MODELING THE BEHAVIOUR OF PAKISTANI MARRIED WOMEN TOWARDS CONTRACEPTIVE USE

Abstract

A number of logistic regression and log-linear models were fitted to determine the factors, affecting the contraceptive use and impact of women’s reproductive autonomy on contraceptive use in relation with husband-wife education attainments in Pakistan by using the data from 2000-01 Pakistan Reproductive Health and Family Planning Survey. In exploratory multivariable logistic regression models’ analysis of a large nationwide survey of women in Pakistan, there was found that most of socio-demographic, socio- economic and cultural variables were significantly associated with use of contraception. A number of logistic regression models were fitted by adjusting different, demographic, socioeconomic, cultural and women’s reproductive autonomy variables, to examine the effect of these variables on ever and currently use of contraceptive methods by the respondents. Predicted probabilities of ever and current contraceptive use through logistic regression models were also worked out by taking the most prominent independent variables from each model, separately. Number of living children, respondents’ education attainments, husbands’ desire for more children, Sex preference for the next child and spouses’ discussion about family planning were found more prominent predictor variables for both ever and currently use of contraception in this study. Women’s reproductive autonomy variables; decision to have no iiimore children, decision to delay the next child, and decision to use the family planning methods, were also found significantly associated with use of contraception. A number of log-linear models were fitted with the help of partial and marginal association tests for the variables, which emerged significantly associated in the logistic regression modeling process in case of ever and currently use of contraceptive, to analyze the factors affecting the contraceptive use in Pakistan. Standardized values of the models’ parameters estimates were computed to see the pair wise association within all levels of the variables. Women’s reproductive autonomy in relation with husband-wife education attainments and contraceptive use were examined through log-linear modeling process and predicted probabilities of contraceptive use. The results of this studies revealed that the prevalence of contraceptive use increases with the increase in the number of living children and education attainment levels of the respondent. From standardized values of models’ parameter estimates found a significant positive association between duration of education and prevalence of contraceptive use and number of living children only when that duration had exceeded six years. It was perceived from the results that Pakistani women who lived in urban area are more likely to have used contraception than those who lived in rural area, even though they were more educated. From the results of predicted probabilities of contraceptive use, it was found that spouses’ communication about family planning and women’s education emerged important predictors of contraceptive use. Contraceptive use was found consistently associated with husband-wife discussion about family planning and their education attainments through out logistic regression and log-linear modeling process. From the predicted probabilities of contraceptive use and standardized values of hierarchical log-linear models’ parameters estimates it is ivperceived that the husband’s desire for more children and his preference for children of a given sex, particularly sons, are obstacles to the application of contraceptive measures. Women’s reproductive autonomy appeared to be strongly associated with contraceptive use and spouses’ education attainments. Although lower fertility is commonly associated with women’s reproductive autonomy, the husband’s education influences his wife’s fertility intentions more than her own education does. Our studies were shown a strong relationship between women’s education and knowledge about family planning clinic and contraceptive use. The findings of this study reaffirm the overarching importance of women’s education, which influences both directly and indirectly the implementation of birth control measures. Improvement of women’s education can also help in reducing the cultural factor of sex preference for the next child, which is a strong barrier to contraceptive use. The data also suggest the need for giving great attention in future studies to the relationship between spouses, which can solve conflicting fertility desires. To further promote contraception, a greater awareness must be achieved among women with regard to birth control. Programmes should be set up to encourage women to practice family planning measures more widely in their early reproductive lives and not to seek so strongly a male child. Despite highlighting a number of determinants of contraceptive use in Pakistan, this study also demonstrates the importance of methodological issues in the contraceptive prevalence studies. It is suggested that the predicted probabilities must be computed in the prevalence of contraceptive studies, as these probabilities precisely provide the degree of assessment of the acceptability of contraception and log-linear model analysis technique also recommended for the future studies because it enhances confidence in the results and vprovides a specific measure of associations within all variables under study in multi-way contingency table for large scale data set.