland, 21044, U.S.A. (Telex 87775). FOREWORD The Demographic and Health Survey (DHS) is an important link in a chain of surveys carried out in Sri Lanka in the past decade or so. Having been designed as part of an international survey program and modelled on the lines of the well renowned World Fertility Survey (WFS) program, the DHS provides an exceptionally valuable source of data for the estimation of trends over time within Sri Lanka as well as for cross national comparison. The survey focussed primarily on fertility, contraception and child mortality as did WFS but. also measured several indicators of child health, particularly immunization coverage and nutrition status. The inclusion of health sector information has been welcome and fruitful, for improve- ment of nutrition status is a subject to which the Government of Sri Lanka has accorded high priority. The DHS has been an exception in the series of demographic surveys in Sri Lanka for two reasons - firstly for the rapid speed with which it was completed and secondly for the rapid dissemination of its findings. The findings from a preliminary analysis have already been presented to those concerned at two seminars, and through two advanced publications before the publication of this survey Report. To see the findings of a survey being discussed and utilized is indeed a great satisfaction. We hope that the data would be further analyzed and put to pragmatic use in impro- ving the situation of all people. I must thank the United States Agency for International Development (USAID) for funding this survey, the Institute for Resource Development (Westinghouse) for implementing the survey pro- gram and the survey staff of the Department of census and statistics for their contribution to conducting the survey in Sri Lanka. My 'thanks are also due to the survey respondents for their excellent cooperation. Survey Director and Director of Census and Statistics. 4 May 1988. ACKNOWLEDGEMENTS Many individuals and institutions made valuable contributions to the implementation of the Sri Lanka Demographic and Health Survey. While all such contributions are gratefully acknowledged, some major contributions are mentioned below. The USAID provided the funding for the survey. The staff of USAID, Colombo efficiently coor- dinated administrative activities between the Institute for Resource Development and the Depart- ment of Census and Statistics. Special thanks are due to Dr. Gnani Thenabadu of USAID for her continued interest and assistance. The Director of Census and Statistics, Mr. R.B.M. Korale was the National Director of the survey. Very special thanks are due to him for his able administration of the survey. Thanks are also due to Mr. A.A.D.C. Yasasiri, the Additional Director for his administrative assistance. The Institute for Resource Development made highly efficient arrangements for technical colla- boration. Very special thanks are due to Dr. Andrew Fisher for his able coordination and hard work, and to Ms. Jeanne Cushine for her expert assistance in setting up the data processing pro- cedures which proved to be extremely efficient. Thanks are also due to Dr. J. Sullivan, Ms. Amy Shoen, Mr. Roger Pearson. The Population Census Division of the Department of Census and Statistics conducted the survey. Thanks are due to the staff of this Division for their commitment and hard work. Mr. K.H.W. Gaminirathna the Deputy Director and Mrs. Some De Silva, Acting Deputy Director functioned as Project Manager and Assistant Project Manager of the survey. The District Statistical Officers made logistical arrangements for field work within their respective districts. The field work was carried out by the Statistical Investigators of the Department. Supporting services were provided by the Accounts and Transport Divisions of the Head Office of the Department. This Report was prepared by a team and was edited by Dr. Andrew Fisher assisted by Mrs. Soma De Silva. Chapter 1 'Background' was contributed by Mrs. Swarna Ukwatta, chapter 2 'Nuptiality and Exp'osure to Risk of Conception' by Mrs. Some De Silva, chapter 3 'Fertility' and chapter 6 'Mortality and Health' by Dr. Andrew Fisher, chapter 4 'Fertility Regulation' by Mr. K.H.W. Gami- nirathne, chapter 5 'Fertility Preferences' by Miss Padmini De Silva. The Printing Division of the Department printed this Report in a short period of time. Special thanks are due to the staff of this Division and to the Assistant Director Mr. K.D. Siripala. TABLE OF CONTENTS Page FOREWORD - , , , , , , . , , , , ~ , o , , , , , , o , o o o o , , , , , o o , o , , , , , , o o , o , , , , , i ACKNOWLEDGEMENTS '~ " ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' " i i i L IST OF TABLES v i i i " O O ' ' ' ' ' O O 6 O D J O , , O , O m O O O J O O O , , , , , , Q D , , , O , e L IST OF F IGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv i CHAPTER 1 : Background i . i H i s to ry , geography , and economy . . . . . . . . . . . . 1.2 Popu la t ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 Fami ly p lann ing po l i c ies and programmes . 1.4 Hea l th pr io r i t ies and programmes . . . . . . . . . . . 1.5 Ob ject ives of the survey . . . . . . . . . . . . . . . . . . . 1.6 Survey sample des ign . . . . . . . . . . . . . . . . . . . 1.7 Survey organ izat ion and f ie ld work . . . . . . . . . . 1.8 The SLDHS quest ionna i re and respondents . . . . . . . 1.9 O D O D Q Q O l l m O 9 Q O O I Q O O O O 1 0 Background Var iab les . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 3 4 5 5 9 13 15 CHAPTER 2: Nupt ia l i ty and Exposure to R isk of Concept ion 2.1 Cur rent mar i ta l s ta tus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 2.3 2.4 Age at f i r s t un ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cur rent exposure s ta tus , , o e i , , e , , , , , e q , e o o , o o o o , , . , , Breast feed ing , amenorrhea , abst inence * 19 20 22 27 CHAPTER 3 : 3 .1 3 .2 3 .3 3 .4 3 .5 3 .6 Fer t i l i ty In t roduct ion . . . . ~ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cur rent and cumulat ive fe r t i l i ty . . . . . . . . . . . . . . . . . . . Fer t i l i ty t rends . . . . . . . . . . . . . . . ~ . . . . . . . . . . . . . . . . . . . Ch i ld ren ever born . . . . . . . . . . . . . ~. . . . . . . . . . . . . . . . . . . Ch i ld ren ever born and age a t marr iage . . . . . . . . . . . . . Age a t f i r s t b i r th . . . . . . . . . . . . . . . . . . . . . ~.~ . . . . . . . . . 31 32 36 40 44 46 CHAPTER 4 : 4 .1 4 .2 4 .3 4 .4 4 .5 4 .6 4 .7 4.8 4.9 4 .10 4 .11 4 .12 4 .13 4 .14 4 .15 4 .16 Fer t i l i ty Regu la t ion In t roduct ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ~.; . 49 Cont racept ive knowledge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acceptab i l i ty o f methods . . . . . . . . . . . . . . . . . . . . . . . . . . Source knowledge .~ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ever use o f cont racept ives . o o o a o o o e o e o e o e e . o o e o o o o a urrent ~ " " Trends in c use o f cont racept lo n . . . . o. o o o o. Cur rent use by method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D i f fe rent ia l s in cont racept iv e method use . . . . . . . . . . N, Imher o f ch i ld ren a t t ime o f f i r s t use . . . . . . . . . . . . Knowledge o f fe r t i l e per iod . . . . . . . . . . . . . . . . . . . . . . . . . Age a t s te r i l i za t ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Source o f cont racept ive supp ly and sat i s fac t ion w i th source . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reasons fo r method d iscont inuat ion . . . . . . . . . . . . . . . . . . A t t i tudes towards becoming pregnant . . . . . . . . . . . . . . . . . Fu ture use . . . . . . . . . . . . . . . . . . . . . ~ . . . . . . . . . . . . . . . . . . Fami ly P lann ing messages on rad io . . . . . . . . . . . . . . . . . . . 49 51 52 54 56 6Q 61 63 66 68 68 69 70 71 74 vi C~ER 5: 5 .1 5.2 5.3 5~4 Fer t i l i ty P re ferences In t roduct ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 " Des i re for more ch i ld ren . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 Women in need of fami ly p lann ing . . . . . . . . . . . . . . . . . . . . 86 Idea l nnmher of ch i ld ren . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 5.5 Fer t i l i ty p lann ing and the s ta tus o f b i r ths . . . . . . . . . 89 CHAPTER 6 : 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6 .9 6 .10 Mor ta l i ty and Hea l th Background . . . . . . . . . . . . . . . . . . . . . ; . . . . . . . . . . . . . . . . . . . . . 94 In fant and ch i ld mor ta l i ty . . . . . . . . . . . . . . . . . . . . . . . . . . 94 P renata l care, te tanus toxo id in jec t ions , and ass i s tance at de l ivery . . . . . . . . . . . . . . . . . . . . . . . . . . 99 Immunizat ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Treatment o f d ia r rhea . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . 108 Anthropometr i c measurements o f length and we ight . 109 He ight - fo r -age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 Weight - fo r -he ight . . . . . . . . . . . . . . . . . . . . : . . . . . . . . . . . . . . 117 Height-for-age by we ight - fo r -he ight . . . . . . . . . . . . . . . . . 118 Weight - fo r -age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 APPENDIX A: Es t imates of sampl ing er ror . . . . . . . . . . . . . . . . . 126 APPENDIX B: APPENDIX C: Quest ionna i re . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 L i s t o f survey s ta f f . . . . . . . . . . . . . . . . . . . . . . . . 180 vii LIST OF TABLES Page TABLE I.I: TABLE 1.2: TABLE 1.3: TABLE 1.4: TABLE 1.5: TABLE 2.1: TABLE 2.2 : TABLE 2.3: Household and individual weights appl ied to SLDHS data, SLDHS 1987 . Percent d is t r ibut ion of the female populat ion by age from all women in the 1987 SLDHS household sample , and from the 1984 est imated female populat ion for a l i sri Lanka . . . . . . . . ~ . . . . . . . . . . . . . . . . . . . . . Resul ts of SLDHS household and indiv idual interviews by sector, SLDHS 1987 . . . . . . . . . . . . . . Percent d is t r ibut ion of ever -marr ied women aged 15-49 by level of educat ion accord ing to se lected background character ist ics, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iO 13 15 17 Percent d is t r ibut ion of eyer -marr fed women by selected background character is t ics , SLDHS 1987. . ~ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Percent d is t r ibut ion of ever -marr ied women aged 15-49 accord ing to age by current mar i ta l status, SLDHS 1987 . . . . . . . . . . . . . . . . 20 Percent d is t r ibut ion of all ever -marr ied and never marr ied Women (from household schedule) according to current age by age at f i rst un ion and~median age at f irst union accord ing to current age, SLDHS 1987 . Proport ion of all women in f ive year age groups who have ever-marr ied, and s ingu la te mean age a t marr iage (SMAM), 1963 census, 1971 census, 1975 SLWFS, 1981 census , and 1987 SLDHS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 22 viii TABLE 2.4: TABLE 2.5: TABLE 2.6: TABLE 2.7: TABLE 3.1: TABLE 3~2: TABLE 3.3: TABLE 3.4 : TABLE 3.5: Med ian age at f i rs t un ion among al l ever -marr ied women (from househo ld schedule) aged 25 - 49 years by sector and zone, SLDHS 1987 . . . . . . . . . . . . . ~ . . . . . . . . . . . . 23 Percent d i s t r ibut ion of cur rent ly marr ied women aged 15-49 by exposure to concept ion s ta tus accord ing to cur rent age, SLDHS 1987 . . . . . . . - . . . . . . . . . . . . . . . . . . . . . . . . 24 P ropor t ion of b i r ths in the las t 36 months who are be ing breast fed and whose mothers are s t i l l amenor rhe ic , absta in ing , o r insuscept ib le by number of months s ince b i r th , SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . •28 Among ever -marr ied Women, the es t imated mean number of months of breastfeeding, amenorrhea , postpar tum abst inence and postpar tum insuscept ib i l i ty by se lec ted background character i s t i cs , SLDHS 1987 . . . . . . . . To ta l Fer t i l i ty Rates (TFR) for ca lendar year per iods and for f ive years preced ing the survey, and mean ch i ld ren ever born (CEB) to a l l never marr ied and ever - marr ied women 45-49 years ( f rom househo ld schedule) • by background character i s t i cs , SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 33 A compar i son of age spec i f i c fe r t i l i ty ra tes f rom var ious sample surveys in Sr i Lanka and f rom the c iv i l reg is t ra t ion sys tem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Age-per iod fe r t i l i ty ra tes (per 1 ,000 women) by materna l age at b i r th of ch i ld ; SLDHS 1987 . . . . . . . . . . . . . . ~ . . . . . . . . . . . . . . ~ . . . . . . . 40 Percent d i s t r ibut ion of ever -marr ied women aged 15-49 by number ch i ld ren ever born (CEB) accord ing toage , SLDHS 1987 and SLWFS 1975 . . . . . . . . . . . . . . . . . ~ . . . . . . . . . . . . . . 42 Percent d i s t r ibut ion of cur rent ly marr ied women aged 15-49 by number of ch i ld ren ever born (CEB) accord ing to cur rent age, SLDHS 1987 and SLWFS 1975 . . . . . . . . . . . . . . . . . . . . . 42 ix TABLE 3.6 : TABLE 3 .7 : TABLE 3.8: TABLE 3.9 : TABLE 4.1: TABLE 4.2: TABLE 4.3: TABLE 4.4: TABLE 4.5: TABLE 4.6: Percent d ist r ibut ion of ever-marr ied and current ly marr ied women aged 45-49 by n,~mher of chi ldren ever born, 1975 SLWFS, 1982 SLCPS and 1987 SLDHS . . . . . . . . . . . . . . . . . . . . . 44 Mean n1*mber of chi ldren ever born (CEB) to ever-marr ied women, by age at f irst marr iage and durat ion since first marrdage, SLDHS 1987 and CEB for all ages and durations, SLWFS 1975 . . . . . . . . . . . . . . . . . . . . . 46 Percent d istr ibut ion of all women by age at f irst birth ( including the category 'no birth'), according to current age, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Median age at f irst b i r th among all women aged 25-49 by current age and background character ist ics, SLDHS 1987 . Percentage of ever-marr ied women aged 15-49 knowing any method, knowing any modern method, and knowing speci f ic ~ontracept ive methods, SLDHS 1987 . . . . . . . . . Percentage of current ly marr ied women aged 15-49 knowing at least one modern method, by number of l iv ing chi ldren and selected background character ist ics, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Percent d istr ibut ion of ever-marr ied women aged 15-49 who have ever heard of a method by the main problem perceived in using part icu lar methods, if any, according to method, SLDHS 1987 . . . . . . . . . . . . . . Percent d ist r ibut ion of ever-marr ied women aged 15-49 knowing a method by supply source named (if any), SLDHS 1987 . Percent of ever-marr ied women 15-49 who have ever used a contracept ive method by type of method, 1975 SLWFS, 1982 SLCPS, and 1987 SLDHS . . . . . . . . . . . . . . . . . . . . . . . . . Percent of ever-marr ied women aged 15-49 who have ever used speci f ic methods of contraception, by current age, SLDHS 1987 . 47 48 51 52 53 54 55 57 X TABLE 4.7: TABLE 4.8: TABLE 4.9: TABLE 4.10: k TABLE 4.11: TABLE 4.12: TABLE 4.13: TABLE 4.14: TABLE 4.15: TABLE 4.16: Percent of current ly marr ied women aged 15-49 who have ever used speci f ic methods of contraception, by current age, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Trends in current contracept ive use by method among current ly marr ied women aged 15-49 from the 1975 SLWFS, 1982 SLCPS, and 1987 SLDHS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Among current users, changes in method mix between the 1975 SLWFS, 1982 SLCPS, and 1987 SLDHS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Percent d ist r ibut ion of current ly marr ied users aged 15-49 by speci f ic method of contracept ion according to current age, 1975 SLWFS, 1982 SLCPS, and 1987 SLDHS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Percent of current ly marr ied women aged 15-49 according to contracept ive method current ly using by age, SLDHS 1987 . . . . . . . . . . Percent d ist r ibut ion of current ly marr ied women aged 15-49 by method of contracept ion current ly using, according to selected background character ist ics, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Percent d ist r ibut ion of ever-marr ied women aged 15-49 by number of l iv ing chi ldren at the t ime of f irst use of contracept ion by current age, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . Percent d istr ibut ion of ever-marr ied women aged 15-49 and women ever us ing per iodic abst inence by knowledge of the fert i le per iod dur ing the ovulatory cycle, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . For ster i l ized women, the percent d istr ibut ion by age at the t ime of steri l ization, according to the number of years since the steri l izat ion, SLDHS 1987 . For all current users of contracept ive methods the percent d ist r ibut ion according to most recent source of supply, by method, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 65 65 66 67 67 68 ×i L TABLE 4.17: TABLE 4.18: TABLE 4.19: TABLE 4.20: TABLE 4.21: TABLE 4.22: TABLE 4.23: TABLE 4.24: Among current users of contracept ion obtain ing a method at a source, the percent d istr ibut ion by type of d issat is fact ion with the service (if any), according to type of source last visited, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Percent d istr ibut ion of women who have d iscont inued a method in the last 5 years by main reason for last discontinuation, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Among current ly marr ied exposed women not using contraception, the percent d ist r ibut ion by att i tude toward becoming pregnant in next few weeks, according to number of l iving chi ldren, SLDHS 1987. . . Among women who are exposed and not using contraception, but who would not be happy if they become pregnant, the percent d istr ibut ion by the main reason for nonuse, according to current age; SLDHS 1987 . Amon~ current ly marr ied women aged 15-49 not Current ly using any method of contraception, the percent d ist r ibut ion by intent ion to use in the future, according to nnmher of l iv ing chi ldren ( including current pregnancy), SLDHS 1987 . Among current ly marr ied women not current ly using any method of contracept ion but who intend to use in the future, the percent d ist r ibut ion by preferred method according to whether they intend to use in the next 12 months or after 12 months, SLDHS 1987 . . . . . . . . . . . . . . . Percent d istr ibut ion of ever-marr ied women by whether they have heard a radio message about family planning, aecording tose lec ted background character ist ics, SLDH~ 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Percentage of ever-marr ied womenbe l iev ing that l t is acceptable to have messages about family p lanning on the radio, by age and selected background character ist ics, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 71 72 73 74 75 L 76 77 ×ii TABLE 5.1 : TABLE 5.2 : TABLE 5.3 : TABLE 5.4: TABLE 5.5: TABLE 5.6: TABLE 5.7: TABLE 5.8: TABLE 5.9: Percent d istr ibut ion of current ly marr ied women aged 15-49 by desire for more chi ldren and the certa inty of their preference, according to number of l iv ing children, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . D Percent d istr ibut ion of current ly marr ied women aged 15-49 by desire for more children, according to number of l iv ing children, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . Percent d ist r ibut ion Of current ly marr ied women aged 15-49 by desire for more children, according to age, SLDHS 1987 . Percent of current ly marr ied women who want no more chi ldren ( including ster i l ized women) by number of l iv ing chi ldren and selected background characterist ics, SLDHS 1987 . . . . . . . . . . . . . . . . . . . Among current ly marr ied women, the percent who are in need of family p lanning , and the percent who are in need and plan to use a contracept ive method in the future, by background characterist ics, SLDHS 1987 . . . . . . . . . . . . . . . . . . . Percent d istr ibut ion of ever-marr ied women by ideal number of chi ldren according to number of l iv ing chi ldren (including any current pregnancy), SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mean ideal number of chi ldren for ever-marr ied women by current age and selected background character ist ics, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Percent d istr ibut ion of all b irths ( including current pregnancy) in last five years by contracept ive pract ice of mother and whether birth was wanted, by birth order, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . Among women having a birth in the last 12 months, the percentage want ing a chi ld then, later, or want ing no more children, by birth order, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . 80 81 84 86 87 88 89 90 91 xiii TABLE 5.10: TABLE 6.1 : TABLE 6.2: TABLE 6.3 : TABLE 6.4: TABLE 6.5: TABLE 6.6: TABLE 6.7: TABLE 6.8: TABLE 6.9: Total wanted fert i l i ty rates and total fert i l i ty rates for all women 15-49 for f ive years preceding the survey, by background character ist ics, SLDHS 1987 . Infant and chi ldhood morta l i ty est imates for selected periods of t ime between 1972 and 1987, by sex of child, SLDHS 1987 . Soc ioeconomic d i f ferent ia ls in infant and chi ld morta l i ty for the ten year per iod 1977-1987, SLDHS 1987 . 2 . . . . . . . . . . . . . . Demographic d i f ferent ia ls in infant and chi ld morta l i ty for the ten year per iod 1977-1987, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . Mean number of chi ldren ever born, surviv ing and dead, and percent dead among ever born chi ldren by age of ever-marr ied women, SLDHS 1987 . . . . . . . . . . . . . . . Percent d istr ibut ion of b ir ths in the last 5 years by type of prenatal care received by mother, by background character ist ics, SLDHS 1987 . . . . . . . . . . . . . . . . . . . Percent of births in the 12 months preceding the survey whose mothers received tetanus toxoid injections, by selected background character ist ics, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Percent d ist r ibut ion of births in the per iod 1-59 months pr ior to the survey by type of ass istance mother received during delivery, by selected background character ist ics, SLDHS 1987~ . . . . . . . . . . . . . . . . . Among all chi ldren under 5 years of age , percent with a heal th card; and among chi ldren with a heal th care, percent who have received specif ic immunizat ions and percent who are fully immunized (BCG, DPT3, Pollo3, Measles), SLDHS 1987 . Among all chi ldren 12-23months of age, the percent with a health card, and among chi ldren with a health card, the percent who have received specif ic immunizat ions and the percent fully immunized (BCG, DPT3, Polio3, measles), by background character ist ics, SLDHS 1987 . . . . . . . . . . . . . . . . . . . ×iv 92 95 97 98 99 I00 i01 102 104 105 TABLE 6.10: TABLE 6~ii: TABLE 6.12: Among chi ldren under 5 years of age , the percent having an episode of d iarrhea 24 hours or two weeks preceeding the survey by selected background characterist ics, SLDHS 1987 . . . . . . . . . . . . . . . . . . . Among chi ldren under 5 years of age who had diarrhea in the past two weeks, the percentage consult ing a medical faci l i ty and the percentage receiv ing d i f ferent treatments as reported by the mother, according to background character is t ics , SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Among mothers with chi ldren under 5 , the percent who know about ORS by educat ional level and selected background character ist ics, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 iii 112 TABLE 6.13: TABLE 6.14: TABLE 6.15: TABLE 6.16: Percent d istr ibut ion of chi ldren aged 3-36 months, by standard deviat ion category of he ight- for -age using the internat ional NCHS/CDC/WHO reference population, according to background character ist ics, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Percent d istr ibut ion of chi ldren aged 3-36 months, by standard deviat ion category of weight - for -he ight us ing the internat ional NCHS/CDC/WHO reference population, according to background character ist ics, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Among chi ldren aged 3-36 months, the percent in each height for age standard deviat ion category by each weight for height standard deviat ion category (Waterlow classif ication) using the NCHS/WHO/CDC internat ional reference population, SLDHS 1987 . . . . . . . . . . . . . . . . . Percent d istr ibut ion of chi ldren aged 3-36 months, by standard deviat ion category of weight- for -age using the internat ional NCHS/CDC/WHO reference population, according to background character ist ics, SLDHS 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 117 122 124 ×¥ F IGURE i . i F IGURE 2.1: F IGURE 3.1: F IGURE 3.2: F IGURE 3.3: F IGURE 3.4: F IGURE 3.5: F IGURE 4.1: F IGURE 4.2: F IGURE 5.1: F IGURE 5.2: F IGURE 5.3: F IGURE 6.1: F IGURE 6.2: L IST OF F IGURES Page Map of Sr i Lanka w i th SLDHS zones . . . . . . . . . . . . . . 6 Exposure to the r i sk o f p regnancy among cur rent ly marr ied women aged 15-49 . . . . . . . . . . . . . 26 Ch i ld ren ever born (CEB) and to ta l fe r t i l i ty ra tes . . . . . . . . . . . . . . . . . ~ . . . . . . . ~ . . . . . . . . . . . . . . . . 34 Age spec i f i c fe r t i l i ty ra tes 1975 SLWFS and 1987 SLDHS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Sr i Lanka c rude b i r th ra tes 1971-1985 (Reg is t ra r Genera l ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Perage age spec i f i c fe r t i l i ty ra tes . . . . . . . . . . . . 41 Par i ty p rogress ion ra t ios for 1975 SLWFS, 1982 SLCPS, and 1987 SLDHS . . . . . . . . . . . . . . . . . . . . . 45 Cont racept ive preva lence by method among cur rent ly marr ied women age 15-49 (exc lud ing nor thern and eastern prov inces) . . . . . . . . . . . . . . . . 59 Percent d i s t r ibut ion of cur rent ly marr ied users accord ing to age group and method used 1975 SLWFS, 1982 SLCPS, and 1987 SLDHS . . . . . . . . . 62 Fer t i l i ty p re ferences among a l l cur rent ly marr ied women age 15-49 . . . . . . . . . . . . . . . . . . . . . . . . 82 Fer t i l i ty p re ferences by par i ty among cur rent ly marr ied women age ~5-49 . . . . . . . . . . . . . . 83 ~er t i l i ty p re ferences by age among cur rent ly marr led women . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 In fant mor ta l i ty ra tes over t ime and by res idence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Immunizat ion coverage among ch i ld ren 12-23 months w i th hea l th cards . . . . . . . . . . . . . . . . . . . . . . . 106 xvi FIGURE 6.3: FIGURE 6.4: FIGURE 6.5: FIGURE 6.6: FIGURE 5.7: FIGURE 6.8: Treatment for d iarrhea by res idence among chi ldren under 5 with d iarrhea in the last two weeks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Age of weighed and measured chi ldren . Average height of chi ldren (3-36 months) compared to the internat ional reference . Stunt ing among chi ldren (3-36 months) by sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stunt ing among chi ldren (3-36 months) by educat ion of mother . . . . . . . . . . . . . . . . . . . . . . . . . . . . Re lat ionship between stunting and wast ing among chi ldren 3-36 months . . . . . . . . . . . . . . . . . . . . . i i0 114 119 120 121 123 xv[i I CHAPTER i. Background I.i History, geography, and economy Sri Lanka, an is land nat ion ear l ier known as Ceylon, has a recorded h is tory that stretches back about 2500 years. Since the 15th century, the country exper ienced several foreign invasions and was ruled in success ion by the Portuguese, the Dutch, and the Br i t ish unt i l 1948 when independence was f inal ly gained. S ince independence, Sri Lanka has been pr imar i ly a wel fare state, which has enabled her to make s ign i f icant advances in social development. Of s ingular importance are the ach ievements in educat ion and publ ic health. Sri Lanka's social indicators such as l i teracy rates, infant morta l i ty rates, l i fe expectancy and total fert i l i ty rates are far bet ter than most countr ies in the region or what one might expect on the basis of her nat ional income. A factor of equal s ign i f icance is the gradual d isappearance of gender inequal i t ies and the convergence of regional levels towards nat ional averages. Geographical ly , the is land of Sri Lanka is s i tuated off the southern coast of India between northern lat i tudes 5 ° 55' and 9 ° 50" and Eastern longitudes 79 ° 42' and 81 ° 52'. The island, an e longated pearl in shape, spans an area of 65,608 sqaare ki lometers, s t retch ing a maximum length of 435 k i lometers and a width of 225 ki lometers. Three d is t inct agro-c l imat ic zones can be identi f ied: I) the wet zone which covers the south-western part of the is land and receives the h ighest ra infal l from two monsoons annual ly; 2) the dry zone which covers the ent i re northern and eastern parts of the is land and is dependent for more than ha l f of each year on surface water for irr igation; and 3) the intermediate zone which is a narrow str ip between the wet and the dry zones on the side of the south-western hil ls. It is in the wet zone and the intermediate zone that the main commercia l estate p lantat ions of tea, rubber, coconut, and spices are situated. The tea estates are located largely in the south central h i l ls and the adjacent southern and western low lands; the coconut estates are a long the coastal belt; and the rubber estates go from the coast up to the central hil ls. The dry zone pr imar i ly supports paddy cult ivat ion. Sri Lanka's economy is predominantly agricultural and is heavily dependent on tea, rubber, and coconut for foreign exchange. In recent years, the government has attempted to diversify th ~ economy in order to relieve the heavy dependence on export ~ops . The present government has ado~£ed the policy 0~ an open economy, encouraging foreign investment and promoting the growth of the private sector. A free trade zone has been set up in the suburbs to the north of the capitol city, Colombo. Rural development has also received high priority. One of the most important rural development efforts is the Mahaweli river diversion project which is intended to provide irrigation to the major part of the dry zone agricultural areas. Rural development is also being achieved through the establishment of integrated rural development projects in several of the less developed, largely rural districts. 1.2 Population Sri Lanka possesses a wealth of demographic data upon which a considerable degree of reliance can be placed. The main sources of population information are the periodic censuses which started in 1871 and the sYstem of vital registration which was established in 1887. In addition, a number of sample surveys carried out in the country in recent years provide an abundance of data on demographic, social, and economic characteristics of the people. Two of the most important of these, particularly for purposes of comparison with the Sri Lanka Demographic and Health Survey (SLDHS) are the Sri Lanka World Fertil ity Survey (SLWFS) in 1975 and the more recent Sri Lanka Contraceptive Prevalence Survey (SLCPS) conducted in 1982. As of mid-1987, Sri Lanka had an estimated total population of 16~2 million. The rate of population growth has fluctuated during this century. Prior to World War II, the growth rate was between 0.9 and 1.7 percent. After 1945 and the dramatic decline in the death rate due to malaria control efforts, the growth rate rose to 2.8 percent. During the period from 1963 to 1971 the population growth rate levelled off and then declined to 1.7 percent during the period from 1971 to 1981. This recent decline in growth has been aided by two large emigration streams. One is employment oriented migration mainly to the Gulf countries. The second is the repatriation of Tamils of Indian origin. Had it not been for this net out migration, it has been estimated that the growth rate would have been 2.2 percent rather than the observed 1.7 percent during the decade of the 1970s (Dept . of Census and Statistics, 1985). -- The distribution of population, as in most countries, is highly uneven. In general, more than half of the population is concentrated in the wet zone districts which account for only 20 2 percent of the total land area.' Since about 1930, there has been a shift of population from the high density wet~zone to the low density dry zone. This transfer was very pronounced in the 1970s. Investments in dry zone agriculture, colonization schemes, improvements in the irrigation faci l i t ies, and the erad icat ion of malar ia in the 1940s have probably been contributory factors w~ich aided the population movements between the climatic zones. The fertility levels in Sri Lanka have shown a steady decline since the end of the World War II. The crude birth rate fell from an average of 38.1 per i000 during the period 1946-50 to 28.7 per i000 in 1971-74. In the latter half of the 1970s, however, the crude birth rate showed signs of levelling off, but by the early 1980s the declining trend resumed, bringing the crude birth rate to an estimated 24.3 per i000 by 1985. Mortality levels have also declined over the past four decades. The crude death rate was approximately 22 per i000 in 1945 but then dropped dramatically in the next five years to approximately 13 per I000 in 1950. Since 1950, the death rate has continued to decline but at a slower pace. In 1985 the crude death rate was estimated to be 6.2 per i000. Sri Lanka is a multiracial country which consists of several ethnic groups. According to the 1981 census, Sinhalese constitute 74 percent of the population, Sri Lankan Tamils 12.7 percent, Indian Tamils 5.5 percent, Sri Lankan Moors 7.1 percent, and Burghers, Malays, and others 0.8 percent. Most Sinhalese are Buddhists, most Tamils are Hindus, Moors are Muslims, and Burghers are Christians. 1.3 Family planning policies and programmes The impact of rapid population growth on economic and social development has long been recognized in Sri Lanka. In 1953, the Fami ly P lann ing Assoc ia t ion o f Sri Lanka (FPASL), a nongovernmental organization, was established. By 1955, the Government provided a grant to the FPASL which was increased in subsequent years. Recognizing the importance of providing family planning services to all women in the reproductive ages, the Government entered into an agreement with the Royal Government of Sweden to arrange a pilot project in community based family planning services. This project demonstrated that it was possible to introduce contraceptive methods acceptable to the people. Consequently, the National Family Planning Programme was inaugurated in 1965 with direct involvement of the government and a view to reducing the crude birth rate to 25 per i000 by the year 1976. To coordinate and direct the activities of the national programme, a Family Planning Bureau was established within the Ministry of Health in 1968 and later 3 renamed the Family Health Bureau (FHB) when the government sought to integrate family p lanning with maternal and chi ld care services. The country's family p lanning program was further strengthened in 1973 when the United Nations Fund for Popu la t ion Act iv i t ies (UNFPA) provided assistance to improve the various support ing services essential for the successful implementat ion of the nat ional family p lanning programme. Besides the FPASL, there are several other non-governmental organizat ions involved with the provis ion of family planning services. Among these are the Sri Lanka Assoc iat ion for Vo luntary Surgical Contracept ion (SLAVSC) establ ished in 1974 and the Community Development Services (CDS) establ ished in 1978. A major thrust of the government 's family p lanning program is to educate people about the effects unplanned populat ion growth has on economic development as well as on the qual i ty of l i fe for individuals. It is also the Government 's pol icy to provide a var iety of family p lanning services to couples so that they can select a contracept ive method free from coerc ion or compulsion. In pursuance of this policy, the Government has establ ished 25 Distr ict Populat ion Committees, which are led by the Head of the distr ict administrat ion, and represented by the relevant ministr ies, departments, and field level officers, to imp lement and mon i to r the popu la t ion pol icy in the 25 administ rat ive distr icts of the country. The family p lanning service for which there has been the h ighest demand is steri l ization. The Government supports ster i l i zat ion through an incentive scheme that provides acceptors with re imbursement of incidental expenses, travel costs, and lost income. The acceptance of modern temporary methods has been far lower than steri l ization. The current focus of the Government, therefore, is on developing a system whereby the use of effect ive temporary methods acceptable to the populat ion can be increased. 1.4 Health pr ior i t ies and programmes The health pol icy of the Government of Sri Lanka is to provide heal th care to the entire populat ion so that all people can lead a social ly and economical ly product ive ~ife. Government provides free health care services through pr imary health care faci l i t ies at the per ipheral level and specia l ized and intensive care at leading hospitals. In the last four to five decades, the heal th care system has been expanded and strengthened. A network of publ ic heal th nurses :and midwives provide maternal and chi ld heal th care to thevast major i ty of the rural population. The cur rent emphas is on immunizat ion against common communicable d iseases has been one of the most ef fect ive health programmes launched in recent years. The programme seeks to 4 immunize all chi ldren against tuberculosis, poliomyelitis, diphtheria, whooping cough, tetanus, and measles. Within the integrated package of maternal and child care services, the promot ion of b reast feed ing and the monitor ing of child nutritional status have also assumed high priority. Besides specific health policies and actions, many other social developments have had a bearing on the promotion of health standards in Sri Lanka. The most important of these perhaps was the introduction of free education in the mid 1940s. The ensuing gradual but steady expansion of school enrolment and achievement of higher educational levels has likely been a major factor influencing reproductive behavior and child care practices. Addi t iona l ly , the nationwide food distribution system has provided food security to the population as a whole. In recent years, the highest priority was. given to local production of r ice which has enabled the country to reach self sufficiency in this staple food. Other efforts designed to improve the rural infrastructure, such as construction of roads, community wells, latrines, and housing have affected rural health conditions and prov ided better access to health service outlets. 1.5 Objectives of the survey The Sri Lanka Demographic following object ives: and Health Survey has the i. To provide policymakers and administrators with current and accurate data on fertility, morbidity, family planning and se lected indicators of health status which could be used for planning new strategies for the wellbeing of the population; . 2. To provide data which can be used to analyze trends over time. The SLDHS examines many of the same fertility, mortality, and health issues that were addressed in earlier surveys, most notably the SLWFS and the more recent SLCPS; and 3. To addto the international body of data which can be used for comparative studies. 1.6 Survey sample design On the basis of socio-economic and ecological criteria, and the experience of the SLWFS, nine zones were created. It was felt that some_of the six SLWFS zones were too heterogenous and should be redrawn as shown in Figure i.i and described below: Zone 1 - Colombo Metropolitan area consisting o f SLWFS zone 1 and parts of zone 2. 5 Figure 1.1 - SLDHS SAMPLE ZONES F.S. Zone Jndary tr ict Boundary rues ~e 1 le 2 ~e 3 ~e 4 ~e 5 ~e 6 ~e 7 le 8 le 9 zone 2 - zone 3 - zone 4 - Zone 5 - Zone 6 - Zone 7 - Zone 8 - Zone 9 - Colombo feeder areas and Northern part of SLWFS zone 2. South Western coastal low lands corresponding to Southern part of SLWFS zone 2. Lower South Central hi l l country corresponding to Western and Southern part of SLWFS zone 6, exc luding d istr icts with a concentrat ion of estates. South Central hi l l country corresponding to part of SLWFS zone 5 with a concentrat ion of estates. I r r igated Dry Zone corresponding to SLWFS zone 3, with major or minor i r r igat ion schemes. Rain fed Dry Zone cover ing the rest of SLWFS zone 3. Eastern Coastal Belt, corresponding to SLWFS zone 4 (not included in SLDNS). Northern Province corresponding to SLWFS zone 5 (not included in SLDHS). The changes SLDHS made to the SLWFS zones were designed: a) to separate the Co lombo urban feeder areas from rural h inter lands; b) to separate rural areas with predominant ly estate populat ions from other rural areas; and c) to d ist inguish between irr igated dry zone areas which are new sett lements under development projects from those areas which rely pr imar i ly on rains for cult ivation. A l though the survey or ig inal ly p lanned to conduct interviews in all nine zones, Civil d isturbances in zones 9 and 8 (the Northern and Eastern provinces) prevented interviews from being conducted there. These zones, which contain approx imate ly 14 percent of the 1986 est imated populat ion of Sri Lanka, have been exc luded from the SLDHS. With the except ion of zone 5, the sample was al located equal ly between zones with an est imated target 900 completed individual interviews per zone. Zone 5 was given a larger target sample size of 1,350 to permit over sampl ing of the estate p lantat ion workers. In principle, the sample was des igned to cover pr ivate households in the areas sampled. The populat ion residing in inst i tut ions and inst i tut ional households was excluded. For the deta i led individual interview, the e l ig ib i l i ty cr i ter ia were: ever-marr ied women aged 15 through 49 who slept in the household the previous night. For the select ion of area units, the sample frame was based on block stat ist ics from the 1981 Census of Populat ion and Housing. However, these f igures were updated where poss ib le on the basis of the work done in connect ion with a 1985-86 labour force survey. This appl ied in part icu lar to newly sett led areas with the development of i rr igat ion schemes in the dry zone. For the final select ion of housing units with in ult imate area units, a special operat ion was undertaken before the survey to update household l ists with in selected census blocks. The zones created by the SLDHS, which were designed to capture re lat ively homogeneous subgroups of the population, served as the pr imary strata. Each zone was further strat i f ied into (up to) three strata: urban, rural, and estate areas. Further impl ic it s t rat i f icat ion was achieved by order ing the sampl ing areas according to administrat ive and geographical location. Simi lar systematic sampl ing procedures were fol lowed at all stages up to and including the select ion of housing units. The sampl ing of housing units was undertaken in two or three stages depending upon the stratum. In densely populated zones i, 2, and 3, and in urban strata of all zones a three stage design was used: i. At the first stage, a strat i f ied sample of Gram Savaka or equivalent areas (waras or estates) with probabi l i ty proport ional to size (PPS) was selected. The number of pr imary sampl ing units (PSIs) selected was 54 in zones 5 and 36 in each of the other zones. Within a given zone, the number to be selected in a stratum was a l located proport ionate ly to the strata populations. 2. With in each PSU, two census blocks were selected with PPS, systemat ica l ly without replacement. 3. The final stage consisted of the select ion of the housing units in selected blocks with inverse PPS so as to yield a self we ight ing sample within each stratum. For the main survey, there was no further sampl ing as all e l ig ib le women in each selected housing unit were taken into the sample. Also, for the anthropometr ic measurements, all chi ldren 3 through 36 months of e l ig ib le women were taken. In the non-urban strata in zones 3 through 7, the only d i f ference in procedures was that general ly only one block was selected per PSU. This procedure ef fect ive ly reduced the number of stages to two: blocks as the first stage and housing units as the second stage. Since zones were al located general ly uni form sample sizes, the overal l sampl ing fract ions var ied in inverse proport ion to 8 the zone population. The overal l sampl ing fract ions were as fol lows for the var ious zones: Zone Est imated 1986 populat ion in thousands Overal l sampl ing fract ion x 103 1 1600 4.3 2 1830 3.5 3 2290 2.8 4 2700 2.4 5 estates 680 7.0 Other strata 2030 2.3 6 1160 5.5 7 1630 3.9 The overal l average sampling fract ion is 3.95 per thousand. The sampl ing weights appl ied in the analysis are inversely proport ional to the above fractions, with a further modi f icat ion for household and individual non-response. These weights are appl ied to all data presented in this report and are shown in Table i.i. The sample was designed to be se l f -weight ing with in strata as well as across strata within a given zone (with the except ion of the estate area in zone 5 which was over sampled). 1.7 Survey organizat ion and field work The Demograph ic and Health Survey in Sri Lanka was implemented by the Department of Census and Stat ist ics (DCS) of the Min is t ry of Plan Implementat ion in close co l laborat ion with the Inst i tute for Resource Development (IRD) /West inghouse, USA. Funding for the survey was provided by the U.S. Agency for Internat ional Development. The survey organizat ion in Sri Lanka was headed by the Director of the Department of Census and Statist ics. A Deputy Director of the Department funct ioned as Project Manager and an Ass istant Director as Ass istant Project Manager. Other DCS staff including two statist ic ians, stat ist ical support ing staff, and adminis£rat ive and secretar ial staff worked dur ing var ious stages of the survey. IRD/West inghouse provided technical co l laborat ion through per iod ic consu l tant vis i ts concerned with sample selection, quest ionnaire design, anthropometr ic measurements, 9 TABLE 1.1: Household and individual weights applied to SLDHS data, SLDHS 1987. Inverse of Adjustment Adjustment Scaled Seated SLDHS sampling for household for individuat individual household zones strata fraction non-response non-response ueight weight zone 1 Urban 114.3 9561887 7031642 0.858937 0.810151 Zone 2 Urban 113.5 1421137 1131107 0.978745 0.957194 Rural 1/3.5 117':,1125 8561814 0.982068 0.964534 Zone 3 Urban 1/2.8 111/111 84180 1.173562 1.154363 Rural 112.8 8631848 6011594 1.150853 1.174782 Estates 1 /2 .8 49146 37137 1.190571 1.2296{,7 Zone 4 Urban 112.4 62/62 51/51 1.303958 1.346757 Rural 112.4 1125/1103 904/889 1.352407 1.373618 Estates 112.4 58/57 44144 1.326835 1.370384 Zone 5 Urban 112.3 83181 57154 1.471707 1.440010 Rural 1/2.3 7691724 5701533 1.5455/8 1.492658 Estates 117.0 6631651 5811578 0.456756 0.470257 Zone 6 Urban I /5 .5 85/8] 62/61 0.592263 0.601836 Rural 115.5 8001771 620/591 0.619373 0.609780 Zone 7 Urban 113.9 27127 16/16 0.802436 0.828773 Rural 1/3.9 10011955 8061774 0.875861 0.868693 data processing, and analysis. The f ield staff for the survey were drawn from among Stat ist ica l Invest igators of the DCS, the major i ty of whom had field exper ience in previous fert i l i ty surveys such as the SLWFS and SLCPS. Nine survey teams were formed. A team consisted of four to seven female interviewers and a supervisor for a total of 48 interviewers and 9 superv isors (5 female and 4 male). In addition, 18 measurers col lected length and weight data from chi ldren 3 through 36 months of age. A measurer and assistant measurer accompan ied each team of in terv iewers . The interviewers, supervisors, and measurers were tra ined in January 1987 for I0 days. The measurers were given a standardizat ion test by an exper ienced nutr i t ionist immediately fo l lowing their t ra in ing and again midway through field data col lect ion. The tra in ing for the interviewers was conducted over a period of two weeks and consisted pr imar i ly of role playing sessions and other exercises deigned to famil iar ize them with the questionnaire. In addition, each interviewer completed at least three pract ice interviews in the field. All the supervisors had previousl) been involved with the SLWFS and/or the SLCPS. Among i0 the interviewers, about three-fourths had exper ience conduct ing interviews for SLWFS and/or SLCPS. During the f irst week of data col lection, all new interviewers were teamed wi th an exper ienced interviewer and two quest ionnaires were completed for each respondent. The quest ionnaire from the new interviewer was then compared against the quest ionnaire completed by the exper ienced person and necessary correct ions made. This procedure provided addit ional on the job training for new interviewers and helped to bui ld their confidence. The tra in ing of heights and weights measurers and assistants was carr ied out with special care as this was the f irst exper ience with anthropometr ic measurements for the survey staff. The tra in ing was conducted by a specia l ist from IRD West inghouse and fo l lowed guidel ines developed by the Uni ted Nat ions (United Nat ions Department of Technical Co-operat ion for Development Stat ist ical office, New York 1986). The 18 measurers and assistants part ic ipated in the relevant t ra in ing sessions for the interviewers. In addition, for over a week, they received pract ical instruct ion outside the c lassroom in the techn iques of tak ing accurate height and weight measurements of children. Pract ice sessions were conducted at two chi ldren's homes in Colombo. At the complet ion of the tra in ing per iod all measurers were given a test which consisted of all t ra inees and the supervisor measur ing a group of i0 ch i ldren twice. The measurements were graded for prec is ion (the abi l i ty to record the same measurement twice on the same child) and for accuracy (the c loseness of the reading to that of the supervisor). One of the measurers fai led to pass the test and was ass igned to other survey duties. To ensure that standards were mainta ined throughout the f ield work, the test was readministered midway through f ield work. Analys is of the test data indicated that 55 percent of the measurers overest imated length, 29 percent underest imated and 16 percent showed no var iat ion from the supervisor 's measurements. The average error among those who underest imated length was 2.7 mil l imeters. The average error among those who overest imated length was 1.9 mil l imeters. During the test, 91 percent of the chi ldren were measured to with in hal f a cent imeter of the supervisor 's measurement. For the purpose of coordinat ing the f ield work and implement ing qual i ty control procedures, f ive f ield coordinators were appointed. The coordinators v is i ted the interviewers in the f ield and reviewed quest ionnaires completed by each interviewer both at the beginning and midstream of f ie ld work. Data col lect ion began January 18, 1987, and was essent ia l ly completed by the end of March except for a few areas which began late and were completed by May. Each quest ionnaire was edited in the f ield dur ing the evening fol lowing the interview. In ii addition, all quest ionnaires were further reviewed and edited in Colombo. Data were entered onto microcomputers start ing just two weeks after the commencement of field work. The ISSA (Integrated System for Survey Analysis) software package of programs developed by IRD/West inghouse was used for data entry, machine editing, and tabulation. An especia l ly effect ive procedure for correct ing errors and inconsistencies detected dur ing off ice edit ing and data entry was to relay information about problems in a quest ionnai re to the interviewers whi le they were stil l in the field. In most cases the problem could be corrected by going back to a respondent. As a result of this procedure, field errors d imin ished considerabiy With t ime ~hd the volume af edi t ing was reduced to a minimal level. Al l survey act iv i t ies were completed on schedule and within a re lat ive ly short per iod of t ime from October 1986 to March 1988. The calender of act iv i t ies is shown below: Survey design Quest ionnai re design and t rans lat ion Preparat ion of manuals Pretest of quest ionnaires Select ion of f ield staff Quest ionnai re and manual pr int ing Supervisor, interviewer, and measurers t ra in ing Field work Data process ing and edit ing Tabulat ions Preparat ion of pre l iminary report Preparat ion of draft final report Seminar on pr incipal f indings Preparat ion of final report Oct 1986 Nov 1986 Oct to Nov 1986 Nov 1986 Dec 1986 Dec 1986 Jan 1987 Jan to Apri l 1987 Feb to May 1987 Ju ly to August 1987 August 1987 September to Nov 1987 November 1987 Dec 1987 to March 1988 12 TABLE 1.2: Percent d i s t r ibut ion of the female popu la t ion by age f rom al l women in the 1987 SLDHS househo ld sample, and f rom the 1984 es t imated female popu la t ion for a l l Sr i Lanka. AGE GROUPS PERCENT D ISTRIBUT ION OF FEMALES A l l Sr i Lanka 1984 es t imate* 1987 SLDHS Househo lds 00-04 12.5 9.7 05-09 11.4 11.5 10-14 ii. 3 ii. 0 15-19 10.9 10.7 20-24 10.4 i0 .0 25-29 8.7 8.4 30-34 7.6 7.5 35-39 5.7 6.8 40-44 4.6 5.1 45-49 4.1 3.6 50-54 3.6 4.5 55-59 2.8 3.3 60-64 2.2 2.3 65+ 4.2 5.5 Percent to ta l i00 i00 N = 7 ,648 ,000 20 ,232 .4** * Source: Depar tment of Census and Stat i s t i cs , S ta t i s t i ca l Abst rac t of the Democrat i c Soc ia l i s t Repub l i c of Sr i Lanka, 1985, Co lombo, Sr i Lanka, 1987. ** We ighted number of females in SLDHS househo lds . 1.8 The SLDHS quest ionna i re and respondents The Sr i Lanka Demograph ic and Hea l th Survey used two quest ionna i res each of wh ich was pretes ted . The f i rst , ca l led the Househo ld Quest ionna i re , was used to l i s t a l l usua l househo ld members and any v i s i to rs who s lept in the househo ld the preced ing n ight . For each person l is ted, in fo rmat ion on age, sex, and mar i ta l s ta tus and whether or not he /she s lept in the househo ld the prev ious n ight was recorded. F rom th is l i s t e l ig ib le respondents were se lec ted for in terv iew. An e l ig ib le respondent is de f ined as a woman cur rent ly marr ied , d ivorced , separated , or w idowed between the ages of 15 and 49 who s lept in the househo ld the prev ious n ight . Tab le 1.2 compares the SLDHS age d is t r ibut ion of a l l women ident i f ied through the Househo ld Quest ionna i re w i th the age 13 dis t r ibut ion of the 1984 est imated female popu lat ion in Sri Lanka. Among women aged 15 through 49, the SLDHS age d is t r ibut ion is a lmost identical wi th the all Sri Lanka age d is t r ibut ion upto age 35. In the age categor ies 35-39 and 40-44, the percent of women in the SLDHS sample is h igher than the percent in the all Sri Lanka 1984 estimate, but in the category 45-49 the percent of women in the SLDHS sample is s l ight ly lower. The second or Indiv idual Quest ionna i re was each e l ig ib le respondent. On the average, interv iew took approx imate ly 35 to 40 minutes. Quest ionna i re cons isted of nine sections: admin is tered to an individual The Individual i. Respondents background 2. Bir th h is tory - dates of all l ive b i r ths and infant and chi ld deaths 3. Contracept ion - knowledge, ever use, current use and a deta i led h is tory of inter b ir th use in the last 5 years 4. Chi ld hea l th - immunizat ion status, ep isodes of diarrhea, breast feeding, the use of supplementary foods, prenata l care, and ass is tance at de l ivery 5. Marr iage and migrat ion 6. Fer t i l i ty preferences 7. Husband's background and respondent 's work 8. Soc io -economic indicators 9. Length and weight - measurements of all ch i ldren 3 through 36 months. More than in s imi lar fert i l i ty and family p lanning surveys conducted in the past, the SLDHS devoted cons iderab le t ime and at tent ion to obta in ing informat ion on the heal th status of mothers and chi ldren. In addi t ion to many hea l th re lated questions, anthropometr ic length ~nd weight measurements were taken on all ch i ldren 3 months through 36 months. The resul ts from the Household and Indiv idual Quest ionnai res are shown in Table 1.3. A total of 8,119 households were l isted in seven zones. Not a l l of these turned out to be val id househo lds in the f ie ld and the complet ion rate (response rate) at the househo ld level was 96.3 percent. From these households, 6,170 e l ig ib le respondents were ident i f ied and interv iews were completed among 5,865 ever -marr ied women for a response rate of 95.1 percent. 14 It is important to note once again that the districts in the northern and eastern portions of the country were not covered by the SLDHS because of civil disturbances. Whenever comparisons are made between the SLDHS and the earlier SLWFS and SLCPS, the differences in areas covered by the surveys should be kept in mind. 1.9 Background Variables The survey 's pr imary f indings concern ing fertility, contraception, fertility preferences, child mortality, and health are presented in the following chapters. The major background TABLE 1.3: ResuLts of 5LDRS household and ind iv idua l interviews by sector, SLDHS 1987. ALL Other sampled urban Rural ResuLt of Interv iew areas CoLombo areas areas Estates HousehoLd Interviews CompLeted household interviews 94.5 91.4 95.4 94.7 95.7 No competent respondent at home 1.4 2.5 1.1 1.4 0.4 Head of household absent 0.4 0.2 0.2 0.3 0.9 Refused in terv ieu 0.04 0.2 0.0 0.0 0.1 *D~ett ing vacant 1.3 1.0 1.9 1.1 2.5 *DueLling destroyed 0.6 0.5 1.0 0.6 0.0 DweLLing not found 0.5 1.1 0.2 0.5 0.4 Other 1.3 3.1 0.2 1.3 0.0 TOTAL PERCENT 100 100 100 100 100 HOUSEHOLD RESPOMSE RATE 96.3 92.8 98.2 96.3 98.2 UNWEIGHTED NLMBER OF HOUSEHOLDS 8,119 972 525 5,8.~ 788 Individual lnterv|ews CompLeted ind iv idua l interviews 95.1 89.9 95.8 95.2 99.6 Res~-~nt not at home 3.6 7.1 3.6 3.5 0.5 Refused ln terv ieu 0.2 1;3 0.0 0.1 0.0 PartLy completed 0.1 0.1 0.0 0.1 0.0 Other 1.1 1.5 0.5 1.2 0.0 TOTAL PERCE~T 100 100 100 100 100 TOTAL UNMEIGHTED NUMBER OF ELIGIBLE WONEN 6,170 714 385 4,409 662 * DweLLing vacant and d~tt ing destroyed are not included in the caLcuLations the response rate. 15 var iables used throughout this zone, and education. Each descr ibed below: report are age, parity, sector, of these var iab les is br ief ly AGE. Age is computed from the year and month of birth as reported by the respondent. Among all women interviewed, 90 percent were able to give both the year and month of birth and this date was almost always supported with some documentary evidence such as a National Identity Card or occas ional ly a birth cert i f icate. Six percent of the women were able to give only the year of birth but not the month and 4 percent were unable to give e i ther the year or month of birth. Whenever the year or month of b i r th is missing, an imputed age has been assigned. PARITY. The informat ion on par i ty was col lected using a series of quest ions to ascertain the number of chi ldren l iving with the mother, l iv ing elsewhere, or dead. Each series of quest ions was asked separately for sons and daughters. The interv iewer tota l led the chi ldren in all 3 categor ies and then ver i f ied with the woman whether the total was correct. Any d iscrepancy was resolved immediately through probing and by checking against the birth history. SECTOR. The var iable sector identi f ies the respondent's place of residence. Four categor ies were used: Colombo metropol i tan, other urban," rural, and estate. The metropol i tan Colombo area was separated from the other urban areas because of its d is t inct ion as the capitol city. An urban area is one which is def ined for administrat ive purposes as being under the admin is t rat ion of a Municipal Counci l or an Urban Council. Estates are the p lantat ions of tea, rubber and coconut. The remaining areas fall into the rural category. ZONE. Init ial ly, nine fair ly homogeneous geographical zones were created for the purpose of provid ing est imates in the SLDHS. Because of civi l disturbances, only seven zones could be included in the survey. EDUCATION. The level of educat ion was def ined in terms of the grades completed in school. Four major groups were identif ied: no school ing or never attended formal school, pr imary (grades 1 through 5), secondary (grades 6 through 9), and more than secondary (grades i0 and above). The re lat ionship between educat ion of respondent and the other six background var iables is shown in Table 1.4. A unique achievement of Sri Lanka is the re lat ively high proport ion of women who have some educat ion compared with other countr ies in the South Asia region. Only ii percent of the SLDHS sample of ever-marr ied women 15-49 have no education. Many of these women are on the estates where about 44 percent have no education. In other rural areas, approximately i0 percent of the ever-marr ied 16 TABLE 1.4: Percent d i s t r ibut ion of ever -marr ied women aged 15-49 by (eve( of educat ion according to se lected background character i s t i cs , SLDHS 1987. EDUCATION At( BACKGHQUND More CHARACTER- No then Percent Weighted Ur~eighted ISTICS Education Primary Secoc~lery Secondary Tote( Number Number AGE 15-19 15.0 24.1 54.8 5.3 100.0 133.9 139.0 20-24 11.6 29.9 36.0 22.6 100.0 723.3 723.0 25-29 7.7 26.7 41.1 24.5 100.0 1112.7 1126.0 30-34 8.4 25.8 40.4 25.5 100.0 1203.4 1185.0 35-39 9.8 28.8 35.3 26.0 100.0 1130.7 1133.0 40-44 13.9 34.4 30.4 21.3 100.0 921.9 925.0 45-49 19.8 37.6 26.1 16.5 100.0 638.7 634.0 PARITY 0 10.4 20.4 37.5 31.7 100.0 517.0 510.0 1 7.8 19.1 38.7 34.5 100.0 1042.9 1038.0 2 8.9 24.2 37.7 29.1 100.0 1340.2 1344.0 3 8.8 30.8 39.0 21.4 100.0 1165.6 1167.0 4 + 16.7 41.8 31.4 10.1 100.0 17~B.8 1806.0 SECTOR Cot omba C i ty 5.3 19.2 45.5 30.1 100.0 551.4 642.0 Other Urban 6.4 20.1 41.1 32.4 100.0 393.6 369.0 Rurat 9.7 30.9 36.6 22.7 100.0 4553.0 4195.0 Estates 43.6 40.4 11.3 4.7 100.0 366.4 639.0 ZONE zone 1 5.3 19.2 45.5 30.1 100.0 551.4 642,0 Zone 2 3 .7 19.1 45.9 " 31.3 100.0 904.1 921.0 Zone 3 9.2 31.5 35.0 24.3 100.0 821.5 711.0, Zone 4 10.2 32.4 35.5 21.8 100.0 1327.2 984.0 Zone 5 21.6 30.4 28.4 19.7 100.0 1167.3 1165.0 zone 6 10.5 33.9 33.7 21.9 100.0 402.2 652.0 zone 7 12.9 40.6 33.5 13.0 100.0 690.8 790.0 Tote( 11.2 29.7 36.2 22.9 100.0 5864.5 5865.0 women have no education. Table 1.5 provides further information on basic background var iables and also shows the weighted and unweighted number of respondents for each category of the variable. With the except ion of religion, these background var iab les are used throughout this report. Rel ig ion has not been used because the exclus ion of the northern and eastern prov inces resulted in an under representat ion of Hindus and Muslims. 17 TABLE 1.5: Percent d i s t r ibut ion of ever -marr ied women by se lected background character is t ics , SLDHS 1987. BACKGROUND CHARACTER- ISTICS NUmber of Respondents Percent Weighted Unweighted AGE 15-19 2.3 13i.9 139.0 20-24 12.3 72!.3 723.0 25-29 19.0 1112.7 1126.0 30-34 20.5 1203.4 1185.0 35-39 19.3 1130.7 1133.0 40-44 15.7 921.9 925.0 45-49 10.9 638.7 634.0 PARITY 0 8.8 siZ.o 510<0 1 17.8 1042.9 1038.0 2 22.9 1340.2 1344.0 3 19.9 1165.6 1167.0 "4 12.8 748.0 761 .0 5 8.5 496.4 487.0 6+ 9.5 554.2 558.0 EDUCATION No educat ion 11.2 657.2 737.0 Pr imary 29.7 1739.1 1777.0 Secondary 36.2 2122.5 2062.0 More than set. 22.9 1345.6 1289.0 REL IG ION Buddh is t 81.9 4802.5 4554.0 H indu 5.8 338.1 546.0 Mus l im - 4.8 280.4 259.0 Catho l iC 6.7 390.2 448.0 Other Chr i s t ian 0.8 45.3 51.0 Other ~.i 5.6 5.0 SECTOR Co lombo 9.4 551.4 642.0 Other Urban 6.7 393.6 369~0 Rural 77.6 4553.0 4195.0 Estates 6~2 366.4 659.0 ZONE Zone 1 9.4 551.4 642.0 Zone 2 15.4 904.1 921.0 Zone 3 14.0 821.5 711.0 Zone 4 22.6 1327.2 984.0 Zone 5 19.9 1167.3 1165.0 Zone 6 6.9 402.2 652.0 Zone 7 11.8 690.8 790.0 Total 100 5864.5 5865.0 18 CHAPTER 2. Nuptiality And Exposure to Risk of Conception 2.1 Current marital status Current marital status is of basic importance in its own r ight and in the determinat ion of many fer t i l i ty and contraceptive prevalence indicators discussed later in this report. The denominator for some of these indicators is all ever-married women, while for Others it is currently married women. The SLDHS identified five categories of current marital statds: currently married and living together, currently married but not living together, widowed, divorced, and separated. The term married is intended to mean lega l or customary union. Living together outside such legal or customary marriages is socially unacceptable and is almost non-existent in Sri Lanka; hence, there is no marital status category such as "living together" that might be found in other countries. "Currentl# married" are those women who are married and their~husbands are at present either living with them or are temporarily away for reasons such as employment. Since short term spousal separations have been noticed in recent years due to migration abroad and to other parts of the country, SLDHS identified two categories of currently married; one "living with the husband," and the other "not living together with the husband.~ Divorce is Complete dissolution of marriage, while separation is legal separation which does not pe~i t either party to remarry. The current marital status distr ibut ion of the SLDHS sample is given in Table 2.1. O f the total sample, 92.8 percent are currently married and the proportion of women whose husbands are away is only 1.4 percent. Marital dissolution by divorce is extremely low at 0.3 percent, but separation is higher at 3.2 percent and almost as high as that caused by widowhood which is 3.8 percent. Widowhood increases with age to a level of 12 percent among women 45-49 years. Divorce too tends to occur at older ages past 30 years but within any age group is less than 1 percent. 19 TABLE 2.1: Percent d i s t r ibut ion of ever-married ~a~en aged 15-49 according to age by current marital status, SLDHS 1987. CURRENT ~R%TAL STATUS Current ly CurrentLy married married CURRENT t iv ing not t |v|ng Percent Weighted AGE together together Widowed Divorced Separated to ta l N 15-19 93.1 2.1 1.2 0 .0 3.6 100 134 20-24 95.1 0.9 1.1 O. 1 2.7 100 723 25-29 94.8 1.4 1.0 0.0 2.8 100 1113 30-34 93.8 1.4 2.0 0.3 2.5 100 1203 35-39 90.4 1.9 3.6 0.2 3.9 100 1131 40-46 88.3 1.0 6.6 0.7 3.6 100 922 45-49 82.1 1.8 11.9 0.3 3.9 100 639 Totat 91.4 1.4 3.8 0.3 3.2 100 5~65 Separat ion tends to be s l ight ly h igher among older age groups but l ike divorce, the di f ferences between age groups are not outstanding. 2.2 Age at f irst union The age at first marr iage is an important social event and demographic indicator of fert i l i ty since chi ld-bear ing in Sri Lanka takes p lace predominant ly within marriage. A r is ing age at marr iage was one of the most important determinants of the fert i l i ty decl ine that commenced in Sri Lanka in the 1950s. One approach to ident i fy ing trends in age at marr iage from the SLDHS survey data is to examine the exper ience of di f ferent cohorts. For all women in the sampled households (ever marr ied as wel l as never marr ied women), Table 2.2 presents the percent d ist r ibut ion by age at f irst marr iage according to current age. Also presented in this table is the median age at f irst marr iage for cohorts aged 25 and older. The median is not shown for the two youngest cohorts because in each of these age groups over f i fty percent of the women have not been married. Table 2.2 suggests that women who are current ly between the ages of 25 and 40 marr iedat an older age than women who are current ly aged 40 and above. For example, the median age at f irst marr iage among younger cohorts is approximately 23 years. This is two to three years higher than the median age at first marr iage among women who are current ly 40 and above. However, the median for the older cohorts needs to be v iewed with some caut ion because these women may tend to underreport their age at marriage. A summary measure der ived from Table 2.2 that is useful for compar ing cohort trends is the percent of women who have marr ied 20 by a certain age in all cohorts who have passed that age. For example, the proport ion of women marr ied by age 20 in the age cohorts 20-24, 25-29, 30-34, 35-39 has remained approx imate ly the same at 30 percent. On the other hand, among the cohorts aged 40-44 and 45-49, the proport ion marr ied by age 20 is much higher at 41 and 50 percent respectively. Since age at marr iage in Sri Lanka tends to be high, it is more meaningful to examine the proport ions marr ied at an older age. ~hen the pivotal age is shifted to 25, the proport ion marr ied before this age decl ines steadi ly from 76 percent for the 45-49 group to 61 percent for the 25-29 age cohort. More women in the younger cohorts appear to have delayed marr iage than women in older cohorts. Another way of est imat ing trends in the age at marr iage is by compar ing data from earl ier censuses and surveys. Table 2.3 suggests that the age at marr iage has increased in recent years. The table shows the percent of women who have ever been marr ied TABLE 2.2: Percent d i s t r ibut i c¢ l of a l l ever-marr ied and never mrr ied wo~en (from household schedule)* according to current age by age at f i r s t union and median age at f i r s t un ion according to cur rent age, SLDHS 198Z. AGE AT FIRST UNION CURRENT Never < 15 15-17 18-19 20-21 22-24 25-27 28-29 30+ Total Weighted Median AGE married Percent N Age ** 15-19 92.7 0.6 5.1 1.6 0.0 0.0 0.0 O.O 0.0 100 1824 *** 20-24 57.1 1.1 12.8 14.3 10.5 4.3 0.0 0.0 0.0 100 1685 *** 25-29 30.0 2.9 11.5 15.9 14.6 16.0 8.5 0.7 0.0 100 1589 23.1 30-3G 1&.2 2.3 11.& 15.2 16.8 18.1 12.2 6 .0 3 .8 100 1403 22.6 35-39 9.1 4.2 13.4 12.6 12.8 20.0 12.7 7.1 8.1 100 1243 23.1 40-44 6.2 8.0 18.4 14.3 13.5 16.3 11.1 3.8 8.4 100 983 21.4 45-49 3.5 11.4 22.1 16.3 13.1 13.2 9.0 5.0 6.4 100 662 20.0 Total 37.5 3.4 12.2 12.2 10.9 11.5 6 .7 2.7 3.0 100 9389 *** * A l l women (taken from the household schedule) inc ludes women ever-marr ied, cur rent ly marr ied, and never marr ied. ** Defined as the age by b~hich one-ha l f of a l l women have ever-married. *** Omitted due to censor ing. among five year age cohorts for the censuses of 1963, 1971, and 1981 and the 1975 SLWFS and 1987 SLDHS. The table also shows the s ingulate mean age of marr iage (SMAM). The SMAM is a per iod rather than a cohort measure. It can be interpreted as the mean age of marr iage among women who marry before the age of 50. The s ingulate mean age at marr iage increased from 22 years in 1963 to 25 years in 1975. Since 1975, the SMAM has remained re lat ively constant at about 25 years. S imi lar ly the proport ion marr ied among the three youngest cohorts aged 15 though 29 decl ined between the 1963 census and the 1975 SLWFS but then increased s l ight ly or at least level led off in the 1980s. 21 TABLE 2.3: Proport ion of all women in' f ive year age groups who have ever-marr ied, and s ingulate mean age at marr iage (SMAM), 1963 census, 1971 census, 1975 SLWFS, 1981 census, and 1987 SLDHS. PROPORTION EVER-MARRIED AGE 1963 1971 1975 1981 1987 GROUP Census Census SLWFS Census SLDHS 15-19 14.8 10.6 6.8 9.9 7.3 20-24 57.6 46.8 39.4 44.7 42.9 25-29 81.0 75.4 68.1 69.6 70.0 30-34 88.6 89.1 86.3 84.2 85.8 35-39 89.8 94.2 94.2 91.1 90.9 40-44 86.1 95.3 95.4 94.1 93.8 45-49 81.6 95.9 97.9 95.5 96.5 SMAM 22.1 23.5 25.1 24.4 24.8 Trends and d i f ferent ia ls in the median age at marr iage in 1987 among al~ women are shown in Table 2.4. The h ighest median age at marr iage is 26.5 years among the cohort aged 25-29 l iv ing in urban areas other than Colombo. Indeed, across all age groups, women l iv ing in other urban areas have the h ighest median age at marr iage, except for the age group 45-49. Conversely, the lowest median age of marr iage for all cohorts is among women l iv ing on the estates and women in Zones 6 and 7. 2.3 Current exposure status Women who are current ly exposed to the r isk of pregnancy are the potent ia l c l ient group for family p lanning programs. These women const i tute the denominator used by many family p lanning programmes to es t imate measures such as unmet need for contracept ion. The SLDHS def ined "exposed to the r isk of pregnancy" as women who: a) menst ruated in the last s ix weeks and b) had sex in the last four weeks and c) have an open birth interval of less than f ive years or, if longer, d) used cont racept ion dur ing the interval. In other words, these are women who are sexual ly active, presumably ovulating, and presumably fecund. This is a more r igid def in i t ion than adopted in prev ious fert i l i ty and contracept ive surveys which c lass i f ied women as "exposed" so le ly on the bas is of the woman's percept ion of her ab i l i ty to conceive. The SLDHS inc luded more quest ions re lated to exposure which enabled the co l lect ion of greater detai l regard ing a woman's abi l i ty to conceive. These quest ions focussed on the durat ion of t ime s ince last sexual intercourse 22 and s ince las t menst ruat ion . In add i t ion , comprehens ive prob ing on the use of cont racept ion dur ing the las t 5 years was under taken . The data f rom the SLDHS on exposure are presented in Tab le 2.5. The categor ies in th i s tab le a re h ie rarch ica l in the order p resented . That is, f i rst , p regnant women are se lec ted f rom among a l l cur rent ly marr ied women aged 15-49. Nonpregnant women are then checked to see whether they are amenorrhe ic . Those not amenorrhe ic a re then checked to see if they are in fecund (i.e. TABLE 2.4: Med ian age at f i r s t un ion among a l l ever -marr ied women ( f rom househo ld schedu le )* aged 25 - 49 years by sector and zone, SLDHS 1987. BACKGROUND CURRENT AGE CHARACTER- A l l Ages IST ICS 25-29 30-34 35-39 40-44 45-49 25-49 SECTOR Co lombo c i ty 24 .6 24 .0 24.5 22 .5 21 .9 23 .8 Other Urban 26.5 25.2 24.9 24.8 21.4 24 .9 Rura l 23 .1 22.4 22 .9 20 .9 20 .0 22 .1 Es ta tes 20.4 21 .6 20.9 20 .1 19 .1 20 .6 ZONE Zone 1 24 .6 24 .0 24 .5 22 .5 21 .9 23.8 Zone 2 23.9 22.9 24.4 23.3 23 .6 23.7 Zone 3 24 .1 25 .0 24.8 24 .8 23 .0 24 .6 Zone 4 23 .1 22.6 22.7 19.1 19.4 21.7 Zone 5 22 .6 22.5 21.9 21.2 19.1 21.8 Zone 6 21.7 21.5 21 .1 18.8 18.2 20 .7 Zone 7 21.3 20.3 20.5 18.6 17.4 19.9 A l l Groups 23 .1 22 .6 23 .1 21.4 20 .0 22 .4 * A l l women ( taken f rom the househo ld schedu le ) inc ludes women ever -marr ied , cur rent ly marr ied , and never marr ied . 23 TABLE 2.5: Percent d i s t r ibut ion of cur rent ly marr ied women aged 15-49 by exposure to concept ion s tatus accord ing to cur rent age, SLDNS 1987. EXPOSURE TO CONCEPT ION STATUS CURRENT AGE 15-19 20-24 25-29 30-34 35-39 40-44 45-49 TOTAL Cur rent ly 30.8 16.0 10.6 7.0 2.5 0.7 0. i 6.9 pregnant Amenorrhe ic 16.0 17.5 14.6 9.3 5.2 2.2 0.i 8.8 In fecund * 0.0 0.9 2.6 6.0 9.2 16.4 32.9 9.4 No sex in 5.8 9.5 9.9 10.3 13.3 21.0 26.0 13.8 last 4 weeks No per iod 0.0 4.3 3.4 1.4 1.9 2.7 9.4 3.2 in last 6 wks. EXPOSED WOMEN** Us ing s ter i l - 1.0 5.0 16.3 29.1 35.1 30.8 15.7 22.9 i zat ion Us ing modern 8.3 14.0 13.1 9.5 6.1 4.8 3.8 8.8 temporary Us ing 6.1 13.2 14.7 17.2 19.4 15.6 8.0 15.2 t rad i t iona l Not cont ra - 32.0 19.7 15.0 i0. i 7.3 5.8 4.0 i i .0 cept ing TOTAL % i00 i00 i00 i00 I00 i00 i00 i00 WEIGHTED N = 127 695 1071 1146 1044 824 535 5442 * Has a non-cont racept ive open interva l of at least f ive years ** Exposed to the r i sk of p regnancy is the sum of the percent us ing s ter i l i zat ion , modern methods, t rad i t iona l methods or not us ing any method. It is de f ined as women who have: -- menst ruated in last s ix weeks and -- had sex in last four weeks and -- an open interva l of less than f ive years or, used cont racept ion dur ing the interval . if longer 24 whether they have had a non-contracept ive open interval of at least five years). Women who are found to be fecund are checked to see if they report having sexual intercourse in the last month, a category that wil l include postpartum abstinence. Women who have had sex in the last month are then checked to see if they reported having a menstrual per iod in the last six weeks, a group that could include some menopausal women not already c lass i f ied in the preceding categor ies as wel l as premenarchia l women, and women who have just become pregnant but do not recognize their condition. All remaining women are regarded as exposed. Exposure as used here is a theoret ical concept. Many of these exposed women are protected through the use of some form of cont racept ion . Because of this , Tab le 2.5 checks the contracept ive status of the exposed group and l ists those using either male or female steri l ization, modern temporary methods, tradit ional methods, or no method. Figure 2.1 graphica l ly shows the percent of exposed women protected through the use of contraception. It is important to note that among the women c lass i f ied as exposed, a high proport ion are using either ster i l izat ion, a modern temporary method of contraception, or a t radit ional method. Also, because of the hierarchia l sort ing procedure used to c reate Table 2.5, there can be some over lap between categories. For example, a woman could be c lass i f ied in the infecund category who also did not have sex in the past four weeks, or who did not have a per iod in the last six weeks. And, a woman could be c lass i f ied as not having had sex in the last four weeks or a period in the last six weeks who is also using ster i l izat ion. Thus the order of the category in the table wil l affect the frequency of "no sex in the last 4 weeks," "no per iod in the last 6 weeks," "using ster i l izat ion," "using modern temporary," and "using tradit ionals." The categor ies pregnant, amenorrheic, infecund, and exposed are mutual ly exclusive. The prevalence of pregnancy is 6.9 percent. These are conf i rmed pregnancies based on a report by a woman which is then checked for consistency against her last date of menstruat ion. However, some pregnancies of very early durat ion may not have been conf i rmed and hence not reported. Since the proport ion of nonpregnant women who had no per iod in the last six weeks before interview was 3.2 percent, the level of unreported early ~regnancies probably do not exceed this figure. The reported 6.9 percent of women who are current ly pregnant represents a decl ine from the i0 percent reported by SLWFS and the 9 percent reported by SLCPS, which is in keeping with the observed decl ines in fert i l i ty d iscussed in chapter 3. Pregnancy rates are higher among the youngest age groups, par t i cu la r ly women below age 30. A high proport ion of 25 Exposure to Among Currently Figure 2.1 the Risk of Pregnancy Married Women Aged 15-49 ¢b PrA~n~ Amenorrl 8 .8~ Infecund 9.4% Not Using a Method 1 1 .OK Traditional Methods 15.2% Modern Methods 8.8% No Sex In Last 4 Weeks 1~ No . . . . . . in Last 6 Weeks 3.2~ All Currently Married Women Sterlilzed 23.0% Women Exposed to the Risk of Pregnancy Sri Lanka DHS 1987 pregnancies among the youngest age groups and re lat ive ly few among those over age 30 suggests that women in Sri Lanka not only are having smal ler families, as wil l be d iscussed in Chapter 3, but also are complet ing the process of ch i ldbear ing earl ier. The overal l level of infecundity in the sample is 9.4 percent. This is a lower proport ion than the self reported infecundity level of 14 percent in the SLWFS. As expected. infecundity is h ighest (32.9 percent) among women 45-49 years and then shows a monotonic decl ine among younger age groups. Among al l current ly marr ied women, 57.9 percent are potent ia l ly exposed to concept ion but a s igni f icant proport ion of these exposed women are using some form of contraception. Indeed, only ii percent of all women are both exposed and not current ly using any method of contraception. Most of these women are in the younger age categor ies and are probably seeking pregnancy. There are only a re lat ively few older women who are exposed and not current ly using any method of contraception. 2.4 Breastfeeding, amenorrhea, abst inence With the establ ishment of fert i l i ty surveys, and the improved poss ib i l i t i es for making more indepth inquiries, demographers have increas ing ly turned their attent ion to proximate determinants of fert i l i ty such as exposure to r isk of concept ion . The SLDHS col lected detai led informat ion on postpar tum abst inence and postpartum amenorrhea which is pro longed through breastfeeding. The proport ion of births in the last 36 months that are be ing breast fed or whose mothers are s t i l l amenorrheic, abstaining, or insuscept ib le to concept ion is shown in Table 2.6. Insuscept ib i l i ty is def ined as the per iod before mothers resume both menstruat ion and sexual intercourse after the birth of a child. The proport ion of chi ldren whose mothers are current ly breastfeeding, amenorrheic, and abstain ing from sexual relat ions is h ighest during the first two months after a chi ld is born. Among chi ldren born in the past two months, 98 percent of the mothers were breastfeeding, 87 percent were amenorrheic, and 93 percent were abstaining from sex. Breast feeding cont inues for a re lat ive ly long per iod in Sri Lanka and fully 81 percent of chi ldren born i0-ii months ago are sti l l breastfed. Beyond the first year, however, there is a rapid increase in weaning and by the end of the second year less than 40 percent are stil l being breastfed. Postpartum amenorrhea combined with postpartum abst inence from sex leads to a re lat ively high proport ion of women who are 27 TABLE 2.6: Proportion of births in the last 36 months who are being breastfed and whose mothers are still amenorrheic, abstaining, or insusceptible by number of months since birth, SLDHS 1987. MONTHS S INCE BIRTH Weighted nnmher Breast - Amenorr- of feeding heic Abstaining Insusceptible* births < 2 97.5 87.4 92.9 97.7 108 2-3 93.6 74.9 75.1 90.8 143 4-5 91.6 50.9 38.0 6~.0 137 6-7 80.0 36.0 23.7 45.1 131 8-9 85.6 37.0 21.7 46.5 114 I0-ii 80.8 32.4 8.1 34.7 132 12-13 71.9 17.3 13.3 26.9 128 14-15 66.5 14.3 8.8 20.7 149 16-17 69.3 14.5 6.2 18.6 135 18-19 59.0 6.6 2.8 8.5 124 20-21 48 .4 3 .6 7.2 10 .5 138 22-23 38.8 2.2 3.9 6.1 121 24-25 40.7 1.0 5.2 6.2 133 26-27 37.8 3.1 0.6 3.1 156 28-29 34.1 1.3 2.0 3.4 119 30-31 24.6 0.0 3.5 3.5 107 32-33 27.7 0.0 3.6 3.6 135 34-35 20.2 0.0 0.8 0.8 150 Total 59.0 20.8 17.1 26.8 2358 Median 20.2 4.6 3.9 6.0 insuscept ib le to concept ion . Indeed, approx imate ly one th i rd of the mothers who de l ivered i0 - i i months pr io r to in terv iew were insuscept ib le to concept ion because of amenorrhea or abst inence . Tab le 2 .7 examines d i f fe rent ia l s in b reast feed ing , amenorrhea , abst inence , and insuscept ib i l i ty . A cur rent s ta tus es t imate of the mean is used in th i s tab le and is ca lcu la ted by d iv id ing the preva lence of a cond i t ion (breast feed ing , amenorrhea , etc) by the inc idence of the cond i t ion . P reva lence is de f ined as the number of ch i ld ren whose mothers are breast feed ing , amenorrhe ic , etc. at the t ime of the survey. Ignor ing the s l ight d i sc repancy caused by mul t ip le b i r ths , the number of ch i ld ren be ing breast fed is the same as the number of b reast feed ing mothers . Inc idence is de f ined as the average number of b i r ths per month . Th is average is es t imated by summing the number of b i r ths over the last 36 months to overcome prob lems of seasona l i ty . A 28 TABLE 2.7: Among ever -marr ied women, the est imated mean number of months of breastfeeding, amenorrhea, postpar tum abst inence and postpar tum insuscept- ib i l i ty by se lec ted background character ist ics , SLDHS 1987. BACKGROUND CHARACTER- ISTICS MEAN NUMBER OF MONTHS OF: Breast- Amenorr - Insuscep- feeding hea Abst inence t ib i l i ty* AGE < 30 21.8 7.8 6.7 10.2 30 + 24.1 7.0 6.5 i0. i SECTOR Colombo 18.1 6.0 7.0 i0. i Other~Urban 17.8 6.4 5.0 8.3 Rural 23.6 7.7 6.7 10.3 Estates 21.0 7.7 7.0 9.9 ZONE Zone 1 18.1 6.0 7.0 i0.I Zone 2 20.5 6.8 6.2 9.8 Zone 3 21.1 7.0 6.5 9.9 Zone 4 24.3 7.4 6.7 9.7 Zone 5 22.4 8.7 5.9 10.6 Zone 6 25.0 7.4 7.4 10.6 Zone 7 26.0 7.5 7.6 10.5 EDUCATION No educat ion 26.9 9.2 8.7 13.0 Pr imary 23.6 8.0 7.5 10.9 Secondary 23.0 7.6 6.4 10.2 More than sec. 19.6 6.0 5.1 8.2 Total 22.7 7.5 6.6 i0.2 The per iod of insuscept ib i l i ty is def ined as the t ime between b i r th and the resumpt ion of both menst ruat ion and sexual intercourse. It is a f fected by both amenorrhea and postpar tum abstinence. 29 simple d iv is ion of the number of mothers breast feed ing at the t ime of the survey by the average number of b ir ths per month provides a current status est imate of the mean durat ion in months of breastfeeding. A s imi lar ca lculat ion can be made for amenorrhea or other condit ions. Overal l , the mean durat ion of breast feeding is 22.7 months. This relatively long per iod of t ime affects amenorrhea which lasts on the average for seven and a hal f months, an~ when combined with 6.6 months of abst inence produces an average durat ion of insuscept ib i l i ty of 10.2 months. Several important d i f ferent ia ls are also apparent in Table 2.7. The table indicates that women under age 30 breast feed their ch i ldren on the average approx imate ly two months less than mothers over the age of 30. By sector, urban res idents in Colombo as wel l as other c it ies breast feed their chi ldren 5 to 6 months less than women in rural areas who breast feed for an average of 23.6 months. The estates have an intermediate mean durat ion of breast feeding of 21 months. Cons iderab le d i f fe rent ia l s in the mean durat ion of breast feed ing exist by the zone and level of mother 's education. Zone I, wh ich is metropol i tan Colombo, has the lowest mean durat ion whi le zones 6 and 7 which are most ly the dry areas have s l ight ly h igher mean values than other zones. By education, the main d is t inct ions are found between mothers wi th no schooling, mothers wi th ei ther a pr imary or secondary education, and mothers with more than a secondary education. Babies whose mothers have had no school ing are breast fed for a mean durat ion of 26.9 months. As educat ion increases, the mean durat ion decreases to a low of 19.6 months for chi ldren whose mothers have a post secondary education. The pat terns fo r amenorrhea , abst inence , and insuscept ib i l i ty genera l ly fo l low those for breast feeding, a l though the d i f ferent ia ls are less pronounced. 30 CHAPTER 3 Fert i l i ty 3.1 Introduct ion One major object ive of the SLDHS is to est imate fert i l i ty leve ls , trends, and dif ferentials. In the SLDHS survey quest ionnaire, detai led information was col lected from all ever- marr ied women on current, cumulative, and past levels of ferti l ity. Each woman was asked a series of quest ions about the number of sons and daughters l iv ing with her, the number l iv ing elsewhere, and the number who had died. Fol lowing these questions, a full b i r th h istory was obtained from each woman. Interv iewers were tra ined to pay part icu lar attent ion to the col lect ion of accurate dates of b irth (day, month, and year) and death. The use of a full b ir th h istory in the SLDHS yields extensive data on fert i l i ty and chi ld morta l i ty in Sri Lanka. This chapter presents the data on ferti l ity. Two important measures der ived from the b i r th h istory information are the total fert i l i ty rates (TFRs) for calendar year per iods and the mean number of chi ldren ever born (CEB) to women current ly aged 45-49. The numerator for the TFRs consists of l ive births c lass i f ied by (i) segments of t ime preceding the survey us ing the date of interview and the date of birth, and (2) by age of the mother at the t ime of b irth (in convent ional f ive year groupings) using the date of b irth of the mother. The denominator is the number of women-years l ived in the speci f ied f ive year age interval for each t ime segment. The tables on cumulat ive fert i l i ty (children ever born) are taken from a ser ies of quest ions asked about the number of boys and gir ls l iv ing and not l iv ing in the household and the n,,mher of chi ldren who have died. This data was checked for consistency with the data from the birth history. When examined together with var iables such as age, age at marriage, durat ion of marriage, and selected background character ist ics of the mother, the TFR, CEB, and other measures provide a fair ly detai led account of current and past fert i l i ty levels. The accuracy of ~ fert i l i ty data is af fected pr imar i ly by underreport ing of births and deaths and misreport ing of age. These two problems, if substantial , can ser ious ly affect the 31 s tandard demograph ic measures used to descr ibe ferti l i ty. Underreport ing, for example, af fects measures which indicate how many ch i ldren (the quantity) women have had, whi le misreport ing af fects measures which indicate the t iming or "tempo" of chi ldbear ing. Both of these problems are less ser ious for recent t ime per iods when recal l of b i r th and death events is l ikely to be reasonably accurate, but more ser ious for d istant periods. Also, because of re lat ive ly h igh levels of l i teracy and the use of documents (birth cert i f icates and hea l th cards) to record v i ta levents , these two problems are probably less extensive in Sri Lanka than in many other countries. This chapter begins with a summary of total fert i l i ty rates (TFRs) and ch i ldren ever born (CEB) by background character is t ics of the respondent. Trends in age-spec i f ic fert i l i ty are then examined for f ive-year t ime in terva ls before the survey. The data on chi ldren ever born by current age of the mother, age at f i rst marr iage, and durat ion s ince f irst marr iage are also reviewed. The chapter concludes with tabulat ions of the age of mother at the t ime of her f irst birth. Whenever appropriate, the f indings from the SLDHS are compared with data from other, independent sources, most notably, the v i ta l reg is t rat ion system and s imi lar sample surveys such as the Sri Lanka Wor ld Fert i l i ty Survey (SLWFS) conducted in 1975 and the Sri Lanka Contracept ive Prevalence Survey (SLCPS) conducted in 1982. Whi le such compar isons can suggest trends, it is a lways important to note that fert i l i ty est imates der ived from d i f ferent data bases are never ent i re ly comparable. For example, surveys d i f fer in coverage, sample size, data co l lect ion methods, and other areas. As we have noted earl ier, the SLWFS and SLCPS se lected sample areas throughout the ent i re country, whi le the SLDHS exc luded areas in the north and the east. 3.2 Current and cumulat ive fert i l i ty Tab le 3.1 summarizes current and cumulat ive fert i l i ty for res ident ia l sectors and survey sample zones. The table al lows for an init ial assessment of fert i l i ty trends. Current levels of fer t i l i ty as est imated by the TFR can be compared wi th completed levels of fer t i l i ty as est imated by the average number o f ch i ldren ever born (CEB) to all women current ly aged 45-49. This compar ison is best seen in F igure 3.1. It should be noted that the CEB measures shown in Table 3.1 and in F igure 3.1 were ca lcu lated on the basis of all women who were ident i f ied in the househo ld schedule ( including those who were never-marr ied) . In later tables, the same CEB measure wi l l be used but ca lcu lated on the bas is of ever -marr ied or current ly marr ied women wi th completed interviews. The TFR measures in Table 3.1 are presented for women aged 32 TABLE 3.1: Tota l Fer t i l i ty Rates (TFR) for ca lendar year per iods and for f ive years preced ing the survey, and mean chi ldren ever born (CEB) to a l l never marr ied and ever - marr ied women 45-49 years ( from househo ld schedu le )* by background character i s t i cs , SLDHS 1987. F ive Mean BACKGROUND 1984 1981 years CEB CHARACTER- th rough through pr io r to women IST ICS 1987,* 1983 survey 45-49 SECTOR Co lombo 2.1 2.3 2.2 4.3 Other Urban 2.1 2.4 2.3 4.9 Rura l 2.8 3.2 2.9 5.0 Es ta tes 3.3 3.4 3.4 4.9 ZONE Zone 1 2 Zone 2 2 Zone 3 2 Zone 4 2 Zone 5 3 Zone 6 2 Zone 7 3 2 3 2 7 2 8 3 0 3 5 3 6 3 6 22 4.3 24 3.8 26 3.9 27 5.0 32 5.3 30 6 .1 34 6.6 Tota l 2.7 3.1 2.8 4.9 * A l l women (taken f rom the househo ld schedule) inc ludes women ever marr ied , cur rent ly marr ied , and never marr ied . ** Inc ludes completed months in 1987 before respondent was in terv iewed, usua l ly between one and four months . 15-49 for th ree t ime per iods : (i) the three year per iod immediate ly preced ing the survey f rom ca lendar year 1984 through 1986 (plus the completed months in 1987 before a respondent was in terv iewed, usua l ly an add i t iona l one to th ree months) , (2) the three year ca lendar per iod 1981 through 1983, and (3) the f ive year per iod pr io r to the survey, a t ime wh ich over laps w i th per iods 1 and 2. Across a l l sec tors andzones , the TFR has dec l ined f rom 3.1 in the three year per iod 1981-83 to the cur rent leve l of 2.7 in the three year per iod 1984-87. Whi le a s imi la r and cons is tent t rend of fe r t i l i ty dec l ine is ev ident for each res ident ia l sec tor and sample zone, there are impor tant d i f fe rences in both the cur rent leve l of fe r t i l i ty and the magn i tude of the dec l ine . 33 Figure 3.1 Children Ever Botln (CEB) and Total Fertility Rates 5 4 3 2 1 0 W 6 Estates Rural -• ' :7 : . . Other Urban Residence Colombo ! Mean CEB:Women 45-49 TFR 1982-1986 (5 Yrs prior to survey) TFR 1984-1987 Sri Lanka DHS 1987 For the most recent t ime period, the Colombo metropo l i tan area and other urban areas have the lowest total fer t i l i ty rate at 2.1. In contrast, the estates have the h ighest TFR at 3.3 wi th other rural areas at an intermediate level of 2.8. By zone, the h ighest TFRs are 3.2 in zone 7 which covers the ra infed dry areas and 3.1 in zone 5 which covers the upper south central hi l l country where many of the estates are located. The lowest TFR of 2.1 is in zone 1 which is the Colombo metropo l i tan area fo l lowed by zone 2 which cons ists of the ColombQ feeder areas wi th a TFR of 2.3. At an intermediate level of 2.5 is zone 3 (the south western coastal low lands), zone 4 (the lower south central hi l l country) wi th a TFR of 2.6, and zone 6 (the i r r igated dry areas) with a TFR of 2.7. By res ident ia l sectors, the magni tude of the fert i l i ty dec l ine betweenthe per iod 1981-83 and 1984-86 has been greatest in the rural areas where the TFR dropped from 3.2 to 2.8. Since this area has the largest percentage of the country 's populat ion, it a lso has the largest impact on the overal l level of fert i l i ty in the country. The Colombo metropo l i tan area and other urban areas have also exper ienced a fert i l i ty decl ine but somewhat less than the rural areas. The total fert i l i ty rate in the estates, on the other hand, has remained v i r tua l ly unchanged, dropp ing only s l ight ly from 3.4 to 3.3. The total fert i l i ty rate is a synthet ic measure of current fert i l i ty which can be interpreted as the average number of b i r ths a woman would have if she surv ived throughout the reproduct ive per iod and she exper ienced the same age-spec i f ic fert i l i ty rates that all women are current ly exper iencing. One method of inferr ing poss ib le changes in fert i l i ty is to compare current levels as indicated by the TFRs in Table 3.1 wi th the average number of ch i ldren ever born (CEB) to women aged 45-49 shown in the last co lumn of Table 3.1. Women 45 and over have essent ia l ly completed thei r ch i ldbear ing and thus CEB can be cons idered a measure of past or completed fert i l i ty. A compar ison of the TFRs for the most recent per iod 1984-1987 wi th the average number of ch i ldren ever born suggests that Sri Lanka is exper ienc ing a t rans i t ion to substant ia l ly lower levels of ferti l i ty. For all res ident ia l sectors and sample zones, the current level of fert i l i ty as measured by the TFR for the per iod 1984-1987 is cons iderably be low the level of past fert i l i ty as measured by the CEB for women aged 45-49. If current age-spec i f ic fert i l i ty rates remain unchanged in the future, then new cohorts of women enter ing the reproduct ive per iod wi l l have approx imate ly 2.7 ch i ldren ever born by the t ime they are aged 45-49 compared wi th 4.9 for women who are current ly in this age group. 35 3.3 Fert i l i ty trends The SLDHS used a complete b i r th h is tory to co l lect fert i l i ty in format ion and generate age-spec i f ic fert i l i ty rates (ASFR). Table 3.2 and F igure 3.2 compare the ASFRs der ived from the SLDHS against s imi lar rates obta ined from other sources. The purpose of this compar ison is to prov ide an init ial examinat ion of fe r t i l i ty trends. The interpretat ion of the data should be undertaken with caut ion s ince some of the measures are based on data from the ent ire country and ca lcu lated for s ingle years whi le other measures are based on data from a sample and ca lcu lated for a per iod of several years. Over the past two plus decades, there has been a substant ia l dec l ine in the total fe r t i l i ty rate in Sri Lanka from 5.0 in 1963 to 2.8 in the per iod 1982-87. The one notable except ion to this t rend appears in the per iod towards the end of the 1970s and the ear ly 1980s (approximately 1976 through 1981) when there was an apparent but temporary increase in fert i l i ty par t icu lar ly among women aged 15-35. This increase is ref lected in the age-spec i f ic rates from the 1981 SLCPS and from the Regist rar Genera l 's Off ice for the ca lendar years 1980 and 1981 combined, par t icu lar ly among women under age thirty. The increase is a lso apparent in the crude b i r th rates shown in F igure 3.3 for the per iod from 1971 through 1985. By the per iod 1983-86, however, fert i l i ty began to decl ine once again, due a lmost ent i re ly to a substant ia l fert i l i ty reduct ion among women aged 30 and over. In the 1983-86 period, the age-spec i f i c rates for these women are cons iderab ly lower than the comparable rates dur ing ear l ier periods. Among women under 30, the age-spec i f ic rates from the SLDHS are essent ia l ly the same as the rates reported twelve years ear l ier by the SLWFS. The one except ion is among the youngest SLDHS age group 15-19 which has a s l ight ly h igher age-spec i f ic fert i l i ty rate compared wi th the SLWFS. The complete b i r th h is tory informat ion co l lected from all ' respondents in the SLDHS is further examined in Table 3.3 which presents age per iod rates by maternal age at birth. The data in Table 3.3 are also d isp layed in graphic form in F igure 3.4. The schedule of rates d isp layed in Table 3.3 and F igure 3.4 are progress ive ly t runcated as the t ime before the survey increases. The SLDHS co l lected informat ion from ever -marr ied women under age 50. The further one goes back in t ime the less complete the age-spec i f ic fert i l i ty schedule becomes. It is not possible, for example, to est imate the fert i l i ty of women aged 45-49 for a per iod more than f ive years before the survey. One reason to present Table 3.3 is to assess the qual i ty of the SLDHS data. In a data array such as this, normal ly one would 36 TABLE 3.2: A compar i son of age spec i f i c fe r t i l i ty ra tes f rom var ious sample surveys in Sr i Lanka and f rom the c iv i l reg is t ra t ion system. AGE SPECIF IC FERT IL ITY RATES a a b c d e AGE SPECIF IC 1963 1970 1974 1981 1980- 1982-87 GROUP SLWFS SLCPS 1981 SLDHS 15-19 52 38 31 34 38 38 20-24 228 172 146 172 173 147 25-29 278 238 161 222 197 161 30-34 240 219 158 177 149 122 35-39 157 134 126 99 89 71 40-44 46 38 43 37 26 23 45-49 7 6 6 0 4 3 Tota l Fer t i l i ty Rate 5.0 4.2 3.4 3.7 3.4 2.8 a. Based on v i ta l s ta t i s t i cs data for 1963 and 1970, see D.F.S. Fernando, " A Note on D i f fe rent ia l Fer t i l i ty in Sr i Lanka , " Demography II, August , 1974. b. Based on SLWFS data for the ca lendar year 1974. c. Based on SLCPS data for ca lendar year 1982. d. Based on data f rom the Reg is t ra r Genera l ' s Of f i ce for ca lendar years 1980 and 1981 combined . e. Based on SLDHS data for the f ive complete years (60 months) p reced ing the survey. 37 Figure ,3.2 Age Specific Fertility Rates 1975 SLWFS and 1987 SLDHS 200 150 100 50 0 Births per 1,000 Women 1987 SLDHS 1975 SLWFS 15-19 20-24 25-29 30-34 Age Group 35-39 40-44 45-49 Sri Lanka Figure 5.5 Sri Lanka Crude Birth 1971 - 1985 (Registrar Rates General) Births per 1,000 Population 52 50 28 26 24 22 20 I I I I I I I I I I I I I 1971 1975 1980 1985 Year 39 expect to find monotonic trends, that is to say, constant ly increasing fert i l i ty over time. Whi le in general these trends are apparent in the table, the except ions occur in the period 5-9 years pr ior to the survey among women who gave birth at ages 15-19 and 20-24. For both of these two age groups, fert i l i ty r ises in the per iod 5-9 years before the survey (1978-1982), decl ines s l ight ly in the per iod 10-14 years before the survey, and then cont inues to increase in the more distant past. There are several poss ib le explanat ions for a peaking of fert i l i ty 5-9 years pr ior to the SLDHS survey. Since the di f ferences between the per iod 5-9 years ago and 10-14 are not that great, one poss ib le explanat ion is sampl ing error. A second possible explanat ion is that births have been misdated from the past to the more recent present and/or births by older women have been omitted. In the case of Sri Lanka, however, an equal ly p lausib le explanat ion and one that is consistent with the data presented in Tables 3.2 and Figure 3.3 is that fert i l i ty rose s l ight ly in the per iod five to ten years pr ior to the survey but has now begun to decl ine once again. As noted earlier, the temporary increase in fert i l i ty is ref lected in independent est imates der ived from other surveys and from the country's vital registrat ion system. Also, Table 2.3 presented in Chapter 2 indicates that the proport ion of women marr ied among the cohorts 15-19 and 20-24 was h igher in 1981 than it "was in either 1975 or 1987. We can conclude that the fert i l i ty trends from the SLDHS are consistent with the trends obtained from other sources. This suggests that the birth h istory data from the SLDHS is of reasonably good qual i ty and a fair ly accurate indicator of current and past reproduct ive behavior. 3.4 Chi ldren ever born Tables 3.4 (for ever-marr ied women) and 3.5 (for current ly TABLE 3.3: Age-per iod fert i l i ty rates (per 1,000 women) by maternal age at birth of child, SLDHS 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 38 47 44 51 70 20-24 147 171 168 178 236 25-29 161 203 216 253 (274) 30--34 122 157 185 (238) 35-39 71 91 (114) 40-44 23 (31) 45-49 (3) 105 (100) (242) ( ) Indicates t runcated rates.40 Figure 3.4 Period Age-Specific Ferfilify Rafes 300 250 Births per 1,000 Women 20-24 Years Ago 15-19 Years Ago 200 150 100 50 10-14 Years Ago ~5-9 Years Ago 0-4 Years Ago 0 I I I I I I 15-I 9 20-24 25-29 30-34 35-39 40-44 Age Group Sri Lanka DHS 1987 TABLE 3.4: Percent distr ibut ion of ever-married eeo~n aged 15-49 by number children ever born (CEB) according to age, SLDHg 1987 and SLWFS 1975. CI~|LDREN EVER BORN CURRENT Percent ~/eighted Mean AGE 0 1 2 3 4 5 6 7 8 9 10+ ToteL Humber CEB 15-19 48.9 41.8 7.9 1.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 100 134 0.6 20-24 21.4 40.5 27.4 9.1 1.3 0.1 0.1 0.0 0.0 0.0 0,0 100 723 1.3 25-29 9.7 25.7 29.2 24.3 8.0 2.7 0.4 0.0 0.0 0.0 0.0 100 1113 2.0 30-34 5.0 14.4 26.0 26.6 15.4 7.9 3.3 0.7 0.4 0.1 0.I 100 1203 2.8 35-39 4.5 9.3 21.5. 22.4 18.1 12.1 6.7 3.0 1.6 0.6 0,3 100 1131 3.3 40 "/~ 2.8 7.0 12.3 17.9 15 .5 17.0 11.2 6.8 4.6 2.3 2,6 100 922 4.3 45-49 3.3 4.7 9.1 9.1 20.1 13.4 12.5 8.7 9.0 4.2 5.9 100 639 5.1 ALl Ages SLDHS 8.3 17.2 21.5 19.3 12.9 8.6 5.2 2.7 2.1 1.0 1,1 100 5865 3.0 ALL Age8 SLWFS 8.4 14.6 14.0 13.5 11.8 9.9 8.4 6.5 5.2 7 .7" --- 100 6813 3.9 * For 9+ CEB TABLE 3.5: Percent distr ibution of currentty married women aged 15-49 by number of children ever born (EEB) according to current age, SLDHS 1987 and SLWFS 197~. CHXLDREM EVER BORN CtJI~EEMT Percent Weighted Mean AGE 0 1 2 3 4 5 6 7 8 9 10+ Total Murder CER 15-19 50.2 40.1 8.~ 1.5 O.O 0.0 0,0 0.0 O.O O.O 0.0 100 127 0.6 20-24 22.0 39.5 27.8 9.1 1.3 0.1 0 ,1 0.0 0.0 0.0 0.0 100 695 1.3 25-29 9.9 24.8 29.5 24.5 8.2 2.7 0.5 0.0 0.0 0.0 0.0 100 1071 2.1 30-36 5.2 13.7 26.4 26.5 15.5 8.0 3,4 0.7 0.4 0.1 0.1 100 1146 2.8 35-39 4.2 8.6 21.5 23.3 17.7 11.9 7.2 3.1 1.7 0.6 0.3 100 1044 3.4 40-44 2.7 6.4 12.1 18.1 13.5 17.2 11,2 6.9 4.8 2.5 2.8 100 824 4.4 45-49 2.6 4.4 9.5 9.4 21.2 13.7 12,3 8.9 7.8 4.5 5.7 100 535 5.1 Ati Ages SLDHS 8.5 16.8 22.0 19.7 12.9 8.5 5,1 2.7 1.9 1.0 1.1 100 5442 3.0 ALL Ages SLWFS 8.5 14.2 14.0 13.5 12.0 10.1 8,5 6.3 5.1 7.9* --- 100 6163 4.0 * For 9+ CEB 42 marr ied women) show the number of chi ldren ever born by current age of the mother . These two tables are v i r tua l ly identical which indicates that most mari ta l d isso lut ion occurs among older women who have already completed their chi ldbearing. Consider ing the entire sample, both ever-marr ied and current ly marr ied women have had an average of 3.0 chi ldren ever born. Th is can be compared with the 1975 SLWFS f indings of 3.9 chi ldren ever born to ever-marr ied women and 4.0 to current ly marr ied women. Across all ages, only one-th i rd of the women in the SLDHS have 4 or more chi ldren compared with one hal f of the women in the SLWFS. Clearly, the greatest interest in these two tables centers on women aged 45-49 who have essent ia l ly completed their ferti l ity. Table 3.6 compares the completed fert i l i ty exper ience of women in the SLDHS, SLCPS and SLWFS. For both ever-marr ied and current ly marr ied women in the 45-49 age cohort, the average number of chi ldren ever bore is 5.1 in the SLDHS. This is a decl ine from 6.0 for ever-marr ied and 6.3 for current ly marr ied women reported by the SLWFS, and a dec l ine from the 5.8 reported by the 1981 SLCPS for both ever-marr ied and current ly marr ied women. The reduct ion in completed fert i l i ty is ref lected in the par i ty d ist r ibut ion d i f ferences between the three surveys. In the SLDHS, there most often is an increase in the proport ion of women with par i t ies 0 through 5 and a concomitant decrease in the proport ions with parit ies 6 or h igher compared with both the SLWFS and the SLCPS. This d i f ference between the SLDHS and the other two surveys holds for both ever-marr ied and current ly marr ied women and indicates that fewer women are moving on to higher par i t ies than in the past. Table 3.6 also provides an indicat ion of pr imary steri l ity. Among current ly marr ied women 45-49, 2.6 percent in the SLDHS sample have had no chi ldren compared with 2.1 percent in SLCPS and 2.3 percen£ in the SLWFS sample. These percents have remained re lat ively constant over the past twelve years and are lower than in many other developing countr ies where pr imary ster i l i ty typical ly is found among 3 to 5 percent of current ly marr ied women aged 45-49. The changes in completed fert i l i ty between the three surveys can be seen graphica l ly in Figure 3.5 which shows parity progress ion rations. These ratios indicate the proport ion of women at a given parity who subsequent ly wil l go on to a higher parity. Fo r example, if the par i ty progress ion rat io at par i ty 2 is .90 this means that 90 percent of par i ty 2 women wil l subsequent ly go on to have three or more children. Typically, in non-contracept ing populat ions par i ty progress ion ratios decl ine gradual ly whi le in contracept ing populat ions there is a sharper 43 TABLE 3.6: Percent d i s t r ibut ion of ever -marr ied and cur rent ly marr ied women aged 45-49 by number of ch i ld ren ever born, 1975 SLWFS, 1982 SLCPS and 1987 SLDHS. EVER-MARRIED WOMEN NUMBER OF CHILDREN SLWFS EVER BORN 1975 CURRENTLY MARRIED WOMEN SLCPS SLDHS SLWFS SLCPS SLDHS 1982 1987 1975 1982 1987 0 3.2 1 5.4 2 7.8 3 7 .5 4 8.8 5 10.4 6 13.3 7 11.6 8 9.3 9+ 22.8 2.8 3.3 2.3 2.1 2.6 5.4 4.7 3.6 5 .0 4.4 8.0 9.1 6.4 8.5 9.5 8.2 9.1 6.9 8.5 9.4 i i .0 20 .1 8.5 .11.0 21.2 12.7 13.4 11.2 13.3 13.7 13.2 12.5 14.4 13.3 12.3 9.7 8~7 i i .0 8.7 8.9 7.1 9.0 9.7 7.3 7.8 21.8 i0 . i 26.0 22.2 10.2 Tota l i00 Number of Women 995 Mean CEB 6.0 i00 i00 i00 i00 i00 536 639 817 436 535 5.8 5.1 6.3 5.8 5.1 and more abrupt dec l ine . F igure 3 .5 ind icates that in a l l th ree surveys , approx imate ly the same propor t ions o f women progressed f rom par i ty 0 to I, 1 to 2, and 2 to 3. Thereaf ter , wh i le the SLWFS and SLCPS ra t ios show a gradua l dec l ine and s l ight separat ion , the DHS ra t ios revea l an abrupt dec l ine , most notab le between par i ty 3 and 4 and 7 and 8. The d i f fe rences between the ra t ios are what one wou ld expect in a popu la t ion where cont racept ive preva lence o f modern methods (and in par t i cu la r female s ter i l i za t ion) has increased sharp ly and more than doub led over the past twe lve years . 3 .5 Ch i ld ren ever born and age at marr iage Tab le 3.7 permi ts an examinat ion of the re la t ionsh ip between age of marr iage and the ra te of ch i ldbear ing . The data ind icate that across a l l mar r iage durat ions , age at f i r s t marr iage has a s t rong e f fec t on fe r t i l i ty . The las t row in the tab le shows that the average number of ch i ld ren ever born dec l ines f rom 5.6 for women marr ied be fore age 15 to 2.1 for women marr ied between 25 44 Parity 1975 SLWFS, Figure ,5.5 Progression 1982 SLCPS, Ratios for and 1987 SLDHS Parity Progression Ratio 1.0 . . . . . . . . . . . . . . . . ~ . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 .9 0 .8 0 .7 0 .6 0 .5 0 .4 I I I I t I O. 1 2 3 4 5 6 7 8 Pa rity • 1975 SLWFS I 1982 SLCPS 1987 SLDHS Sri Lanka to 27 years. The reasons for this are fair ly straight forward. An ear ly age of marr iage results in more years of exposure to concept ion and chi ldbearing. Conversely, a late age of marr iage results in fewer years of exposure before dec l in ing fecundity after age 40 and subsequent menopause. The ef fect age of marr iage has on fert i l i ty is far more pronounced at h igher marr iage durat ions and less so at lower marr iage durations. For example, at durat ions 0-4 years and 5-9 years, fert i l i ty is essent ia l ly unaf fected by the age at which a woman f irst marries, s imi lar ly , for those marr ied 5-9 years, age at marr iage has almost no effect on the average number of ch i ldren ever born. Conversely, at h igher durat ions of marr iage (15 years and over) an older age of marr iage means a woman has had fewer years of exposure to concept ion and poss ib ly some loss of fecundity after age 40. These two exper iences combine to reduce ferti l ity. 3.6 Age at f irst b i r th The age at which women first give birth is an important indicator of fert i l i ty trends. A high age at f irst birth often is assoc iated with an increase in the age at marr iage and a subsequent decl ine in ferti l ity. Conversely, an ear ly age at f irst b i r th is often associated with high infant and maternal mortal i ty. Table 3.8 is a percent d ist r ibut ion of all women (taken from the Household Questionnaire) according to age at f irst birth by TABLE 3 .7 : Mean number of ch i ld ren ever born (CEB) to ever -merr ied women, by age at f i r s t marr iage and durat ion s ince f i r s t marr iage, SLDHS 1987 and CEB for eLL ages and durat ions , SLWFS 1975. DURATION AGE AT FIRST MARRIAGE Tote[ SINCE FIRST a l l ages MARRIAGE < 15 " 15-17 18-19 20-21 22-24 25-27 SLDHS SLWFS 0-4 0.8 1.0 1.0 0.9 1.0 0.9 1.0 0.9 5-9 2.4 2.3 2.3 2.3 2.3 2.0 2.3 2.5 10-14 3,3 3.4 3.2 3.1 2.9 2.7 3.1 3.7 15-19 4.0 4.3 4.1 3.8 3.4 3.5 3.8 4.9 20-24 5.2 5.1 4.6 4.5 4.0 3.6 4.6 5.7 25-~ 6.1 5.7 5.4 4.9 3,9 5.4 6.5 30+ 7.4 6.3 5.6 6.8 6.9 AIL Marr iage 5.6 3.8 3.2 2.9 2.6 2.1 3.1 Durat ions SLDNS Art Marr iage Durat ions SLWFS 5 .7 4.8 4.1 3.5 ~.8 2.4" 3.9 f * For age group 25-29 46 TABLE 3 .8 : Percent d i sc r ibut ion of s i t ~¢lnen* by age at f i r s t b i r th t fnc lud~n9 the category ,no b i r th ' ) , according to cur rent age, $LDHS ~Q87. AGE AT FIRST BIRTH Ned{an CURRENT No Total Weighted age at AGE b f r ths <15 15-17 18-19 20-21 22-26 25-27 28-30 31+ ~,ercent number 1st b i r th 15-19 96.4 0.1 2.0 1.5 0.0 0.0 0.0 0.0 0.0 100 1923 ** 20-24 67.0 0.5 4.8 11.1 11.1 5.4 0.0 0.0 0.0 100 1723 ** 25-29 37.7 0.7 6.1 12.2 15.2 17.4 9.5 1.2 0.0 100 1613 24.7 30-34 19.2 1.0 5.2 10.6 17.9 20.7 13.9 8.3 3.0 100 1416 24.1 35-39 14.2 1.5 8.5 10.5 11.0 19.1 16.1 11.8 7.2 100 1258 24.9 40-44 9.4 5.3 12.4 12.9 13.0 17.9 13.1 6 .8 9.2 100 989 23.1 45-49 6 .9 4.5 18.0 13.5 15.2 16.9 10.3 7.1 7.6 100 663 21.8 A l l Ages 43.9 1.5 6 .7 9.6 11.0 12.5 7.8 4.2 2.9 100 9585 ** * Taken from Rousehotd Ouest ionna|re ** Omitted due to censor ing current age. Whi le the Household Quest ionnai re did not record informat ion on births to women never married, these women are assumed to have had no births. The last column shows the median age at f irst birth. Medians for women current ly aged 15-19 and 20-24 and for the total have been omitted to avoid the censor ing prob lem for women who have their f irst b i r th at an older age. The median age at f irst birth among women 25 through 39 has f luctuated between approximately 24 years to a lmost 25 years. For o lder cohorts 40-44 and 45-49 the median age at f irst birth is lower at 23.1 and 21.8 years respectively. What is noteworthy in Table 3.8 is the re lat ively large percent of women who have given no births. Well over a third (37 percent) of women 25-29 have had no births and approximately one- f l f th of women aged 30- 34 have had no births. An important health indicator for any country is the percent of women whose first b i r th occurs before they reach the age o f 18. These women usual ly are c lass i f ied as a high r isk group because they tend to have higher infant and maternal morta l i ty rates than women between the ages of 18 and 35. Table 3.8 indicates that the percent of women whose f irst birth occurred before they reached the age of 18 decl ines wi th each five year age cohort from women aged 45-49 to women aged 20-24. Among the former group, approximately 23 percent had their f irst chi ld before they reached the age of 18. Among women 30-34, 6 percent had their f irst chi ld before the age of 18 and among those 20-24, only 5 percent fall into this category. Clearly, the percent of womem who are at h igh r isk because of ear ly age at f i rst birth has decl ined s igni f icant ly in recent years. Table 3.9 summarizes the median age at f irst b i r th among all women (taken from the Household Quest ionnaire) for d i f ferent cohorts and subgroups. Urban women have a h igher median age at 47 f i rst b i r th than women in rural areas or on the estates. Among urban women, those l iv ing in "other urban" areas have a median age at f irst b i r th of 26.5 years. This is four years h igher than the median among women on the estates. By zone, women in Colombo zone 1 and women in the south western coastal low lands, zone 3, have the h ighest median age at f irst b i r th whi le women in zone 7, the rain fed dry zone, have the lowest median age at f irst birth. TABLE 3.9: Med ian age at f i rs t b i r th among al l women* aged 25-49 by cur rent age and background character i s t i cs , SLDHS 1987. CURRENT AGE BACKGROUND CHARACTER- 25-29 ~ 30-34 35-39 40-44 45-49 Al l ages IST ICS 25-49 SECTOR Co lombo 26.9 25.4 25~7 24.1 23.4 25.3 Other Urban ** 26.8 26.4 26.0 23.9 26.5 Rura l 24.5 23.8 24.7 22.5 21.7 23.7 Es ta tes 22.0 23.1 23.5 23.2 21.1 22.5 ZONE Zone 1 26 .9 25.4 25.7 24.1 23.4 25.3 Zone 2 25.6 24.4 26.6 24.8 25.0 25.2 Zone 3 26.4 26.5 26.9 26.5 24.9 26.4 Zone 4 24.7 24.0 24.5 21. i 21.2 23.5 Zone 5 24.2 23.9 24.3 22.9 20.5 23.5 Zone 6 22.8 23.1 23.0 20.4 19.6 22.2 Zone 7 22.7 21.8 22.0 19.8 18.9 21.4 Al l Women 24.7 24.1 24.9 23.1 21.8 24.0 * Taken f rom Househo ld Quest ionna i re . ** Med ian cou ld not be ca lcu la ted because more than 50 percent of women in th i s g roup have not had a f i rs t b i r th . 48 CHAPTER 4 Fert i l i ty Regulat ion i 4.1 Introduct ion The data on fert i l i ty regulation, which is the subject of this chapter, is drawn from the information col lected in sect ion 3 of the SLDHS questionnaire. The quest ions in that sect ion did not deviate from those in the standard DHS "A" core quest ionnaire used in h igh contracept ive prevalence countr ies except for a few addit ions designed to determine the prevalence of pro longed abst inence (rhythm or the safe period). Past surveys such as the SLWFS and the SLCPS have revealed a re lat ive ly large proport ion of marr ied women use so cal led tradit ional methods including pro longed abstinence. The SLDHS included several quest ions intended to provide information on this area of research and program interest. 4.2 Contracept ive knowledge Data On Contracept ive knowledge and use were co l lected from ever -marr ied women aged 15-49. Knowledge of var ious cont racept ive methods was measured in two ways. First, unprompted knowledge about speci f ic contracept ive methods was obtained by asking each respondent if she knew of any methods a couple could use to delay or avoid pregnancy. If the respondent reported in the aff irmative, she was asked to name all of the methods she knew. For every method not ment ioned spontaneously, the interv iewer would prompt the respondent by reading the name and a br ief descr ipt ion of the method. The methods and descr ipt ion read were the fo l low ing : 1. Pil l 2. IUD 3. In ject ions "Women can take a pi l l every day." "Women can have a loop or coil p laced inside them by a doctor or a nurse." "Women can have an in ject ion by a doctor or nurse which stops them from becoming pregnant for several months." 49 4. Diaphragm, Foam, Je l ly 5. Condom 6. Female s ter i l i zat ion 7. Male ster l l izat lon 8. Safe Per iod 9. Wi thdrawal i0. Norp lant ii. Any other methods "Women can place a sponge or supposi tory or d iaphragm or je l ly or cream inside them immediate ly before intercourse." "Men can use a rubber sheath dur ing sexual intercourse." "Women can have an operat ion to avoid having any more chi ldren." "Men can have an operat ion to avoid having any more chi ldren." "Couples can avoid having sexual intercourse on certa in d@ys of each month when the woman is more l ikely to get pregnant ." "Men can be careful and pul l out before cl imax." "Women can have a tube inserted into their arms and avoid pregnancy for many years." "Have you heard of any other ways or methods that women or men can use to avoid pregnancy?" Knowledge of contracept ion as used in this report is the propor t ion of women who spontaneous ly ment ion a method plus those who were prompted. The percent of ever -marr ied women who know any method, any modern method, and any speci f ic method is shown in Table 4.1. It is c lear from this table that knowledge of at least one or more methods is near universal . Indeed, knowledge of most speci f ic methods) par t icu lar ly modern methods, is also very high. For example, 98 percent of the women interv iewed (the h ighest p ropor t ion fo r any s ing le method) knew about female ster i l izat ion. Knowledge of the pi l l and male s ter i l i zat ion is a lso high, each method known to 91 percent or more. Except for Norp lant (which is st i l l in an exper imental stage) and vaginal methods , modern methods are better known to women than t rad i t ional methods. Among the t radi t ional methods, fewer than two th i rds knew about per iod ic abst inence, and fewer than two f i f ths knew about withdrawal. By age of woman, knowledge levels are un i formly high among all groups and there are very few outstanding di f ferent ia ls . As expected, the youngest age group 15-19 years and the o ldest 45-49 years tend to show somewhat lower levels of knowledge par t i cu lar ly for temporary methods. 50 TABLE 4.1: Percentage of ever-married wceen aged 15-49 knowing any method, knowing any modern method, and kne~in9 specific contraceptive methods, SLDHS 1987, PERCENTAGE OF EVER-MARR%ED 5/Of4Eg WHO KNOW: Any Diaphragm, Per iodic CURRENT Any modern Injec- foam, FemaLe Male abstin- With- Norpiant Other Weighted AGE method method* P i l l IUO table je l l y CarOm s ter i l , s te r i i , efice drawai nurber 15-19 97.4 97.4 80.7 47.7 72.5 10.6 48,4 95.1 82.9 33.2 26.5 1,7 0.9 1~ 20-Z4 97.5 97.5 90.0 74.0 81.9 12.1 66,0 95.1 86.9 53.9 33.0 4.5 1.2 25-29 99.3 99.2 94.1 83.3 87,5 14.9 76,4 98.7 92.4 59.A 39.9 6.7 0.5 1113 30-34 99.5 99.5 95.2 86.7 87.5 14.6 77.7 98.8 93.9 64.5 42.4 6.1 1,8 1203 35-39 99.2 98.9 93.0 87.1 83.7 16.8 76,9 98.1 93.4 65.8 61.5 6.1 1.3 1131 40-46 99.3 99.3 93.8 84.6 81.5 15.5 72,0 98.5 90.8 63.8 36.9 5.3 2.1 922 45-49 97.3 97.0 89.1 70.1 73.8 9.6 59,1 95.4" 83.8 55.5 28.6 3.8 1.4 639 At[ Ages 98.8 98.7 92.7 82.4 83.3 14.3 72,3 97.7 90.8 60.7 37.8 5.5 1,4 5865 * Modern methods inc L~ pl , IUD, in jectables , diaphragm, foam, jd ty , condom, s ter i l i za t ion , and Norptant. The data on knowledge from the SLDHS is certa in ly not surprising. Knowledge of contracept ive methods has been high among Sri Lankan women for at least the past two decades. In 1975, SLWFS reported that 91 percent of ever-marr ied women knew at least one method of contraception. By 1982, the SLCPS reported that 99 percent of ever-marr ied women knew at least one method of family planning. Table 4.2 shows the percent of current ly marr ied women who know at least one modern method of contracept ion cross c lass i f ied by the number of l iv ing chi ldren and by selected background character i s t i cs . The table reveals that there are no s igni f icant d i f ferent ia ls in knowledge according to the place of current residence, socio-economic zones, or education- For instance, the level of knowledge var ies in a narrow range from 97.2 percent for no school ing to 99.8 percent for women with more than a secondary education. Clearly, knowledge about contracept ives is widespread throughout the entire populat ion of ever-marr ied women in the reproduct ive ages. 4.3 Acceptabi l i ty of methods Ever-marr ied women who have heard of a method were also asked about the main problem, if any, they perceive in using part icu lar methods. The responses to this quest ion are tabulated in Table 4.3. Among women who had ever heard of a part icu lar method, the major i ty either stated that there was no perceived problem in gett ing or using the method, or they had no opinion about the 51 method (as indicated by a response of don't know or not stated). Of the spec i f ic concerns mentioned, hea l th is perce ived to be a concern by over a th i rd of the women who have ever heard about the pi l l ; over a fourth for the IUD and female ster i l izat ion; and about a f i f th for in ject ib les and male ster i l izat ion. Health concerns are not a perce ived prob lem for t rad i t iona l methods, a l though ine f fec t iveness and inconvenience were ment ioned s l ight ly more often for these methods than for others. Other poss ib le concerns such as access ib i l i ty and ava i lab i l i ty of supplies, the cost of supplies, and d isapproval of husband are s imply not perce ive~ prob lem areas for the vast major i ty of women in Sr i Lanka. 4.4 Source knowledge The extent to which a woman knows a source or an out let to obta in a spec i f ic contracept ive method is an important aspect of cont racept ive knowledge. For each method a respondent knew, the SLDHS asked "where would you go to obtain (the method) if you wanted to use it?" The responses to this quest ion cross TABLE 4.2: Percentage of cur rent ly married wo~en aged 15-69 knowing at least one modern method, by number of l i v ing ch i ldren and selected back- ground character is t ics , SLDHS 1987. BACKGROUND NUMBER OF LIVING CHILDREN CHARACTER- ISTICS 0 1 2 3 4 5 6* Total SECTOR Co I~ 96.1 99.2 100.0 100.0 100.0 100.0 100.0 99.5 Other Urban 100.0 100.0 100.0 99.3 100.0 100.0 100.0 99.8 Rural 96.2 99.0 99.5 99.7 99.6 99.0 99.1 99.1 Estates 91.8 95.2 98.4 98.6 100.0 95.8 100.0 97.2 ZONE Zone 1 96.1 99.2 100.0 100.0 100.0 100.0 100.0 99.5 Zone 2 95.7 100.0 99.6 99.4 100.0 100.0 1~.0 99.4 Zone 3 95.9 96.6 98.7 98.2 98.9 98.2 97.0 97.8 Zone 4 ¢P5.3 99.3 99.5 100.0 100.0 97.7 98.8 99.0 Zone 5 96.8 98.2 99.3 100.0 100.0 100.0 100.0 99.2 Zone 6 100.0 100.0 100.0 99.2 100.0 100.0 99.0 99.7 Zone 7 95.5 98.9 100.0 100.0 99.0 98.5 99.1 99.1 EDUCATION No education 92.9 89.5 100.0 98.1 100.0 96.5 98.7 97.2 Primary 93.3 99.3 98.8 99.4 99.2 100.0 99.4 99.0 Secondary 95.6 99.2 99.7 100.0 100.0 99.1 99.0 99.2 More than sec. 99.4 100.0 99.7 100.0 100.0 98.0 100.0 99.8 Total 96.2 98.8 99.5 99.6 99.7 99.0 99.2 99.1 52 TABLE 4.3: Percent d i s t r ibut ion of ever-married women abed 15-49 who have ever heard of a method by the main problem perceived in using par t i cu la r methods, i f any, according to method# SLDBS 1987. CONTRACEPTIVE NETHOOS Diaphragm, Periodic HAEB P i l l IUD in jec t - foam, C~-,~,,, Female Hate abst in- With- PROBLEM able je t ty s te r i l , a te r i t , ence drawal Morptant No problem 18.0 15.4 21.0 13.9 27.9 37.9 24.0 66.2 52.2 13.3 Husband disapproves 0.2 0.3 0.2 0.3 2.2 0.3 0.8 0.7 2.2 0.0 Health 36.1 27.2 17.5 8.6 3.9 26.0 18.3 0.1 0.5 6.3 Access/ ava i lab i l i ty O.O 0.1 0.5 O.Z 0.1 0.1 0.0 0.1 O.O 0.6 Costs too much 0.0 0.0 0.3 0.1 0.2 0.0 0.0 0.0 0.0 0.0 Not e f fect ive 1.4 5.2 1.9 2.2 3.4 2.2 1.8 7.1 7.2 0.3 I nc onven i ant to use 1.7 2.2 1.2 1.2 1.7 0.4 0.3 3.9 3.1 0.7 Other 0.7 1.3 0.7 1.5 0.7 1.2 0.7 0.4 0.5 1.2 DZ(, Not stated 41.9 48.3 56.7 72.0 59.9 31.9 54.1 21.5 34.3 77.6 Total Percent 100 100 100 100 100 100 100 100 100 100 Weighted B 5436 4833 4883 838 4237 5728 5325 3558 2218 324 class i f ied by type of method are shown in Table 4.4. Most women who know a method of contracept ion also know a source of supply for that method. This is part icu lar ly true for major methods such as the pill, IUD, injectibles, condom, and ster i l i zat ion but somewhat less so for the diaphragm, foam, jelly, and Norplant. Government hospita ls and clinics, and government midwives and nursess tand out as the major source of supply women would use to obtain most modern methods of contraception. Approx imate ly 69 percent of women who know about the pil l would use a government source of supply for this method. An even higher proport ion of women who know about the IUD, injectihles, and ster i l i zat ion would use a government source of supply. Condoms stand out as the single except ion to this. Sri Lanka has an extremely active social market ing programme and this programme probably accounts for the fact that almost 55 percent of women who know about condoms would obtain this method from a pharmacy or shop. Pharmacies and shops were also ment ioned as a source for pi l ls and for vaginal methods such as foam and jelly. S ince vaginal 53 TABLE 4.4: Percent d i s t r ibut ion of ever-married woraen aged 15-49 knowing a method by supply source named ( i f any), SLDHS 1987. CONTRACEPTIVE METHOOS Diaphragm, Periodic SUPPLY ;n jec t - foam, FefllaLe MaLe abst in- SOURCE PiLL ILID able jeLLy Condom s ter i L , s te r i [ , ence Morpient Govt. hosp. MCH center 46.0 80.1 68.5 34.7 17.8 95.4 92.0 7.6 51.1 Pr ivate doctor 3.3 2.0 15.5 7.2 1.0 1.1 1.0 2.3 5.7 Non-govt. c t in ic 1.6 1.4 1.9 1.8 0.8 0. 8 1.2 0.9 4.3 Mobi te c l in i c 0.8 0.8 0.5 0 .6 0.3 0.1 0.3 0.6 0.3 Govt PH mid- wife, nurse 22.5 2.2 1.9 6.9 8.7 0.1 0.0 28.6 1.8 Other f ie ld source 0.1 0.1 0.1 0.1 0.1 0.0 0.0 0.8 0.0 Ayurvedic doctor 0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.1 0.0 Friend, re la t ive 0.2 0.1 0.1 0.2 0.6 0.0 0.0 32.3 1.3 Pharmacy shop 13.0 0.7 0.6 18.0 54.5 0.0 0.0 0.2 2.1 Other 0.4 0.3 0.1 0.3 0.4 0.1 0.1 19.2 5.4 Nowhere 12.1 12.4 11.7 30.0 15.7 2.4 5.3 7.3 28.0 DK/Not stated 0.0 O.O 0.0 0.0 O.O 0.0 0.0 0.1 0.0 • Total percent 100 100 100 100 100 100 100 100 100 WeightedN 5436 4833 4883 838 4237 5728 5325 3558 324 methods are no longer d is t r ibuted by government midwives, pharmac ies and shops wi l l p robably be used more f requent ly by women for these methods. 4.5 Ever use of contracept ives Tab le 4.5 shows the changes in ever use of contracept ive methods between the 1975 SLWFS, 1982 SLCPS, and the 1987 SLDHS. The data for both the SLWFS and SLCPS excludes the northern and eastern prov inces in order to make the rates comparable with the 1987 SLDHS. 54 TABLE 4.5: Percent of ever -marr ied women 15-49 who have ever used a contracept ive method* by type of method, 1975 SLWFS, 1982 SLCPS, and 1987 SLDHS. TYPE OF METHOD PERCENT EVER USE SLWFS SLCPS SLDHS 1975 1982 1987 Any method 46.5 Any modern method** 27.4 Any t radi t ional 30.0 method *** 69.2 71.8 41.0 50.4 46.6 44.3 Data from the northern and eastern prov inces have been exc luded from the SLWFS and SLCPS in order to make these two surveys comparable with the geographic areas covered by the SLDHS. Modern methods are pill, IUD, injectible, diaphragm, foam, jelly, condoms, male and female ster i l izat ion. Trad i t ional methods include withdrawal, rhythm, and per iod ic abst inence for pregnancy avoidance purposes. Ever use of any method of contracept ion rose from 46.5 to 69.2 percent between the per iod 1975-1982, an increase of 49 percent. Ever use of the two major categor ies of methods also increased sharply dur ing this per iod from 27.4 to 41.0 percent for modern methods and from 30.0 to 46.6 percent for t radi t ional methods. Between 1982 and the SLDHS in 1987 ever use of any method increased only s l ight ly from 69.2 to 71.8 percent. Dur ing this same five year period, ever use of modern methods increased from 41 to 50.4 percent whi le ever use of t rad i t ional methods dec l ined s l ight ly from 46.6 to 44.3 percent. Certa in ly one poss ib le reason for this decl ine is that pro longed abst inence was not included in the 1987 SLDHS l ist of contracept ive methods a l though it was inc luded in the 1982 SLCPS list. Tables 4.6 (for ever-married) and 4.7 (for current ly married) show the percent of women who have ever used speci f ic methods of contracept ion by current age. In both tables, for any g iven age group, from the 15-19 year olds through those current ly 55 aged 44-49, more women have used per iodic abst inence at one t ime or another than any other s ingle contracept ive method. For women current ly aged 30 or more the next most wide ly ever used method is s ter i l i zat ion whi le for those under age 30 it is e i ther wi thdrawal or the pill. In Tables 4.6 and 4.7, ever use of any contracept ive method fol lows an inverted 'U' pattern when examined by age. Ever use is lowest, as one might expect, among women current ly aged 15-19 who are in the process of bu i ld ing their famil ies. Thereafter, ever use increases to a peak in the group aged 35-39 and then begins to dec l ine in the older age groups. In general, this pat tern of overal l ever use by age is common to al l spec i f ic methods except for the minor v~r iat ions in the peak age of use which di f fers accord ing to the method. For instance, the peak age in pi l l use tends to be between the ages of 25 and 34, whi le the peak in ever use of female ster i l i zat ions and per iod ic abst inence is among women between the ages of 35 and 44. 4.6 Trends in current use of contracept ion Table 4.8 and Figure 4.1 prov ide an overv iew of trends in contracept ive preva lence and method use among current ly marr ied women aged 15-49. In the table and the figure, the data from the nor thern and eastern prov inces of the country have been excluded from the 1974 SLWFS and the 1982 SLCPS in order to make al l three data sets comparable. Tab le 4.8 and F igure 4.1 reveal that 61.7 percent of current ly marr ied women are current ly us ing some method of contracept ion. Among these users, approx imate ly two th i rds rely on a modern method and a th ird rely on a t radi t ional method. Whi le the preva lence rate for al l methods has near ly doubled s ince 1974, current use of all modern temporary methods (pill, IUD, In ject ibles, and condom) has remained v i r tua l ly constant at approx imate ly i0 percent. Among speci f ic modern temporary methods, pi l l and in jectable use have increased, IUD use has decl ined, and condom use has f luctuated. Tradi t ional method use increased in i t ia l ly between 1975 and 1982 from 14.2 to 26.0 percent but has subsequent ly dec l ined to 21.1 percent in 1987 (as noted earl ier, however, it is l ike ly that changes in the preva lence of t radi t ional method use may reflect, at least to some degree, d i f ferences in quest ion word ing between the three surveys). Undoubtedly, the major change in the contracept ive preva lence rate has been a three fold increase in ster i l i zat ion use from 10.6 percent in 1975 to 29.8 percent in 1987. Changes in method mix s ince 1975 are apparent in the summary 56 TABLE 4.6: Percent of ever-married t~oeenaged 15-49 who have ever used specific methods of contraception, by current 8ge, SLDHS 1987. PERCENT EVER USED SPEC|FtC RETHOOS Diaphragm, CURRENT Any foam, FemaLe Rate Periodic With- ~eighted AGE method Pitt lUD Injectable jetty COndOm sterit, sterit . Abstinence drawer Norptant Other number 15-19 32,1 11.9 3.0 1.9 0.0 1.7 1.0 0.0 13.0 9.6 0.0 0.0 134 20-24 55.4 17.0 4.1 11.4 0.0 9.1 5.9 2.2 26.9 14.3 0.1 1.0 ;'23 25"29 71.7 17,7 7.9 9.1 0.3 9.3 17.1 5.7 37.0 19.4 0.1 0.3 1113 30-34 79.0 17.7 9.5 6.5 0.3 12.4 26.6 7.3 41.9 21.4 0.2 1.0 1203 35-39 79,8 15.2 9.6 4.3 0.2 9.9 34.6 6.9 43.8 18.7 0.0 0.7 1131 40"44 76.6 12.0 11,2 2,8 0,3 8.8 34,5 5.6 43.4 14.4 O.O 1.2 922 45-49 63.9 9.6 9.6 1.4 0.0 5.9 24.8 3.2 37.6 11.5 O.O 1.0 639 Atl 71.8 15,2 8.7 5.9 0.2 9.4 24.2 5.4 38.6 17,2 0.1 0.8 5665 ages TABLE 4.7: Percent of cur rent ly married women aged 15-49 who have ever used spec i f ic methods of contraception, by current age, SLDHS 1987. PERCENT EVER USED A SPECIFIC METHOD Diaphragm, CURRENT Any foam, FemaLe Mate Periodic ~ i th - Nor- ~etghted AGE method PiLL ]UD InjectabLe je t ty Co~lom s ter i t , s te r i t , ab~tin, drawer p lant Other Number 15-19 31.5 12.5 1.9 1.9 0.0 1.8 1.0 0.0 13.6 9.1 0.0 0.0 127 20-24 56.1 17.3 4.3 11.7 0.0 9.2 5.9 2.3 27.2 14.4 0.1 1.0 69S 25-29 72.9 10.2 8.0 9.3 0.3 9.6 17.3 5.7 37.9 19.9 0.1 0.3 1071 30-34 80.2 18.3 9.7 6.8 0.3 12.7 26.9 7.6 42.7 21.5 0.2 1.0 1146 35-39 82.2 15.6 10.2 4.7 0.2 10.7 35.9 7.4 44.7 19.9 0.0 0.0 1044 40-44 80.6 12.0 11.8 3.1 0.3 9.0 37.0 5.8 45.3 14,5 0.0 1.4 024 45-49 69.0 11.1 9.8 1.6 0.0 7.1 26.3 3.6 40.4 12.6 0.0 1.2 535 ALL 73.9 15.8 8.9 6.3 0.2 9.9 24.9 5.7 39.6 17.8 0.1 0.9 5442 ages 57 TABLE 4.8: Trends in current contracept ive use by method among current ly marr ied women aged 15-49 from the 1975 SLWFS, 1982 SLCPS, and 1987 SLDHS. * PERCENT CURRENTLY USING CONTRACEPTIVE SLWFS SLCPS SLDHS METHODS 1975 1982 1987 Pil l 1.7 2.7 4.1 IUD 5.2 2.9 2.1 In jectable 0.4 1.0 2.7 Condoms ~ 2.3 3.3 1.9 Ster i l i zat ion 10.6 22.0 29.8 Rhythm (periodic abstin.) 8.9 14.2 14.9 Withdrawal 1.6 5.1 3.4 Other 3.7 6.7 2.8 All modern temporary 9.6 9.9 I0.8 All s ter i l i zat ion i0.6 22.0 29.8 All t radi t ional 14.2 26.0 21.1 ALL METHODS 34.4 57.8 61.7 . i * * Data from the northern and eastern provinces have been exc luded from the SLWFS and SLCPS in order to make these two surveys comparable with the geographic areas covered by the SLDHS. Table 4.9. This table shows the percent of current ly marr ied users who are us ing speci f ic contracept ive methods by age in 197S, 1982, and 1987. As in the previous Table 4.8, the data in the SLWFS and SLCPS have been made comparable with the SLDHS by exc lud ing the northern and eastern provinces. It is apparent in Table 4.9 that the proport ion of users re ly ing on modern temporary methods dropped from 27.7 percent in 1975 to 17.1 ~in 1982 and has remained at that level (17.4 percent ) in 1987. The proport ion of users relying on ster i l i zat ion has shown an opposite trend, increasing from 30.0 percent in 1975 to 38.0 in 1982, and to the current level of 48.3 percent of all users in 1987. Finally, whi le tradit ional method users accounted for 41 percent of all use in 1975 and 45 percent in 1982, they now account for only 34 percent in 1987. Again, it is important to note that the di f ferences in tradit ional method use may be due at least in part to quest ionnaire design d i f ferences between the three surveys. 58 Cont racept ive among cur rent ly (excluding Figure 4.1 Preva lence by Method marr ied women age 15-49 nor thern and eastern provinces) 1975 SLWFS 1982 SLCPS 1987 SLDHS 0 10 20 30 40 50 60 70 80 90 100 Percent Sterilized ~ Other Modern Traditional ~ Not Using Sri Lanka Table 4.10 and Figure 4.2 indicate that by age, there was a decl ine between 1975 and 1982 in the proport ion of users under age 25 who were using modern methods, part icu lar ly the pil l and IUD, and a increase in the proport ions of users 35 and over who were re ly ing on steri l izat ion. For example, in 1975, 71.4 percent of current ly marr ied users aged 15-19 and 63.6 percent of users aged 20-24 were re ly ing on a modern method of contraception. By 1982, the proport ions decl ined substant ia l ly to 27.0 and 40.6 percent respect ively for these two age groups. Conversely, in 1975 among the older age groups, 35.4 percent of those aged 40-44 and 29.4 percent of those aged 45-49 were using ster i l izat ion. By 1982 the percent rely ing on ster i l i zat ion in these two age groups rose to 44.8 and 41.7 percent respectively. The trend towards increasing use of ster i l i zat ion cont inued in 1987 among all age groups, but the decl ine noted in 1982 among the younger age groups in the proport ions using modern temporary methods was reversed. The SLDHS reveals that 54.3 percent of current ly marr ied users aged 15-19 were using some method of modern contraception. Among users aged 20-24, 60.6 percent were using a modern method. Of part icu lar interest is the increase in pil l and in jectable use among women under age 25 and the decl ine in IUD use compared with the 1975 SLWFS. Because of d i f ferences noted ear l ier in the way each of the three surveys asked quest ions about tradit ional methods and probed for responses, trends in use are somewhat d i f f icu l t to interpret. For example, whi le both the SLWFS and SLDHS spec i f ica l ly asked respondents about the use of prolonged abst inence to avoid pregnancy, the SLCPS did not. Spontaneous responses about the use of pro longed abst inence were included in the category of "other" methods by the SLCPS. All three surveys did ask respondents speci f ical ly about withdrawal and periodic abst inence (safe per iod or rhythm) a l though considerably d i f ferent probing techniques were used. In rev iewing the data presented in Table 4.10 on the use of t radi t ional methods by age, there is very l i t t le d i f ference between the three surveys in the proport ions using per iodic abst inence or rhythm. In all three surveys, a re lat ively constant one quarter of all contracept ive users rely on this method. A s imi lar constant trend is not apparent between the three surveys with regard to the proport ions us ing withdrawal or other t radi t ional methods. 4.7 Current use by method. Table 4.11 indicates that 29.8 percent of all current ly marr ied women rely on ster i l i zat ion as a method of contraceptioh. The second most widely used method is per iodic abst inence (rhythm or safe period) used by 14.9 percent of the couples. 60 TABLE 4.9: Among current users, changes in method mix between the 1975 SLWFS, 1982 SLCPS, and 1987 SLDHS.* PERCENT CURRENTLY USING CONTRACEPTIVE SLWFS SLCPS SLDHS METHOD 1975 1982 1987 ALL MODERN METHODS 58.7 55.2 65.8 Pil l 4.9 4.7 6.6 IUD 15.1 5.0 3.4 In jectable 1.2 1.7 4.4 Condom 6.7 5.7 3.1 Ster i l izat ion 30.8 38.0 48.3 ALL TRADIT IONAL METHODS 41.3 45.0 34.2 Rhythm 25.9 24.6 24.1 Withdrawal 4.7 8.8 5.5 Other tradit ional 10.8 11.6 4.5 ALL METHODS i00 i00 I00 Data from the northern and eastern provinces has b~en excluded from the SLWFS and SLCPS in order to make these two surveys comparable with the geographic areas covered by the SLDHS. These two individual methods account for 44.7 percent of all use among current ly marr ied women. The balance 55.3 percent is d i s t r ibuted among all the other methods both modern and tradit ional . It is noteworthy that despite cons iderable probing in the SLDHS questionnaire, only 2.8 percent of all current ly marr ied women said they were pract ic ing pro longed abst inence for reasons of avoid ing pregnancy. 4.8 Di f ferent ia ls in contracept ive method use Table 4.12 shows the percent d is t r ibut ion of current ly marr ied women by method current ly us ing accord ing to selected background character ist ics. Other urban areas show the highest current use at 68.4 percent. The Colombo metropol i tan area and the rura l a reas have the same rate of current use at approx imate ly 62 percent. The estate sector is notable in at least four respects. First, estate women have the lowest overal l prevalence rate at 61 Figure 4.2 Confracepfive Mefhod Mix Among Currenfly Married Users 1 5 -49 1975 SLWFS 1982 SLCPS 1987 SLDHS M lCX~ eOli- 4G11. i I~r~t FaJrmmt Permnt :" iiiiiiiiiiiiiii . t~-~1) 20-R4 2ErRg 20"64 ~ 40"44 Age , . : : : : : : : iiiiiiiiiiiiiiiiiiiiiiiiiiiii oi~>::::: ~je ~. ========================= ======================================== Age Controcept;ve Method I PILL ~"~/'~J IUD ~ INJECT/VAG. STERIUZATION I J PERIODIC ~ WITHDRAWAL ABSllNENCE K.///A CONDOM F--q OmER * Excludes northern and eastern areas 51.2 percent . Second, 40 percent of estate women use steri l izat ion, the highest among all women c lass i f ied by sectors. Third, only 8 percent of estate women use tradit ional methods, the lowest among all sectors. F inal ly, less than 4 percent of estate women use modern temporary methods, again, the lowest among all sectors. There are very few major d i f ferent ia ls "by socio economic zones which indicates that contracept ive use is pervasive throughout Sri Lanka. Overall, the lowest prevalence area is zone 5 which is not unexpected since it is dominated by the estate p lantat ion workers. This zone also shows a high prevalence of ster i l i zat ion use and a low prevalence of per iodic abst inence use which is s imi lar to the estate sector. The highest prevalence area is zone 2 which consists pr imar i ly of feeder areas for Colombo and thus contains some of the areas c lass i f ied as "other urban." There is a direct re lat ionship between increasing educat ion and increas ing use of both modern temporary methods and tradit ional methods. For example, only 4.7 percent of women with no educat ion use modern temporary methods and 10.4 percent use tradit ional methods. On the other hand, among women with more than a secondary education, 14.9 percent use modern temporary methods and 31 percent use tradit ional methods. Women with a h igher educat ion are therefore three t imes more l ikely to be using a modern temporary method or a t radit ional method than women with no education. Finally, there is a notable inverse re lat ionship between educat ion and ster i l izat ion. A lmost 40 percent of women with no educat ion or only a pr imary educat ion rely on ster i l i zat ion compared with only 16 percent of women with more than a secondary education. As expected, there is a strong re lat ionship between par i ty and current contracept ive method use. This is part icu lar ly evident for s ter i l i za t ion . At par i ty four or more, over 50 percent of all current ly marr ied women use ster i l izat ion. Modern temporary methods as well as tradit ional methods are used pr imar i ly by women with 1 to 3 children. Women with 4 or more chi ldren are less l ikely to use these methods. 4.9 Number of chi ldren at t ime of f irst use AS Sri Lanka has achieved a high level of contracept ive prevalence over the years, it is of interest to know the pattern of b i r th spacing, part icu lar ly the behavior of women with respect to the postponement of f irst births and the spacing of second or h igher order births. Table 4.13 shows the percent d ist r ibut ion of ever-marr ied women by the number of l iv ing chi ldren at the t ime of the first use of contracept ion according to current age. 63 TABLE 4.10: Percent d i s t r ibut ion of cur rent ly marr ied users aged 15-49 by spec i f i c ~ethod of contracept ion according to cur rent age, 1975 SLWFS, 1982 SLCPS, and 1987 SLDHS. * 1975 SR] LANKA ~LD FERTILITY SURV£~ CONTRACEPTIVE METHODS CURRENTLY USING Prolonged In jec tab le , abs t in , CURRENT or vagina l Female Nate Per iodic With- and other Percent AGE P i l l |UO methods Condom s ter i l , s te r i l , ebs t in , draual methods to ta l 15-19 9.5 38.1 . . . . 19.0 4.8 . . . . 23.8 4.8 . . . . 100 20-24 10.4 29.9 3.9 9.7 9.1 0.6 20.8 7.8 7.8 100 25-29 7.0 15.6 0.3 9.6 25.7 2.6 27.2 6 .7 5.3 100 30-34 4.8 15.1 1.3 8.0 29.2 3.2 26.5. 5.0 6.9 100 35-39 4.6 12,7 0.5 6.8 35.6 2,7 25.8 2.9 8.3 100 40-44 1.5 14.5 0.4 1.5 32.8 2.6 27.2 3.0 16.4 100 45-49 1.2 4.3 1,2 . . . . 29.4 . . . . 27.0 2.5 34.3 100 Total 4 .9 15.1 1.2 6.7 28.7 2.3 25.9 4.7 10.8 100 1982 SRi LANKA CONTRACEPTIVE PREVALENCE SURVEY CONTRACEPTIVE NETHOOS CURRENTLY USING ProLonged In jec tab le , abs t in . CURRENT or vagina l Female Nale Per iodic ~| th - and other Percent AGE P i l l IUD methods Condom s ter i t , s te r i t , abs t in , drawal methods to ta l 15-19 16.2 5.4 . . . . 2 .7 2.7 . . . . 35.1 13.5 24.3 100 20-24 7.1 5.9 6,3 7.5 7.1 6 .7 31.0 11.7 16.7 100 25-29 6.8 6.0 3,3 8.0 23.0 9.0 21.8 11.0 11.2 100 30-34 5.0 3.5 0,4 .6 .2 33.7 8.1 26.9 8.5 7.8 100 35-39 2.0 5.8 1.3 4.9 42.8 6.0 21.1 7.3 8.9 100 40-44 3.8 4.1 0,3 4.1 40.0 4.8 23.7 7.6 11.7 100 45-49 0.7 5.6 . . . . 2.1 39.6 2.1 22.9 4.9 22.2 100 Total 4 .7 5.0 1.7 5.7 31.4 6.6 24.6 8.8 11.6 100 1907 SRI LANKA DENOGRAPHIC AND HEALTH SURVEY CONTRACEPTLVE NETHQOS OJRRENTLY USLNG Prolonged in jec tab le , a~tin. CURRENT or vaginal Female Hate Periodic With- and other Percent AGE P i t t itJO methods Condom s ter i l , s te r i l , abs t in , drawer methods to ta l 15-19 37.5 4.2 4.2 4.2 4.2 . . . . 25.0 16.6 4.2 100 20-24 16.3 5.8 16.7 3.4 13.9 4.4 27.9 7.8 3 .7 100 25-29 12.2 4.1 8.3 2.9 30.2 8.5 25.4 5.9 2.4 100 30-34 6.3 3.0 4.0 3.4 40.3 10.7 23.8 6.4 2.1 100 35-39 3.9 1.9 1.8 3.0 48.6 8.4 23.9 6.1 2.5 100 40-44 1.5 3.6 0.5 3.2 51.6 6.5 24.1 3.6 5.4 100 45-49 1°2 4.8 0.0 1.8 46.9 6.1 19.1 1.4 18.8 100 Total 6 .6 * Data from the to make these 3.4 4.4 3.1 40.4 8.0 24.1 5.5 4.5 100 nor thern and eastern provinces have been excluded from the SLWFS and BLCPS two surveys comparable w i th the geographic areas covered by the SLDHS. 64 TABLE 4.11: Percent of cur rent ly marr ied women aged 15-69 according to contracept ive method cur rent ly us ing by age, SLDHS 1987. PERCENT CURRENTLY USING: CURRENT Any Foam, FemaLe NaLe Per iodic With- Prolonged Weighted AGE method P i l l IUD ln jec tabte Je t ty Condom s ter i [ , s te r f l , abs t in , draws[ Other alostin, number 15-19 20.2 7.2 1.1 0.7 0°0 0.8 1.0 0.0 5.0 3.4 0.0 1.1 127 20-24 42.3 6.9 2.4 7. I 0.0 1.5 5.9 1.9 11.8 3.3 OlO 1.6 695 25-29 57.3 7.0 2.3 4.5 0.1 1.7 17.3 4.8 14.6 3.4 0.0 1.4 1071 30-34 66.8 4.2 2.0 2.7 0.0 2.3 26.9 7.2 15.9 4.3 0.1 1.4 1146 35-39 73.8 2.9 116 1.3 0.0 2.2 35.8 6.2 17.6 4.5 0.0 1.8 1046 40-44 71.5 1.1 2.5 0.3 0.0 2.3 37.0 4 .7 17.3 2.5 0.0 3.9 824 45-49 56.1 0.7 2.7 O l 0 0 . 0 l 1"0 26"3 3"4 10"7 0"8 0"2 10.4 535 Art Ages 61.7 4.1 2.1 2.7 0.0 1.9 24.9 4.9 14.9 3.4 0.1 2.8 5642 TABLE 4.12: Percent d i s t r ibut ion of cur rent ly marr ied women aged 15-49 by method of contracept ion cur rent ly us ing, according to se lected background character i s t i cs , SLDliS 1987. PERCENT CURRENTLY USING: BACKGROUND D iaphra~ Per iodic Prolonged CHARACTER- Any In jec t - foam Con- Female Nate abst in - With abst in - Weighted ISTICS method P i l l ICE) able je t ty dam s ter i [ s te r i t , ence drawer Other erce number SECTOR Colombo 62.6 3.0 0.7 3.1 0,0 3.1 24.2 5.6 18.5 2.5 0.0 2.0 521 Other Urban 68.4 5.3 3.6 1.8 0.0 2.8 22.8 6.5 20.9 2.2 0.0 2.4 Rural 61.9 4.4 2.3 2.8 0.0 1.0 24.7 4.5 15.0 318 0.1 2.7 4210 Estates 51.2 1.5 1.0 1.1 0.0 0.1 30.1 9.9 1.4 0.7 0.0 5.6 ]47 ZONE Zone 1 62.6 3.0 0.7 5.1 0.0 3.1 24.2 5.6 18.5 2.5 0.0 2.0 521 Zone 2 67.1 4.1 2.1 2.5 0.0 2.5 20.4 5.7 19.6 6.2 0.0 5.8 855 Zone 3 63.7 4.0 2.3 0.6 0.0 2.3 18.7 6.0 23.3 4.4 0.2 1.9 752 Zone 4 61.8 4.2 3.0 2.9 0.0 2.0 25.8 3.5 14.8 2.7 0.1 2.7 1245 Zone 5 57.4 5.1 2.;~ 2.4 0.1 1.0 28.9 5.5 6.3 2.5 0.0 3.3 1081 Zone 6 62.3 3.6 1.5 2.5 0.0 1.1 30.1 5.4 12.6 3 .6 0.0 2.0 377 Zone 7 58.1 3.3 1.6 5.1 0.0 1.3 27.2 3.4 11.6 1.7 0.0 2.9 610 EDUCATION No educat ion 54.3 2.1 1.3 1.1 0.0 0.2 32.9 6.3 4.4 1.7 0.0 4.3 579 Primary 63.5 3.3 2.3 1.6 0.1 0.9 33.9 5.9 9.5 2.5 0.1 3.6 1581 Secondary 62.9 5.4 1.7 3.7 0.0 1.6 23.2 5.1 15.7 3.8 0.1 2.6 1990 Rare than sac.61.0 4.0 2.9 3.1 0.0 4.3 12.8 2.9 24.9 4.6 0.0 1.5 1292 PAR l TY No ch i td ren 6.4 0 .7 0.0 O i O O . O O .3 0.7 O " 4 3"3 0.8 0.0 0.1 487 1 ch i ld 43.8 5.6 2.1 4.1 0.0 2 . ] 1.1 1.0 21.2 4.8 0.1 1.6 943 2 ch i ld ren 64.0 7.0 3.6 4.3 0.0 2.7 12.1 6.3 21.6 510 OlO 1.4 1275 3 chi tdren 77.7 4.5 1.7 2.5 0.1 2.6 37.5 7.5 14.7 3 .7 0.0 3.0 1100 4 or more 75.9 1.7 1.9 1.5 0.0 1.0 47.2 5.7 9.6 1.9 0.1 5.2 1637 Total 61.7 4.1 2.1 2.7 OlO 1.9 24.9 4.9 14.9 3.4 0.1 2.8 5442 65 Across all age groups, approximately 6 percent of ever- marr ied women init iated contracept ion before the first birth and 35 percent before the second. About half of all ever-marr ied women init iated contracept ion before the birth of a third child. By age, Table 4.13 suggests that younger cohorts are more l ikely to begin contracept ive use before the first or second bir th than older cohorts. For example, approximately 40 percent of the women between the ages of 20 and 35 init iated contracept ive use before the birth of their second chi ld compared with 33 percent among women current ly aged 35-39, 26 percent among those aged 40- 44, and only 17 percent among the oldest cohort aged 44-49. TABLE 4.13: Percent d i s t r ibut ion of ever-marr ied woTen eged 15-49 by number of l i v ing ch i ld ren at the t ime of f i r s t use of contracept ion by cur rent age, SLDHS 1987. NUMBER OF LIVING CHILDREN WHEN FIRST USED CUMTRACEPTIUM CURRENT Never Don't Tote[ Weighted AGE used None One Two Three Four+ know percent N 15-19 67.9 9.8 18.7 1.4 1.0 0.0 1.4 100 134 20-24 44.6 10.3 32.3 9.1 2.8 0.7 0.2 100 723 25-29 28.3 8.9 33.9 16.9 8.8 3.0 0.2 100 1113 30-34 21.0 5.7 34.2 18.0 11.3 9.6 0.1 100 1203 35-39 20.2 3.6 29.1 17.5 11.9 17.2 0.4 100 1131 40-44 23.4 2.5 24.0 13.2 11.9 24.9 0.2 100 922 45-49 36.1 2.7 14.7 10.2 8 .7 27.2 0.5 100 639 Total 28.2 5.8 28.8 14.6 9.5 12.8 0.3 100 5865 4.10 Knowledge of fert i l~ per iod The prevalence of tradit ional method use is re lat ive ly high in Sri Lanka compared with other countries. For the couples who pract i ce these methods , par t i cu la r ly per iod ic abst inence (rhythm), it is important for them to know when dur ing the ovulatory cycle a woman is most l ikely to conceive. Table 4.14 examines knowledge of the fert i le per iod dur ing the ovulatory cycle among ever users of per iodic abst inence and among never users of this method. Among women who had never used per iodic abstinence, only 31 percent correct ly knew that the most fert i le per iod is dur ing the middle of the ovulatory cycle. Among women who had ever used per iodic abstinence, 64 percent correct ly named the middle of the cycle as the most fert i le time. Whi le this proport ion is more 66 TABLE 4.14: Percent d is t r ibut ion of ever -marr ied women aged 15-49 and women ever us ing per iod ic abst inence by knowledge of the fert i le per iod dur ing the ovu latory cycle, SLDHS 1987. FERTILE PERIOD KNOWLEDGE AMONG: Ever users Never users of per iodic of per iod ic abst inence abst inence Al l ever marr ied women Dur ing per iod 0.8 Af ter per iod 18.4 Midd le of cycle 64.0 Before per iod 3.1 At any t ime 0.0 Other response 2.0 Dont know 11.7 0.5 0.7 10.9 13.8 31.2 43.9 2.1 2.5 0.3 0.2 0.9 1.3 54.0 37.7 Total percent i00 i00 i00 Weighted N 2262 3603 5865 TABLE 4.15: For s te r i l i zed women, the percent d i s t r ibut ion by age at the t ime of s te r i t i za t ion ; according to the number of years s ince the s ter | | | za t ion , SLOHS 1987. AGE AT STERILIZATION YEARS SINCE Totat Weighted Hedian OPERATION <25 25-29 30-34 35-39 40+ percent N ABe < 2 years 17.8 32.1 28.6 15.1 6.3 100 507 30.0 2 - 3 15.3 31.7 31.9 15.0 6.0 100 218 30.3 4 - 5 9.4 33.5 34.3 18.9 4.0 100 245 " 30.9 6 - 7 12.1 30.5 36.4 21.0 0.0 100 117 30.5 8 - 9 9.6 35.6 39.7 15.2 0.0 100 146 30.6 10 + years 12.8 42.9 36.8 7.5 0.0 100 176 29.6 Totat 13.9 33.6 32.7 15.2 3.9 100 1420 30.3 67 than double that of the never users, it indicates third of all ever users of per iodic abst inence are about the t ime of greatest r isk of pregnancy. that over a ill informed 4.11 Age at ster i l i zat ion Table 4.15 indicates that there has been re lat ive ly l itt le change in the median age at which women obtain a ster i l izat ion operation. Among women who have obtained this operat ion within two years, the median age at the t ime of the operat ion was 30. This is only s l ight ly below the median age of women who obtained the operat ion between 2 and I0 years ago and only s l ight ly h igher than the median of 29.6 years among those who obtained the operat ion i0 or more years ago. 4.12 Source of contracept ive supply and sat is fact ion wi th source The data on the most recent source of contracept ive supply used among all current users is d isplayed in Table 4.16. For supply methods, government hospita ls and MCH centers were used by 27 percent of current users and government midwives and nurses by TABLE 4.16: For a l l current users of c~t recept ive methods the percent d i s t r ibut ion according to most recent source of supply, by method, SLDHS 1987. SUPPLY METHOOS CLINICAL METHOUS Diaphragm, Total SOURCE OF foam, Female Hate a l l SUPPLY P i l l In jec t . j e l l y Condom Total IUO s ter i l , s te r i l Total methods Oovt hosp i ta l / HCH center 17.4 54.4 0.0 0.9 26.9 85.4 94.8 78.2 91.7 78.2 Pr ivate doctor 11.2 28.5 100.0 3.0 15.2 4.3 2.0 1.0 2.0 4.7 Non-govt c l in i c 0.6 3.2 0.0 0.0 1.3 0.7 0.9 9.9 2.2 2.0 Nobi le c l in i c 0.4 0.6 0.0 1.3 0.7 0.0 0.5 5.7 1.2 1.1 GOVL PB midwife/ 49.0 12.4 0.0 23.0 31.9 9.6 0.0 O.O 0.6 7.1 r~Jrse Other f ie ld source 0.3 0.0 0.0 1.3 0.4 0.0 0.0 0.0 0.0 0.1 Fr iend/Relat ive 4.1 0.0 0.0 13.2 4.8 0.0 0.0 0.0 0.0 1.0 Phar~cy/Shop 14.0 0.0 O.O 37'.1 14.7 O.O 0.0 O.O O.O $.0 Other 2.6 0.0 0.0 9.6 3.3 0.0 0.0 0.0 0.0 0.7 DK/Not stated 0.5 0.9 0.0 1.0 0.9 0.0 1.9 5.3 2.3 2.0 Total percent 100 100 100 100 100 100 100 100 100 100 WeiBhted number 223 145 2 103 472 116 1410 271 1805 2277 68 another 32 percent. For c l in ic methods, 78 percent of current users re l ied on a government hosp i ta l or MCH Center and another 7 percent on government publ ic heal th midwives or nurses. About 95 percent of the female ster i l i zat ions and 78 percent of male s ter i l i zat ions have been per forme d in government fac i l i t ies. The government 's publ ic heal th f ield staff were used by a lmost hal f of all pi l l users. Besides government sources, pr ivate doctors are an important source for injectables. Pharmacies and shops are important sources for condoms and pil ls. Indeed, 37 percent of all condom users obtain their suppl ies from a pharmacy or shop (which is an indicat ion of the act iv i ty of the contracept ive social market ing program) and the actual percent may be somewhat higher. It is l ikely, for example, that some of the fr iends and re lat ives who are l isted as a source obtain their suppl ies from pharmacies and shops. Current users who obtained a method or contracept ive service from a source were asked about thei r sat is fact ion wi th the source. Table 4.17 indicates that the vast major ity, 95 percent, of all current users did not have any compla int to make against the serv ices provided at a source of supply. 4.13 Reasons for method d iscont inuat ion Women who d iscont inued any contracept ive method dur ing the f ive years preceding the survey were probed for the reasons they d iscont inued the last method used. The results are presented in Table 4.18 according to the method d iscont inued and the main reason for the discont inuat ion. Near ly a th ird of all women (32 percent) who d iscont inued a method did so because they wanted to become pregnant. Another 28 percent of respondents d iscont inued because they -believe the method failed. It should be noted the SLDHS had no way of ver i fy ing method failure. D iscount ing male s ter i l i zat ion with only ten cases, method fai lure is the most l ike ly reason g iven for d i scont inuat ion (37 percent) among users of per iodic abst inence and withdrawal. Method fai lure is a lso an important reason g iven by 23 percent of condom users for d iscont inuat ion. Among all women who d iscont inued any method in the past five years, 12 percent l isted heal th concerns as the main reason. Among women who formerly used the pill, 38 percent said they d iscont inued because of health, 27 percent l is ted this as a reason for d i scont inu ing the IUD, and 29 percent for d i scont inu ing w i th in jectab les . Most other reasons for d iscont inuat ion of a method are tr ivial. For example, less than 1 percent d iscont inued because of lack of ava i lab i l i ty of suppl ies or access ib i l i ty to faci l i t ies or because the method 69 cost too much. Also, only 3 percent d iscont inued because the method was judged inconvenient to use. 4.14 Att i tudes towards becoming pregnant Table 4.19 shows the percent d ist r ibut ion of current ly marr ied exposed women who are not using contracept ives by their att i tudes towards becoming pregnant in the next few weeks according to the number of l iv ing children. Overall , 38.4 percent of these exposed nonusers reported that they would be unhappy if they became pregnant during the course of the next few TABLE 6.17: Among cur rent users of contracept ion obta in ing a method at a source, Phe percent d i s t r ibut ion by type of d i ssat i s fac t ion w i th the serv ice ( i f any) , according to type of source test v i s i ted , SLOHB 1987. DISSATISFACTION WITH SERVICE No Wait Staf f Service Desired Other Mot Total Weighted SOURCE OF complaint too d iscour expen- serv ice stated percent number SUPPLY long t ious s iv~ unavait Bovt hosp i ta l / NCH 95.8 0.5 0.5 0.4 0.2 0.4 2.3 100 1781 center P r ivatedoctor 94.7 0.0 O.O 0.9 O.O 0.8 3.6 100 108 Non-govt c l in i c 83.1 2.2 0.0 3.0 0.0 0.0 11.8 100 46 Mobile c l in i c 70.1 O.O 12.3 O.O O.O 0.0 17.6 100 25 Govt PH midwi fe / 94.7 0.0 0.0 2.1 0.0 0.0 3.2 100 162 nurse Other field 100.0 O.O O.O 0.0 0.0 0.0 O.O 100 2 personnel Total 95.1 0.5 0.6 0.6 0.1 0.3 2.8 100 2127 weeks whi le the majority, 54.4 percent, said they would be happy if they became pregnant. As one might expect, among exposed women with no chi ldren ful ly 95 percent would be happy if they became pregnant in the next few weeks. The percent who would be happy drops sharply to 52 percent among exposed women with one child, 35 percent among those with two children, 26 percent with three, and 13 percent with four or more children. Tab le 4 .20 examines the reasons for the nonuse of contracept ion among current ly marr ied women who I) are exposed, 2) are not current ly using a method, and 3) say they would be unhappy if they became pregnant in the next few weeks. There are only 230 current ly marr ied women who meet all the cr i ter ia for inc lus ion in this table. Approx imate ly half of them are under the age of 30. The reasons for the nonuse of contracept ion among these women are diverse but certa in ly a health concern is an important factor for 17 percent of the women under age 30 and for almost 20 percent of those over age 30. Among the women under 70 TABLE 4.18: Percent d i s t r ibut ion of women who have d iscont inued a method in the test 5 years by main reason for test d i scont inuat ion , SLDHS 1987. Nate Per iodic With- Toter P i t t IUD In jec t . Condom Ster t t . abs t in , drawer Other percent To becon~ pregnant 20:4 25.6 20.7 38.3 0.0 36.5 31.2 15.3 32.0 Method Faired 12.9 15.7 12.0 22.9 70.6 37.4 37.3 21.9 28.0 Infrequent sex 1.5 0.0 0.0 1.9 0.0 3.5 3.8 0.0 2.4 Partner disapproved 0.6 0.5 2.5 ~5.4 0.0 0.8 4.9 22.5 2.0 Hearth concerns 38.3 26.7 29.2 5.4 0.0 0.3 0.0 5.6 11.9 Avsi t abi t i tyl access ib i l i ty 1.2 0.8 5.7 1.4 0.0 0.0 0.0 0.0 0.8 Cost too much 0.0 0.0 1.4 1.7 0.0 0.0 0.0 0.0 0.2 I ncenveni ent to use 3.3 /+.4 1.4 3.5 0.0 3.0 5.8 0.0 3.4 To use perma- nent method 2.9 8 .7 0.7 4.5 0.0 6,9 5.3 15.9 6.2 Other reasons 9.1 16.0 16.9 14.1 19.0 10.0 10.9 14.0 11.4 DK, Not s tated 2.1 1.6 1.5 0.8 10.3 1.7 0.9 4.8 1.7 , Total percent 100 100 100 100 100 100 100 100 100 Weighted N 311 191 117 158 10 804 204 24 1822 30, d isapproval of husband, current ly breastfeeding, and a lack of knowledge about contracept ive methods are three other major reasons for nonuse. Among women over 30, infrequent sex, menopausa l , and husband disapproves were ment ioned fair ly f requent ly by the exposed women. 4.15 Future use ~ All current ly marr ied women not current ly us ing any method of contracept ion were asked about their intent ion to use in the future. The responses to this quest ion tabulated by number of l iv ing chi ldren (including current pregnancy) are shown in Table 4.21. Data in the table are examined according to the intent ion to use in the next 12 months and later. 71 Among all nonusers, approx imate ly one th i rd have a def in i te intent ion to use contracept ion in the future whi le 48 percent have no intent ion to use in the future and 16 percent are undecided. Only 21 percent intend to use in the next twelve months and another 14 percent say they wi l l use af ter 12 months. As the number of l iv ing chi ldren increases from 1 through 3, the percent of current non users who intend to use in the future increases from 19 percent to 36 percent. Among those wi th four or more chi ldren, however, the intent ion to use in the next 12 months drops to 26 percent. The women not current ly us ing a method of contracept ion but who intend to use in the future were further asked which speci f ic method they intend to use. The results are shown in Table 4.22. More than two f i f ths (43.5 percent) of women who are exposed to the r isk of pregnancy and who intend to use in the future TABLE 4.19: Among current ly marr ied exposed* women not us ing contracept ion, the percent d is t r ibut ion by att i tude toward becoming pregnant in next few weeks, according to number of l iv ing chi ldren, SLDHS 1987. ATTITUDE TOWARDS BECOMING PREGNANT Does NUMBER OF Not not Not Percent Weighted CHILDREN Happy happy matter stated total number 0 94.6 3.9 0.8 0.7 I00 187 1 52.1 39.8 4.5 3.6 i00 156 2 35 .1 54.8 3.9 6.2 I00 121 3 26.3 65.8 6.5 1.5 i00 58 4 or more 13.3 72.3 7.5 6.9 I00 78 Total 54.4 38.4 3.8 3.4 i00 599 * Exposed to the r isk of pregnancy is def ined as: -- menst ruated in last s ix weeks and -- had sex in last four weeks and -- has an ope n interval of less than f ive years or, if longer used contracept ion dur ing the interval. 72 TABLE 4.20: Among women who are exposed and not us ing contracept ion, but who would not be happy if they become pregnant, the percent d is t r ibut ion by the main reason for nonuse, accord ing to current age, SLDHS 1987. MAIN AGE REASON Al l FOR NONUSE <30 30+ ages Lack of knowledge 14.4 4.5 9.6 Opposed to FP 2.6 1.2 1.9 Husband Disapproves 19.7 7.2 13.7 Others d isapprove 0.4 0.0 0.2 In f requent Sex 4.3 11.7 7.9 Breast feed ing 16.4 4.3 10.6 Menopausa l 2.0 9.5 5.6 Heal th Concerns 17.4 19.5 18.4 No avai labi l i ty /access. 2.1 3.2 2.6 Re l ig ion 3.2 5.1 4.1 Inconvenient to use 1.8 3.1 2.4 Other reasons II.0 24.3 17.4 Don't know 3.7 5.2 4.4 Not stated I.i 1.2 1.2 Total percent i00 I00 i00 Weighted N 119 IIi 230 73 TABLE 4.21: Among current ly marr ied women aged 15-49 not current ly us ing any method of contracept ion, the percent d is t r ibut ion by intent ion to use in the future, accord ing to number of l iv ing ch i ldren ( including current pregnancy), SLDHS 1987. INTENDS TO USE A METHOD NUMBER OF CHILDREN None 1 2 3 4+ Total Next 12 mos 0.0 18.7 25.1 35.9 25.8 21.4 Af ter 12 mos 18.4 15.9 17.0 i i .0 4.5 13.5 Undec ided when 0.7 3.1 3.1 4.8 2.1 2.8 to use Undec ided if 32.4 17.9 9.1 4.6 4.1 13.2 wi l l use No\ in tent ion 48.5 43.3 45.0 41.5 62.3 48.0 to 'use Not stated . . . . 1.2 0.8 2.2 1.3 I.I Total Percent i00 i00 i00 i00 i00 i00 Weighted N 289 600 487 284 423 2084 report that they wi l l select ster i l izat ion. In ject ions were ment ioned by 20.6 percent of the women and p i l l s by i0 percent. Only 9 percent of current nonusers said that they intend to use t rad i t ional methods such as the safe per iod or wi thdrawal in the future. 4.16 Fami ly P lanning messages on radio The radio is a popular medium of communicat ion in Sri Lanka which is owned by the state but is not w ide ly used in the communicat ion of family p lanning messages to the people. However, the Sri Lanka Fami ly P lanning Assoc ia t ion sponsors a 15 minute programme everyday in the evening at 7:30. This programme is broadcast only in S inhala and includes d iscuss ion on a family p lann ing prob lem and advice by a medical off icer. The Sri Lanka Assoc ia t ion of Vo luntary Surgical Contracept ion a lso runs a radio programme in S inhala and Tamil. The percent of ever -marr ied women who have heard a radio messag e about fami ly p lann ing is shown in Table 4.23. Among all ever -marr ied women, only 36.5 percent have ever heard family p lann ing messages over the radio. Approx imate ly 7 percent have heard a message once whi le 30 percent have heard a message more 74 TABLE 4.22: Among current ly marr ied women not current ly us ing any method of contracept ion but who intend to use in the future, the percent d is t r ibut ion by preferred method accord ing to whether they intend to use in the next 12 months or af ter 12 months, SLDHS 1987. INTENDS TO USE METHOD IN: METHOD Next Af ter INTENDS 12 12 TO USE months months Total Pi l l 13.6 4.3 i0.0 IUD 4.7 1.3 3.4 In jectables 25.1 13.5 20.6 Condom 3.3 2.0 2.8 Female steril. 34.0 58.6 43.5 Male Steril. 1.4 i.I 1.3 Safe Per iod 9.2 4.7 7.5 Withdrawal 1.5 1.3 1.5 ' Norp lant 0.3 0.0 0.2 Other 0.7 0.0 0.4 Not sure 6.1 13.1 8.8 Total percent I00 i00 I00 Weighted N 446 282 727 than once. As fnoted ear l ier in this section, SLDHS respondents have un iversa l ly h igh levels of knowledge about fami ly planning, suggest ing that media other than the radio may be more important sources of in format ion about family planning. By p lace of residence, women l iv ing in urban areas outs ide of metropo l i tan Colombo are more l ike ly to have heard a family p lanning message over the radio than women l iv ing elsewhere. The d i f ferent ia ls by broad soc io-economic zones are not great. 75 Approx imate ly a th i rd of women in each of the zones have heard a family p lann ing message over the radio at least once. By education, there is a d i rect re lat ionsh ip between having heard family p lann ing messages over the radio and h igher education. For example, whi le only 16.5 percent of those wi th no educat ion have heard messages more than once, 41.5 percent of those with more than a secondary educat ion have heard these messages more than once. A l though family p lanning messages have been heard by s l ight ly over one th i rd of all ever -marr ied women, Table 4.24 indicates that ful ly 89 percent of these women bel ieve it is acceptab le to air messages on family p lanning over the radio. By age, the proport ions who be l ieve it is acceptable var ies in a rather narrow range from a low of 82 percent among those aged 15- 19 to 91 percent among those current ly aged 25-29 years. By sector and zone the d i f ferent ia ls are also small, however, by education, the proport ions who approve of radio messages on family p lann ing increases with education. TABLE 4.23: Percent d i s t r ibut ion of ever -marr ied women by whether they have heard a rad io message about fami ly p lann ing , accord ing to se lec ted background character i s t i cs , SLD~S 1987. HEARD F.P. RADIO MESSAGE BACKGROUND More CHARACTER- than Percent Weighted IST IC Never Once once to ta l N SECTOR Co lombo 61.5 8.3 30.2 i00 551 Other Urban 54.4 8.1 37.5 I00 394 Rura l 64.7 6.0 29.3 i00 4553 Es ta tes 61.9 11.8 26.3 i00 366 ZONE Zone 1 61.5 8.3 30~2 i00 551 Zone 2 62.2 6.0 31.8 i00 904 Zone 3 58.9 7.5 33.6 i00 822 Zone 4 65.4 5.8 28.9 I00 1327 Zone 5 65.0 5.9 29.1 100 1167 Zone 6 63.1 9.4 27.6 i00 402 Zon~ 7 66.6 7.1 26.3 i00 691 EDUCATION No educat ion 78.3 5.3 16.5 i00 657 Pr imary 72.8 6.0 21.2 i00 1739 Secondary 59.5 7.0 33.5 i00 2123 More than secondary 50.7 7.8 41.5 i00 1346 Tota l 63.5 6.7 29.8 i00 5865 76 TABLE 4.24: Percentage of ever -marr ied women be l iev ing that i t is acceptab le to have messages about fami ly p lann ing on the radlo, by age and se lec ted background character i s t i cs , SLDHS 1987. BACKGROUND AGE CHARACTER- IST ICS 15-19 20-24 25-29 30-34 35-39 40-44 45-49 Tota l SECTOR Co lombo 85.7 90.7 86.5 85.8 92.1 90.2 90.3 89.1 Other Urban 83.0 96.7 96.2 94.8 91 .6 92.7 87.9 93.1 Rura l 81.7 86.8 90.8 91 .5 89.7 87.0 84.3 88.7 Es ta tes 76.2 83.2 88.7 83 .5 82 .6 78.7 77 .8 82.9 ZONE Zone 1 85.7 90.7 86.5 85.8 92.1 90.2 90.3 89.1 Zone 2 81.2 87.6 94.9 91.9 93.2 88.9 87 .0 91.1 Zone 3 i00.0 88.8 90.0 90.7 86.4 95.3 90 .1 90.4 Zone 4 72.0 81.6 88.6 91.5 86.4 88.4 80 .5 86.5 Zone 5 85.3 94.4 93.4 93.9 94.3 80.1 86.3 91.0 Zone 6 85.3 87.6 93.4 93.3 90.2 85 .0 82.4 89.6 Zone 7 78.4 81.2 85.6 84.4 85.4 81.2 77.4 83.0 EDUCATION NO educat ion 69.1 79.0 78.2 81.3 74 .5 81.1 71.9 77.2 P r imary 81.1 80.4 85.8 83.5 88.7 84.1 82.5 84.4 Secondary 83.5 91.1 92.9 95.0 91.4 89.0 92.9 92.1 More than sec. 100.0 94.5 96.0 94.2 93.9 94.7 92.2 94.5 Tota l 81.5 87.3 90.6 90.7 89.6 87.4 84.7 88.7 77 CHAPTER 5. Fert i l i ty Preferences 5.1 Introduct ion This chapter addresses three quest ions regarding the stated preferences of women for future chi ldbearing: i) How many more chi ldren are desired? 2) What is the desired t iming for having the next chi ld? 3) How many chi ldren are considered ideal for a family? Answers to these quest ions are des igned to provide an assessment of the need for family p lanning contracept ive services. The SLDHS also examines two further issues: l) To what extent do unwanted or mist imed pregnancies occur? 2) What ef fect would the prevent ion of such pregnancies have on the fert i l i ty rate? Given that the expl ic i t ly stated populat ion po l icy o f the Government of Sri Lanka is to provide the services necessary for couples to have the number of chi ldren they desire at the t ime they desire, understanding chi ldbear ing preferences has important program implications. Interpretat ion of data on fert i l i ty preferences, however, has always been the subject of controversy. Survey quest ions have been cr i t ic ized on the grounds that answers are mis leading because: i) they ref lect ephemeral v iews which are held with weak intensity and l i tt le conviction; and 2) they do not take into account the effect of social pressures or the att i tudes of other family members, part icu lar ly the husband, who may exert a major inf luence on reproduct ive decisions. The f irst object ion probably is less important in Sri Lanka than in other countries. The high contracept ive prevalence rate can probably be taken as an indicator of a strong and conscious des i re to control reproduct ive behavior. Thus it is not unreasonable to attach some interpret ive weight to the f indings concerning the stated chi ldbear ing preferences of women in Sri Lanka. Furthermore, unl ike most previous surveys, the SLDHS attempts to measure the intensity of v iews by conf i rming the des i re to have or not to have another child. The second object ion is correct in principle. In practice, 78 however, its importance is doubtful. For instance, the evidence from surveys in which both husbands and wives are interviewed suggests that there is no radical d i f ference between the views of the two sexes. The inc lus ion of women who are cur rent ly pregnant compl icates the measurement of v iews on future chi ldbearing. For these women, the quest ion on desire for more chi ldren is rephrased to refer to desire for another child, after the one they are current ly expecting. To take into account pregnant women, the results are c lassi f ied by number of l iv ing children, including the current pregnancy as equivalent to a l iving child. In addition, the answers of pregnant women on preferred wait ing t ime before the next birth presumably include the remaining gestat ion per iod of the current pregnancy and are thus not str ict ly comparable with the answers of non-pregnant women. Women who have been ster i l ized for contracept ive purposes also require special analyt ic treatment. The general strategy in this chapter is to c lassi fy them as want ing no more children. The va l id i ty of this assumption can be ascerta ined by referr ing to the d ist r ibut ion of answers to the special quest ion for ster i l ized women on change of mind since the operation. 5.2 Desire for more chi ldren Data concerning the desire for more chi ldren and the certa inty of preferences among current ly marr ied women in the sample by the number of l iv ing chi ldren is g iven in Table 5.1. This table is based on a series of quest ions which were designed to obtain information on the att i tudes of women towards having another child. About 29 percent say they def in i te ly want another chi ld whi le 31 percent say they def in i te ly do not want another child. Overall, those who state that they do not want more children, whether they are certain of their desire or not, plus women who have no regret about being ster i l ized const i tute 59 percent of all current ly marr ied women. Among all current ly marr ied women, 2.7 percent regret being ster i l ized and want more children. Another 1.5 percent regret being ster i l ized but are unsure whether they want more children. Whi le these percentages appear small in re lat ion to all current ly marr ied women, when the denominator is all s ter i l i zed women, 14 percent expressed some regret about being steri l ized. As one might expect, among women with no l iving children, 91 percent def in i te ly want a chi ld in the future, and among women with one l iv ing child, 76 percent want another. Thereafter, the percent want ing another chi ld drops sharply to 29 percent among women who current ly have two chi ldren and to 9 percent among women who current ly have three children. Conversely, the 79 TABLE 5.1: Percent d i s t r ibut ion of cur rent ly marr ied women aged 15-49 by des i re for more ch i ld ren and the cer ta in ty of the i r p re ference , accord ing to number of l i v ing ch i ldren , SLDHS 1987. DESIRE FOR MORE CHILDREN AND CERTAINTY OF PREFERENCE NDqMBER OF L IV ING CHILDREN * None 1 2 3 4 5 6+ Total HAVE ANOTHER: Def in i te ly 91.3 75.5 29.1 9.1 4.7 1.2 0.8 29.1 Not sure 1.9 2.7 2.5 0.7 0.5 0.7 0.0 1.5 UNDECIDED 3.5 6.3 10.2 4.4 2.1 1.5 0.7 5.2 WANT NO MORE Def in i te ly 1.2 10.5 32.7 38.9 39.2 40.0 46.0 30.7 No sure 0.4 2.8 7.1 3.2 2.1 2.1 1.2 3.5 NOT ESTABL ISHED 0.0 0.5 0.7 0.9 1.8 2.1 0.8 0.9 STERIL IZED: Regret , want more 0.8 0.8 3.8 3.8 3.5 2.5 0.9 2.7 Regret , no more 0.4 0.3 0.8 2.3 2~3 2.2 3.3 1.5 No regret 0.6 0.8 13.1 36.7 43.8 47.8 46.5 25.0 Not s ta ted 0.0 0. i 0.4 0.7 1.5 1.5 0.8 0.7 Tota l I00 i00 i00 i00 100 100 i00 i00 Weighted N 320 1014 1303 1138 704 461 503 5442 * Inc ludes cur rent p regnancy proport ion who def in i te ly do not want another increases rapidly from ii percent among women with one l iv ing chi ld to 33 percent among women with two children. At par i ty three and higher, over 85 percent of women are ei ther ster i l ized or state that they want no more chi ldren. Table 5.2 examines the desire for more chi ldren and the des i red t iming of the next birth among current ly marr ied women. The marginal percentages in this table are somewhat di f ferent from those in the previous table because women are c lass i f ied accord ing to their desired t iming for the next birth. This table al lows for an examinat ion of the total potent ia l need for contracept ive services. Unti l very recently, family planning programmes in Sri Lanka focussed on women who want no further chi ldren and thus are potent ia l c l ients for steri l izat ion. Spacing methods have not been widely used as was pointed out in Chapter 4. This s i tuat ion has now begun to change. The importance of spacing has been reinforced by recent evidence that short birth intervals are harmful to the wel fare of children. 80 F igure 5.1 summar izes the f ind ings on fe r t i l i ty preferences. A remarkably high 35 percent of current ly marr ied women want no more chi ldren whi le another 30 percent are already steri l ized. Together, these two groups const i tute 65 percent of all current ly marr ied women. The 35 percent who are not ster i l i zed but say they want no more chi ldren are potent ia l c l ients for ster i l izat ion. The 18 percent of all current ly marr ied women who want a chi ld after two or more years are either current ly us ing a contracept ive method or are potent ia l c l ients for spacing methods. Table 5.2 also shows that among current ly marr ied women with no children, 82 percent want a chi ld wi th in the next two years, however, this percent drops sharply to 23 among women with one chi ld and to only 9 percent among women with two children. Among women with one child, 55 percent want to delay the next birth two or more years. The re lat ionship between fert i l i ty preferences and par i ty is graphica l ly shown in Figure 5.2. In Table 5.3 the desire for more chi ldren is examined in re lat ion to current age of women. As age increases, there is a monotonic increase in the percent of current ly marr ied women who want no more children, and a monotonic decrease in the percent who want another child. These re lat ionships are best seen in Figure 5.3 Table 5.4 examines the women who want no more children, inc luding ster i l ized women, by selected background var iab les and number of l iv ing children. The percent who want no more chi ldren or who are already ster i l i zed increases rapidly from 16 percent at par i ty one to s l ight ly over 60 percent at par i ty two and reaches a peak of about 96 percent at par i ty four and higher. TABLE 5.2: Percent d i s t r ibut ion o f cur rent ly marr ied women aged 15-49 by des i re fo r more ch i ld ren , accord ing to number o f l i v ing ch i ldren, SLDHS 1987. DES IRE FOR MORE CHILDREN NUMBER OF L IV ING CHILDREN * None 1 2 3 4 5 6+ Tota l Wants no mote 1.9 1"4.0 42.3 43.4 41 .5 42 .5 47 .3 35.3 s ter i l i zed 1.8 2 .0 18 .1 43 .4 51 .1 53 .9 51 .5 29.8 Wants another w i th in 2 yrs. 81.8 22 .9 9 .1 2 .6 2.0 0 .5 0.8 12.2 Wants another a f te r 2+ yrs. 7 .6 55 .1 23 .4 7 .6 3.2 1.5 0 .0 18.4 Wants another D.K. when 4 .6 2 .5 i . i 0 .6 0.3 0 .0 0.0 1.2 Uns~lre o f p re ferences 2.4 3.2 5.7 2 .0 1 .6 I. I 0 .5 2.9 Not es tab l i shed 0 .0 0.3 0.3 0.3 0.3 0 .6 0 .0 0.3 Tota l 100 100 i00 100 i00 100 I00 i00 Weighted N 320 1014 1303 1138 704 461 503 5442 * Inc ludes cur rent ly p regnant women Among Figure 5.1 Ferfilify Preferences All Currently Morried Age 1 5 -49 Women Undecided Want c. 4% (Within , " 1 erilized 30% Want Later (After 2 Years) 18% Want No More .35% Sri Lanka DHS 1987 Among Figure 5.2 Fertility Preferences by Parity Currently Married Women Age 15-49 Percent 100% 80% 60% 40% 20% 0% 0 1 2 3 4 Number of Living Children 5 6+ [-----] Undecided Want Soon Want Later Want No More Sterilized Sri Lanka DHS 1987 TABLE 5.3: Percent d i s t r ibut ion of cur rent ly marr ied women age~ 15~49 b~ des i re for more ch i ldren, accord ing to age, SLDHS 1987. DESIRE FOR MORE CHILDREN CURRENT AGE 15-19 20-24 25-29 30-34 35-39 40-44 ~5-49~-Tota l Wants no more 9.6 19.4 26.9 33.3 36.1 46.9 63.3 35.3 S ter i l i zed 1.0 7.8 22.1 34.1 42.0 41.6 29 .6 29.8 Wants another w i th in 2 yrs. Wants another a f ter 2+ yrs. 24.1 17.7 14.9 12.3 11.6 7.7 4.3 12.2 57.7 48.9 31.4 15.9 5.9 1.3 0. i 18.4 QO Wants another D.K. when 3.2 3.0 0.7 0.6 1.6 0.5 0.9 1.2 Unsure of p re ferences 3.7 2.9 3.9 3.7 2.7 i. 5 i. 1 2.9 Not es tab l i shed 0.7 0.3 0. i 0.2 0.2 0.5 0.6 0.3 Tota l i00 i00 i00 I00 i00 i00 I00 i00 Weighted N 127 695 1071 1146 1044 824 535 5442 Figure 5.3 Fertility Preferences by Age Among Currently Married Women Percent 100% 80% 60% 40% 20% 0% 15-19 20-24 25-29 30-34 35-39 Age Group 40-44 45-49 Undecided Wont Soon Want Later Wont No More Sterilized Sri Lanka DHS 1987 TABLE 5.4: Percent of cur rent ly marr ied women who want no more ch i ld ren ( inc lud ing s ter i l i zed women) by number of l i v ing ch i ld ren* and se lec ted background character i s t i cs , SLDHS 1987. BACKGROUND CHARACTER- IST ICS NUMBER OF L IV ING CHILDREN None 1 2 3 4+ Tota l SECTOR Co lombo 0.0 19.4 70.3 92.9 97 .1 66.2 Other Urbah 0.0 11.8 69.0 89.4 98.2 68.6 Rura l 3.5 15.7 57.6 85.6 95 .0 64.5 Es ta tes 11.8 17.0 65.7 89.8 97.9 66.8 ZONE Zone 1 Zone 2 Zone 3 Zone 4 Zone 5 Zone 6 Zone 7 0.0 19.4 0 0 16.0 5 9 17.6 5 1 15.2 2 5 16.4 3 0 12.3 9 6 12.6 70.3 92 67.7 93 64.0 84 55.2 84 57.5 90 50.0 76 51.8 79 .9 97 .1 66.2 3 98 .5 65.6 9 96.4 64.1 6 94.2 63.5 2 96 .0 65.7 4 95 .1 65.3 3 93.3 66.7 EDUCATION No educat ion 9.4 22 .1 63.7 83.2 95 .0 72.7 P r imary 3.2 18.3 60.3 89.4 95.9 75 .1 Secondary 4.1 15.5 59.3 88.5 95 .6 63.9 More than sec. 1.8 14.1 61.0 82 .1 94.5 51.3 Tota l 3.6 16.0 60.4 86.9 95 .5 65.1 * Inc ludes cur rent p regnancy Th is genera l pat tern of inc rease is ev ident zones, and educat iona l leve ls . Indeed, the background character i s t i cs are re la t ive ly smal l . for a l l sectors , d i f fe rent ia l s by 5.3 Women in need of fami ly p lann ing Tab le 5 .5 examines the need fo r fami ly p lann ing cont racept ive serv ices among cur rent ly marr ied women. Women are cons idered to be in need if they are fecund and not cont racept ing and e i ther want no more b i r ths or want to postpone the next b i r th for two or more years . Inc luded in th i s de f in i t ion are some women who are not immediate ly at r i sk of a p regnancy and thus not immediate ly in need of cont racept ive serv ices , i .e. , women who are cur rent ly p regnant , not menst ruat ing , or not sexua l ly act ive. The women inc luded in the numerator for the percentages in th is 86 table are now or wi l l be in the near future in need of family p lanning services in order to avoid an unwanted or unplanned pregnancy. Overall , Table 5.5 indicates that 23 percent of current ly marr ied women are in need of family p lanning services. E leven percent or about hal f of the 23 percent in need say they want no more ch i ldren whi le approx imate ly 12 percent say they want to postpone the next b i r th or they are undec ided about the next birth. The second half of Table 5.5 examines the percent of women who are in need and who also say that they intend to use a method of contracept ion in the future. Twelve percent of the current ly marr ied women fall into this category and these women are equal ly d iv ided between women who want no more ch i ldren and women who want to postpone the next b i r th or are undecided. The former represent women who wi l l probably accept s ter i l i zat ion whi le the latter are women who wi l l probably use spacing methods. TABLE 5.5: Among cur rent ly marr ied women, the percent who are in need of fe~nity p lann ing* , and the percent who are in need and p|an to use a cont racept ive method in the fu ture , by background character i s t i cs , SLDNS 1987. IN NEED OF FAMILY PLANNING IN NEED AND INTENDS TO USE CONTRACEPTION BACKGROUND Wants no Wants to Total Wants no Wants to Total CHARACTER- more postpone/ in more postpone/ in ]STICS ch i ld ren undecided** need ch i ld ren undecided** need SECTOR CoLombo 11.0 11.9 22.9 5.4 53 10.7 Other Urban 7.9 7.1 15.0 3.8 1.1 4.9 Rural 11.4 11.7 23.1 6 .7 6.3 12.9 Estates 10.7 14.3 25.0 5.7 6.2 11.9 ZONE Zone 1 11.0 11.9 22.9 5.4 5.3 10.7 Zone 2 9.1 9.1 18.2 4.8 4.5 9.3 Zone 3 12.0 9.5 21.5 7.1 3.7 10.7 Zone 4 9.5 11.5 21.0 5.2 6.1 11.3 Zone 5 12.6 12.5 25.1 8.3 6.5 14.8 Zone 6 10.5 12.8 23.3 5.6 7.7 13.3 Zone 7 13.6 15.2 20.8 7.3 7.9 15.2 EDUCATION No educat ion 12.1 11.3 23.4 6.8 5.3 12.1 Primary 11.4 9.6 21.0 6.3 4.4 10.9 Secondary 11.5 12.2 23.7 7.0 6.2 13.2 Nore than sec. 9 .6 13.2 22.8 4.9 7.2 12.2 Total 11.1 11.6 22.7 63 5.8 12.1 * Women in need are def ined as fecund, not cur rent ly contracept ing and who want no more b i r ths or ~ant to postpone the next b i r th for at Least two or more years, ** IncLudes undecided about whether to have another b i r th or about t iming for the next b i r th . 87 By sector, zone, and educat ional level the d i f ferent ia ls in Table 5.5 are re lat ive ly small. Fewer women in other urban areas are in need of fami ly p lanning services than women in other sectors and fewer say that they plan to use contracept ion in the future. This is probably because more of these women are a lready current ly us ing a method of contracept ion. 5.4 Ideal number of ch i ldren One indicator of family size norms in Sri Lanka is the mean ideal number of ch i ldren desired. In the SLDHS, the ideal number of ch i ldren des i red was obtained by asking one of two questions. Women who had no l iv ing chi ldren were asked "If you could choose exact ly the number of ch i ldren to have in your whole life, how many would that be?" For women wi th ch i ldren the quest ion was var ied sl ightly, "If you could go back to the t ime you did not have any ch i ldren and could choose exact ly the number of ch i ldren to have in your whole life, how many would that be?" Table 5.6 shows the responses to these questions. The mean ideal number of ch i ldren desired, presented at the bot tom of the table, is identical for ever -marr ied and current ly marr ied women. On the average, women in Sri Lanka say they would l ike to have 3.1 children. This number is a lmost exact ly equal to the 3.0 mean number of ch i ldren ever born to ever marr ied women 15-49 reported in Chapter 3. It is a lso a drop from the 3.8 mean ideal number of ch i ldren reported in 1975 by TABLE 5.6: Percent d i s t r ibut ion of ever -marr ied wo~en by ideal number of ch i ld ren according to number of l i v ing ch i ld ren ( inc lud ing any cur rent pregnancy), SLDHS 1987. IDEAL Al l NUMBER OF NU~IBER OF LIV]NG CH]LDREN ever CHILDREN married DESIRED None 1 2 3 4 5 6+ women None 0.3 0.2 0.0 0.1 O.O 0.0 0.2 0.1 1 3.9 8.8 2.3 1 .~ 0.7 0.1 0.5 2.9 2 54.4 54.4 45.2 1el.5 17.5 10.0 4.6 31.1 3 26.1 24.2 35.5 54.8 23.3 28.8 1;'.5 32.9 4 7.8 5.5 11.1 15.5 38.5 18.3 21.5 15.9 5 1.0 1.4 2.5 5.4 7.2 21.5 17.2 6.4 6 + 1.0 0.6 0.6 1.5 4.0 8.4 14.8 3.3 Non-numer ic 5.6 5.0 2.9 5.0 8.9 12.9 23.9 7.5 answer Total Percent 100 100 100 100 100 100 TOO 100 Mean idea[ NO. for ever -marr ied 2.5 2.4 2.7 3.1 3.5 3.9 4.5 3.1 Mean ideal No. for cur rent marr. 2.5 2.4 2.7 3.1 3.5 4.0 4.5 3.1 Weighted N 349 1115 1368 1204 768 502 558 5865 88 TABLE 5.7: Mean ideal number of ch i td ren for ever-marr ied vonen by cur rent age and setected background character i s t i cs , SLDHS 1987. BACKGROUND CURRENT AGE CHARACTER- AlL 15TIES 15"19 20-2& Z5-29 30-34 35-39 &O-&& ~&5-&9 ages TOTAL SECTOR CoLombo 2,6 2.2 2.4 2.5 2.9 3.1 3 .7 2.8 2.8 Other Urban 2.1 2.4 2.4 3.1 2.9 3.4 3.8 3.0 3.0 Rurat 2.5 2.6 2.8 3.1 3.2 3.5 3.7 3.1 3.1 Estates 2.6 2.8 3.0 3.0 3.3 3 .7 4.5 3.3 3.3 ZONE Zone 1 2.6 2.2 2.4 2.5 2.9 3.1 3 .7 2.8 2.8 Zoee 2 2.1 2.4 2.5 2.9 2.7 3.1 3.1 2.8 2.8 Zone 3 2.1 2 .4 2.5 2.8 2.8 3.2 3.6 2.9 2.9 Zone 4 2.6 2.7 2.9 3.0 3.2 3 . ; ' 3.8 3.1 3.1 Zone 5 2.6 2.6 2.8 3.2 3.4 3.7 4.2 3.~ 3.2 Zone 6 2.7 2.9 3.0 3.4 3.5 4.0 4.1 3.4 .~.4 Zone 7 2.6 2.8 3.0 3.4 3.6 3.8 4.0 3.3 3.3 EDUCAT IOM No educat ion 2.9 2.7 3.0 3.5 3.7 3.9 4.5 3.5 3.5 Primary 2.6 2.6 2.9 3.3 3.4 3 .8 4.1 3.3 3.3 Secondary 2.4 2.6 2.7 3.0 3.0 3.2 3.3 2.9 2.9 More than Sec. 2,1 2.4 2.5 2.7 2.7 3.1 3.1 2.7 2.7 TOTAL 2.5 2.6 2.7 3.0 3.1 3.5 3.6 3.1 3.1 the SLWFS. For those women with fewer than three chi ldren, the mean ideal number is s l ight ly h igher than the actual number of chi ldren. For women with more than three chi ldren, the mean ideal number is less than the actual number. Di f ferences by current age, sector, zone, and educat ional level in the mean ideal number of ch i ldren des i red are shown in Table 5.7. This table indicates that younger women are more l ike ly than older women to cons ider between 2 and 3 ch i ldren as ideal. In all sectors, zones, and educat ional levels, women under age 30 have a mean ideal number of ch i ldren be low the nat ional average of 3.1 whi le women over 30 tend to have an ideal that is h igher than the nat ional average. By sector, women in the capitol c i ty of Colombo have the lowest mean ideal number of ch i ldren whi le women on the estates have the highest. By zone, the d i f ferences vary from an ideal of 3.4 in zone 6 to a low of 2.8 in zones 1 and 2. By educat ional level, there is a cons istent dec l ine wi th in any g iven age groUp in the mean ideal number of ch i ldren des i red as educat ion increases. 5.5 Fert i l i ty p lanning and the status of b i r ths In the SLDHS, women were asked a ser ies of quest ions for each chi ld born in the last f ive years and any current pregnancy to determine whether the par t icu lar pregnancy was planned, unplanned but wanted at a later time, or unwanted. The purpose 89 of these quest ions is to gauge the degree to which couples successfu l ly control their ferti l ity. The quest ions require the respondent to recall accurately her wishes at one or more points in the past five years. There is of course always the poss ib i l i ty of rat ional izat ion, an unwanted pregnancy may well become a cher ished child. Table 5.8 is a b i r th-based rather than a woman-based table. It shows the percent of all b irths in the past five years ( including any current pregnancy) that were wanted among women who were not us ing contracept ion during the interval and among those who were using contraception. The response categor ies l isted under the heading "contracept ive interval" require some 47 TABLE 5.8: Percent d i s t r ibut ion of a l l b i r ths ( inc lud ing current pregnancy) in last f i ve years by contraceptive pract ice of mother and whether b i r th was wanted, by b i r th order, SLOHS 1987. CONTRACEPTIVE PRACTICE AND DESIRE FOR BIRTH BIRTH ORDER 1 2 ] 4+ Al l b i r ths Non Contraceptive In terva l Wanted b i r th then 76. ] 36.0 35.7 28.2 46.4 Wanted b i r th la ter 9.4 16.9 15.7 13.7 13.6 Did not want b i r th 0.7 2.5 5.4 20.0 6.4 Contraceptive ln tervs l Wanted b i r th then 8.4 22.8 22.9 11.4 15.8 Wanted b i r th tater 2.4 16.1 13.4 8.4 9.6 Did not want b i r th 0.5 ] .8 7.3 16.9 6.3 Desire not stated 2.4 1.9 1.5 1.4 1.9 Total percent 100 100 100 100 100 Weighted N 1380 1214 856 975 4424 explanation. A birth that fal ls into the category "wanted birth later," or the category "did not want birth" represents a contracept ive failure. That is to say, these are births to women who said they were using a contracept ive method when they became pregnant. A birth that falls into the category "wanted birth then" represents a b i r th to a woman who was us ing contracept ion dur ing the interval but stopped for the purpose of becoming pregnant. Among all 4424 births in the last five years, 62 percent were wanted then, 23 percent were wanted later, and 13 percent were not wanted. Women who have four or more l iving chi ldren are 90 the most l ikely to report an unwanted birth. A lmost 16 percent of the births occurred to women who said they were using a method when they became pregnant. Whether these births are due to true method fai lures or to inappropr iate method use cannot be determined from the data. Table 5.9 is s imi lar to the previous table but restr icted to births that occurred in the past 12 months. Since the information in this table is l imited to 12 months, it is probably a more rel iable and certain ly a more current est imate of unwanted and mis t imed pregnancies. The estimates, however, are remarkably s imi lar to those obtained for the five year period. Overall , 61 percent of the 814 births in the last 12 months were wanted, 26 percent were Wanted l~ter, ~n~ ~bout i3 percent were not wanted. Almost 70 percent of f irst and second order births are wanted at the time of the birth compared to 47 percent of third or h igher order births. Third or h igher order births are far more l ikely to be unwanted than first or second order births. TABLE 5.9: Among women having a b ir th in the last 12 months, the percentage want ing a chi ld then, later, or want ing no more children, by birth order, SLDHS 1987. DESIRE FOR LAST BIRTH BIRTH ORDER OF CHILD 1-2 3+ Total Wanted then 69.6 47.1 60.7 Wanted later 23.8 28.2 25.6 Not Wanted 5.4 23.4 12.5 Not stated 1.2 1.2 1.2 TOTAL I00 i00 i00 493 320 814 Table 5.10 presents another indicator of wanted ferti l ity. In this table, the impact of prevent ing unwanted births can be est imated by computing total wanted fert i l i ty rates. These rates express the level of fert i l i ty that theoret ica l ly would result if all unwanted births were prevented. Compar ison of actual rates with wanted rates indicates the potent ia l demographic impact of prevent ing unwanted births. 91 TABLE 5.10: Total wanted fert i l i ty rates and total fert i l i ty rates for all women 15-49 for five years preceding the survey, by background character ist ics, SLDHS 1987. Total Fert i l i ty Rates Real Wanted SECTOR Colombo 2.2 1.9 Other Urban 2.3 1.9 Rural 2.9 2.4 Es ta tes 3.4 3.2 ZONE Zone 1 2.2 1.9 Zone 2 2.4 2.0 Zone 3 2.6 2.1 Zone 4 2.7 2.4 Zone 5 3.2 2.8 Zone 6 3.0 2.4 Zone 7 3.4 2.6 Total 2.8 2.4 The wanted fert i l i ty rates are calculated in exact ly the same manner as the convent ional age-speci f ic fert i l i ty rates presented in Chapter 3, except that births c lass i f ied as unwanted are omitted from the numerator; the remainder can be cumulated to form a wanted total fert i l i ty rate which is analogous to the convent ional total fert i l i ty rate. The wanted fert i l i ty rate can be interpreted as the total number of wanted births a woman would bear by age 50, if she exper ienced the wanted age specif ic fert i l i ty rates for the past 5 years. The wanted fert i l i ty rate may be a better indicator of desired fert i l i ty than answers to quest ions concerning ideal family size. The answers respondents give to quest ions about wanted births presumably take into account the balance between sons and daughters a lready born and s
Looking for other reproductive health publications?
The Supplies Information Database (SID) is an online reference library with more than 2000 records on the status of reproductive health supplies. The library includes studies, assessments and other publications dating back to 1986, many of which are no longer available even in their country of origin. Explore the database here.