Menu
Menu
Menu
Menu
Background: Poorly spaced pregnancies have been documented worldwide to result in adverse maternal and child health outcomes. The World Health Organization (WHO) recommends a minimum inter-birth interval of 33 months between two consecutive live births in order to reduce the risk of adverse maternal and child health outcomes. However, birth spacing practices in many developing countries, including Tanzania, remain scantly addressed.Methods: Longitudinal data collected in the Rufiji Health and Demographic Surveillance System (HDSS) from January 1999 to December 2010 were analyzed to investigate birth spacing practices among women of childbearing age. The outcome variable, non-adherence to the minimum inter-birth interval, constituted all inter-birth intervals <33 months long. Inter-birth intervals ≥33 months long were considered to be adherent to the recommendation. Chi-Square was used as a test of association between non-adherence and each of the explanatory variables. Factors affecting non-adherence were identified using a multilevel logistic model. Data analysis was conducted using STATA (11) statistical software.Results: A total of 15,373 inter-birth intervals were recorded from 8,980 women aged 15-49 years in Rufiji district over the follow-up period of 11 years. The median inter-birth interval was 33.4 months. Of the 15,373 inter-birth intervals, 48.4% were below the WHO recommended minimum length of 33 months between two live births. Non-adherence was associated with younger maternal age, low maternal education, multiple births from the preceding pregnancy, non-health facility delivery of the preceding birth, being an in-migrant resident, multi-parity and being married.Conclusion: Generally, one in every two inter-birth intervals among 15-49 year-old women in Rufiji district is poorly spaced, with significant variations by socio-demographic and behavioral characteristics of mothers and newborns. Maternal, newborn and child health services should be improved with a special emphasis on community- and health facility-based optimum birth spacing education in order to enhance health outcomes of mothers and their babies, especially in rural settings. © 2012 Exavery et al.; licensee BioMed Central Ltd.
The Rufiji Health and Demographic Surveillance System (HDSS) is located in Rufiji district of the Coast region, 178 kilometres south of Dar es Salaam, Tanzania. A HDSS is a longitudinal, population-based health and vital events registration system that monitors demographic events such as births, deaths, pregnancies, in- and out-migrations and socio-economic status of a geographically well-definedsetting of individuals, households and residential units. The Rufiji HDSS was incepted in September 1998 from the Tanzania Essential Health Interventions Project (TEHIP) and as of 2010, it was made up of 33 villages with over 16,000 households in which more than 80,000 people resided. The area is mainly rural with a scattered population, though clustering around Ikwiriri, Kibiti and Bungu townships is known. The largest and original native ethnic group in the HDSS is Ndengereko. Others include Matumbi, Ngindo and Zaramo. In terms of religion, about 90% of the people are Muslim. Most people speak their ethnic languages, even though the national language, Kiswahili, is well understood and widely spoken. Further details about the study area are available [23]. This study is a secondary analysis of longitudinal data collected by the Ifakara Health Institute (IHI) in its Rufiji HDSS in Tanzania for a period of eleven years from 1st January 1999 to 31st December 2010. Access to the data was permitted by IHI, an institute that owns, manages and maintains the HDSS. The inception of the HDSS was approved by the Medical Research Coordinating Committee (MRCC) of the National Institute for Medical Research (NIMR) in Tanzania. This ethical approval is detailed elsewhere [24]. Data collection procedures of the HDSS require that every household is visited once every four months in order to update previously recorded household information and register new demographic events that may have occurred. Between household visits, community-based key informants report births and deaths as they occur. The Rufiji HDSS is an ongoing longitudinal population-based data generating platform. A particular focus of the current study was on analyzing inter-birth intervals in light of the WHO’s recommendation on birth spacing. Therefore, resident women of the Rufiji HDSS aged 15–49 years who were followed-up for vital statistics, particularly birth history, were of interest. As the focus of this study was on closed inter-birth intervals, only women who had given birth at least twice (i.e. multiparous) were retained for this analysis. Those who had experienced adverse outcomes in any of their two consecutive births were very few and excluded in this analysis to be analyzed separately in light of the second recommendation of the WHO on birth spacing after experiencing an adverse outcome. This study examined inter-birth interval as a dependent (outcome) variable against background characteristics of the mother and the child. The inter-birth interval was collapsed into two categories according to the WHO recommendation: (1) <33 months, which was referred to as “non-adherence” or poor birth spacing, and (2) ≥33 months, referred to as “adherence” or appropriate birth spacing. Independent variables investigated (with their categories in brackets) were (1) maternal age (broken into categories of 5 years interval size starting from 15–19 and ending with 45–49), (2) maternal education (secondary and higher, primary and never been to school), (3) maternal occupation (no job, self employment and formal employment), (4) marital status of the mother (married, single, ever married (i.e. divorced or widowed)), and (5) sex of the index child (female and male). Others were (6) place of residence (urban and rural), (7) number of births of the preceding pregnancy (singleton and multiple), (8) parity (2, 3 and ≥4), (9) place of delivery of the index pregnancy (health facility and elsewhere) and (10) HDSS entry type (enumeration and in-migration). During the start of the Rufiji HDSS, entry type of all people present at that time was enumeration. Entry into the HDSS area was also possible through birth or in-migration (migrating into the study area). No one of those who became members by birth was eligible for the current analysis because all were below 15 years of age throughout the follow-up period. Therefore the variable, HDSS entry type, had two categories only as enumeration and in-migration. An inter-birth interval was defined as a period of time (in months) between two consecutive live births [20]. This suggested that a woman could have several inter-birth intervals depending on her parity. Thus, the inter-birth intervals were calculated as Where In = nth interval length between two consecutive births. k = highest parity a woman has had at a given point in her reproductive lifetime, Dn = date of birth of an nth pregnancy, Dn-1 = date of birth of the preceding ((n-1)th) pregnancy and 30.4 = average number of days in a month During data analysis, the inter-birth intervals were first analyzed descriptively in order to assess their distributional features. Then a binary outcome variable was defined by assigning the inter-birth intervals into one of the two categories according to the WHO recommendation such that Proportions of the inter-birth intervals which were below the WHO recommendation by each of the independent variables were computed and presented, and the degree of association between them was tested using Chi-square (χ2). Factors associated with non-adherence were assessed using a multilevel logistic model in order to account for the fact that inter-birth intervals of the same woman are highly correlated. The intervals were considered to be nested or clustered among women. This procedure was conducted using the STATA command, ‘xtlogit’, to obtain random-effects logistic regression results. Odds ratios (OR), their corresponding 95% confidence intervals (CI) and P-values were calculated and presented as well. In interpreting effects such as OR, confidence intervals among other things play the role of P-values. Therefore, presenting OR and their corresponding confidence intervals without the P-values may suffice. However, we also presented the P-values because some readers prefer them for quick inferences about significance. The whole process of data analysis was conducted using STATA (version 11) statistical software (StataCorp, Texas, USA). A cut-off point (significance level) at which a factor was identified as a predictor of the outcome, non-adherence, was 5%.
N/A
