Harar Bulletin of Health Sciences Extracts Number 4, January 2012

Determinants of Infant Mortality in Kersa District, Eastern Ethiopia: A Case Control Study
Kedir Teji

Abstract
Background: Infant and child mortality rates have long been used as indicators of the level of socio-economic development of a country. Various studies have been conducted to show the factors affecting infant and child mortality in both developed and developing countries. These factors are socio-demographic, socio-economic and environmental with examples including ethnicity, levels of education of the mother and father, housing conditions, crowding, the availability of latrines and early termination of breastfeeding. Objective: To assess the determinants of infant mortality in Kersa District, Oromia Region, eastern Ethiopia from January, 2010 to May, 2011. Methods: A case control study was conducted among 200 cases and 800 controls. Cases were defined as all deaths within the first year of life. As controls, there were four selected children born within one week of a deceased baby who was of the same gender, living in the same residence area that did not die within their first year of life. Socio-demographic, socio-economic and environmental factors were assessed for all live births in the 2007-2010 period. Completed livebirth registrations of babies with those mothers who lived in the
79

Harar Bulletin of Health Sciences

district were used. Multiple variable analyses were performed using logistic regression. Results: Eighty-five (42.5%) of the cases were female and 115 (57.5%) were male. All deaths 200 (100%) were registered among Muslims. They were Oromo by ethnicity 198 (99%) and the majority was married 176 (88%). Gestational age of less than 37 weeks, illiteracy among mothers, seeking health care outside of a health care facility, a monthly income of less than ETB500, an unprotected source of water, unsanitary waste disposal, mothers age of less than 20 years at current delivery, and limited spacing between the mothers previous child were found to be predictors of infant mortality after controlling for possible confounders. Conclusions and Recommendations: Many predictors of infant mortality were identified. The illiteracy of the mothers, early age of delivery, and unsanitary waste disposal were strongly associated with infant mortality among others. Improved maternal and child education of pregnant women and mothers is needed. Women need to be encouraged to delay marriage and pregnancy. Spacing of children should be encouraged. Birth control methods should be provided on request to women. Improved environmental sanitation, a safe water supply, and better service availability for reproductive health are recommended. Finally, delivery in a health facility by a qualified health care provider to improve the infant mortality rate is recommended.

80

Harar Bulletin of Health Sciences

Introduction
Socio-economic development and health conditions of a country can be determined by infant and child mortality rates. Data indicate that eleven million children under the age of five die annually in the world of which over ten million are in the developing world. (1) According to a United Nations Report (2009), the infant and under-five mortality rates of the world were 49.4 and 73.7, respectively. (2) Various studies have been conducted to identify the factors affecting child mortality in both developed and developing countries. These factors have been associated with demographic, socioeconomic and environmental factors such as ethnicity, housing conditions, crowding, the availability of latrines, and early termination of breastfeeding. (3, 4) There is limited research conducted on child mortality in eastern Ethiopia. Most of the information for any program planning and implementation has been based on Ethiopian Demographic Health Survey (EDHS) conducted every five years. EDHS describes only the rate of mortality and does not provide information on the causes of mortality. This study provides information about the causes of infant mortality in the Oromia region, eastern Ethiopia. (5) Given the high priority to child mortality in Ethiopia, an understanding of the differentials and causes of infant mortality is needed. The study results serve as baseline data for further investigations and to learn more about the reasons for the persistently high levels of infant mortality. It provides input for health planners and policy makers. The aim of this study was to assess the
81

Harar Bulletin of Health Sciences

determinants of infant mortality in Kersa District, Oromia Region, eastern Ethiopia from January, 2010 to May, 2011.

Methods
Study Area and Period: The Kersa Demographic Surveillance and Health Research Center (KDS-HRC) is located in the eastern Hararge zone and the Oromia Regional State. This study was conducted from January 15, 2011 to May 15, 2011. Study Design: A community-based, case-control study design was used. Source Population: The source population included all infants born or were born and died located in Kersa District from October, 2007 to December, 2010. Study Population: The cases were all infants who died between October 1, 2007 and December 30, 2010. The controls were infants who were registered in the KDS and were alive at the time their respective cases died. The control group was neighboring households whose identification numbers were nearer to the case in the kebele where live infants of 1 weeks age variation were included in study. Sample Size Determination: The sample size was calculated using Epiinfo 3.5.1 and a consideration of the proportion of poor mothers among the controls of 74% which was the main exposure variable estimated from another study. (6) A 95% CI, a 90% power of the study and a control to case ratio of 4:1 to detect an odds ratio of 2 was determined. Accordingly, 200 cases and 800 controls for a total sample size of 1000 were included from 12 kebeles.

82

Harar Bulletin of Health Sciences

Sampling Technique: Cases that were registered in the data base with complete information were included. The controls were infants randomly selected by age variations within less or greater than one week of birth of the case baby in the same kebele. Data Collection: Data were collected from the KDS Office which had the registrations of all data concerning births, deaths, families and residential housing conditions in the kebeles.

Study Variables
Dependent Variable: Infant mortality was the dependent variable. Independent Variables: Socio-economic, demographic, mother and infant-related variables were the independent variables. Data Quality Management: One supervisor with a degree in computer sciences was assigned to supervise data gathering. A data manager trained for two days was also assigned. She checked each completed questionnaire and conducted data entry. Supervision was performed by the principal investigator on daily basis. Data Analysis: Data were entered using Epi-data software. The data were then exported to SPSS Version 16. Frequencies and cross tabulations were computed to check for consistency and to identify missing values. Bivariate and multivariate analyses were calculated to identify the risk factors associated with infant mortality using logistic regression. Ethical Considerations: Ethical clearance was obtained from the Institutional Research Ethics Review Committee on the Harar campus of Haramaya University. Data collected did not contain
83

Harar Bulletin of Health Sciences

names or identifiers which might identify the study participants. Data were collected with the consent of the participants and community. The study was explained to the subjects. Confidentiality was assured. They were informed that the information they gave would not be used for any purpose other than the study.

Results
Eighty-five (42.5%) of the cases were girls and 115 (57.5%) were boys. The majority 196 (98%) of the deaths occurred in residential homes, two (1%) deaths occurred in the health center and one death (0.5%) in the hospital. Out of a sample of 1000 study subjects, nearly all 980 (98%) were Muslim followed by Orthodox Christian 14 (1.4%) and Protestant 3 (.3%). Oromo was the dominant ethnic group 982 (98.2%). Among the total population, 912 (91.2%) were married and monogamous, 40 (4%) were polygamous and 30 (3%) were widowed, divorced, or separated. To determine the role of factors effecting infant mortality, every variable with the infants death was studied separately by using a logistical regression method with SPSS Version 16. Sociodemographic, socio-economic and infant-related factors were analyzed.

84

Harar Bulletin of Health Sciences

Table 1. Determinants of Infant Mortality in Kersa District, Eastern Ethiopia. February, 2011 to March, 2011
Variables Marital Status Married Unmarried/ divorced/ widowed Mothers Education Literate Illiterate Number of Pregnancies <4 4 Family Expenditure in ETB >500 Case Control C.OR CI 95% 4.4(2.2-8) 1.00 Adjusted OR (95%CI) 1.00 3.6(1.7-7.7) 0.001 P- value

176 24

776 24

22 178

209 591

1.00 2.9(1.7-4.7)

1.00 3.57(2-7)

0.000

120 80

404 396

0.68(0.48-0.94) 1.00

0.76(0.5-1) 1.00

0.13

500
Care Seeking When Baby is Sick Health Institute Other place Mothers Age at First Pregnancy >20
20

83 117

246 554

1 3.2(2- 4.4)

1.00 2.3(1.5-3.5)

0.001

51 149

298 502

0.58(0.4-0.8) 1.00

0.53(0.35-0.8) 1.00

0.02

21 176

96 702

1.00 1.15(0.69-2)

85

Harar Bulletin of Health Sciences

Mothers Age at Current Pregnancy >20

20
Duration of Pregnancy 37 weeks < 37 weeks Place of Childs Birth Health Institution Home Delivery Attendant Health Professional TTBA TBA Previously Born Child Alive Yes No Main Type of Water Supply Protected Source Unprotected Source

163 37

735 64

1.00 2.61(1.6-4.1)

1.00 3.4(2-6)

<0.000

162 38

777 23

1.00 7.3(3.9-13.6)

1.00 5.6(3-11)

0.000

4 196

39 761

1.00 2.5(0.8-7.2)

1.00 1.7(.5-7)

0.4

3 58 139

36 219 545

1.00 .3(.09-1) .3(0.1-1)

1.00 2(.4-9.5) 1(.6-1.4)

0.3 0.9

143 57

727 57

.25(0.2-0.4) 1.00

.29(.18-.5)

0.000

83 117

468 332

.5(.4-.7) 1.00

0.52(.36-.8)

0.001

86

Harar Bulletin of Health Sciences

Refuse Disposal Method Pit/compost/ burning Dispose in open field /farm

40

448

0.196(0.1-.3)

0.2(.13-.3)

0.000

160

352

1.00

1.00

Adjusted for the variables in the same level and for those in Level I

Discussion
This study revealed that illiterate women had a 2.8 times greater chance of having a child die than literate women. Studies conducted in Tanzania and Zambia have shown similar findings. (7, 8, 9, 10) In India, literate women were 44% less likely to have a child die than were illiterate women. Literate women had a 50% lower infant mortality risk. When controlled for other factors, a child living in a neighborhood with more educated mothers had a fifty percent greater chance of survival. (11) One of the most important factors affecting both mother and infant survival was the age of the mother at delivery. Mothers less than age twenty at the birth of their babies had a greater risk of infant mortality when compared to those mothers whose ages were twenty years or greater. It is medically known that those mothers who give birth at less than 20 years of age- and especially before 18 years of age- are at much greater risk for both maternal mortality and infant mortality. Children born before 37 weeks of pregnancy were five times more likely to die than those who had a gestation period of 37 weeks or
87

Harar Bulletin of Health Sciences

more. Studies indicate that ninety percent of neonatal mortality is due to prematurity. With prematurity, death usually results from respiratory distress syndrome. Another study in Ethiopia showed that families who had a monthly income of less than ETB500 experienced a much greater infant mortality rate than those families with incomes of ETB500 and greater. Monthly income was predictive of child mortality. Studies conducted in Kenya showed that women classified as poor had three times the number of children die than those classified as rich. (9, 10) Again, income was a predictor for child mortality. Families who used protected water sources (ex. protected springs or wells and pipe lines) were 48% less likely to have an infant die when compared to those families who used unprotected sources (ex. ponds and lakes). Access to safe water and sanitation reduces child mortality risks in rural areas. (10) The absence of a safe water supply predisposes babies to gastrointestinal diseases which lead to severe dehydration and death. Families who sought care in modern health institutions had greater child survival rates when compared to those families who went to others places. Families who went to modern health care institutions had better health outcomes than those who went other places for care. Among those families who used proper refuse disposal (i.e. disposing their waste in pits, making compost or burning waste) were 80% less likely to lose their babies than those who used improper disposal methods (i.e. disposing refuse in open fields or on farms). Waste disposal was associated with an unclean environment. These environments were a source of morbidity and mortality.
88

Harar Bulletin of Health Sciences

Conclusions and Recommendations

Infants born to illiterate mothers aged less than 20 years were more likely to die. Proper waste disposal and the utilization of protected sources of water reduced the risk of infant mortality. A family monthly income of less than ETB500 led to higher infant mortality. This research clearly indicates the need for increased efforts of educating women in Ethiopia. Women who attend school marry later and therein have children later. Health education activities concerning reproductive health and prenatal care needs, delivering in a health facility, and improved environmental sanitation should be given greater attention by all concerned bodies of the woreda health bureau and woreda administration.

Acknowledgements
I would like to thank Haramaya University, my thesis research advisors, and my family including my sister Fayine Teji for their support.

References
1. Jones G, Steketee RW, Black RE, Bhutta ZA, Morris SS. How Many Child Deaths Can We Prevent This Year? Lancet. 2003; 362:32327. 2. United Nations. World Population Prospects Report for 2005-2010. New York, USA. 2011.
89

Harar Bulletin of Health Sciences

3. Hill K, Amouzou A. Child Mortality and Socio-economic Status in Sub-Saharan Africa. African Population Studies 2004;19(1):1-11. 4. DHS Survey. Central Statistical Agency, Addis Ababa, Ethiopia. ORC Macro. Calverton, Maryland. 2006. 5. Girma B, Berhane Y. Epidemiological Assessment of Determinants of Under-five Mortality in Jimma Town. Master of Public Heath thesis, Addis Ababa University, Addis Ababa, Ethiopia. Unpublished. 6. Brown CA, Sohani SB, Khan K, Lilford R, Mukhwana W. Antenatal Care and Perinatal Outcomes in Kwale District, Kenya. PLoS. 2006. doi: 10.1186/1471-2393-8-2. 7. Fahrmeir L. Analyzing Child Mortality in Nigeria with Geoadditive Discrete-time Survival Models. Statistics in Medicine 2010;24:709-728. 8. Fotso JC. Child Health Inequities in Developing Countries: Differences across Urban and Rural Areas. PLoS. 2006. doi: 10.1186/1475-9276-5-9. 9. Ezeh AC, Fotso JC, Madise NJ, Ciera J. Progress Towards the Child Mortality Millennium Development Goal in Urban Sub-Saharan Africa: Dynamics of Population Growth, Immunization, and Access to Clean Water. PLoS. 2007. doi: 10.1186/1471-2458-7-218. 10. Adhikari R, Podhisita C. Household Headship and Child Death: Evidence from Nepal. PLoS. 2010 doi: 10.1186/1472698X-10-13. 11. Kandala NB, Gebrenegus G. Spatial Modeling of Socioeconomic and Demographic Determinants of Childhood
90

Harar Bulletin of Health Sciences

Under-nutrition and Mortality in Africa: Geo-additive Bayesian Discrete-time. Shaker Verlag, New York, USA. 2008. 12. Rutherford ME, Dockerty JD, Jasseh M, Howie C, Herbison P, Jeffries DJ, Leach M, Stevens W, Mulholland K, Adegbola RA, Hill PC. Access to Health Care and Mortality of Children Under Five Years of Age in the Gambia: A Casecontrol Study. PLoS. 2009. doi: 10.2471/BLT.08.052175.

91