Bangladeshi women's choice of a contraceptive is influenced not only by their personal characteristics but also by their assessments of available options. For example, pill users choose their method, among other reasons, because of its convenience, availability and cost, but these factors do not play a significant role in condom users' method choice. Users of oral contraceptives, long-term reversible methods and, especially, condoms are influenced by concerns about the side effects of other methods. These are among the findings of an analysis of 1996-1997 Demographic and Health Survey data, undertaken to examine a range of factors that were previously overlooked in studies of the determinants of method choice in Bangladesh.¹

Survey participants were a nationally representative sample of women between the ages of 10 and 49; the analyses of factors affecting contraceptive choice were based on the 3,510 respondents who were married, were not pregnant and were using a modern method of contraception at the time they were interviewed. The great majority lived in rural areas and were Muslim (86% and 88%, respectively); most of the women (79%) and their husbands (66%) had less than a complete primary education. On average, the women were nearly 30 years old and had 1.5 living sons; one-third had lost a child.

By far the largest proportion of women (50%) were using the pill. Substantial proportions relied on female sterilization (18%) and inject-ables (15%); the remainder were protected by condoms (9%), IUDs (4%), their husband's vasectomy (3%) or implants (fewer than 1%).

The main reasons pill users gave for choosing their method (cited by 35-41%) were that it is easy to use, a field-worker had delivered it to their home and they had concerns about other methods' side effects. Concern about the side effects of other contraceptives and ease of use were the most frequently mentioned reasons for using IUDs or implants (33-44%) and injectables (38-48%); a substantial proportion of IUD and implant users also cited the desire for a very effective method (27%). Condom users most often mentioned side effects of other methods and their husbands' preference (54% each). Women who had been sterilized were likely to say they had wanted a permanent method (77%); those whose husbands had had a vasectomy most often said that his preference drove the choice (66%). Some 10-16% of users of sterilization and long-term reversible methods (IUDs, implants and injectables), and smaller proportions of pill and condom users (8% and 3%, respectively), cited a family planning worker's recommendation as a reason for their method choice; cost was a factor in 2-8% of choices.

The analyst used logistic regression to examine the effects of six types of variables on modern contraceptive use. The variables reflected women's reproductive-related characteristics and goals, women's and their husbands' education, women's evaluation of available methods, accessibility of contraceptives, women's residence and their exposure to a family planning program.

According to the multivariate results, pill use is influenced by a wide variety of factors. Among variables related to reproductive goals, an increasing number of living children is associated with an increasing likelihood that a woman will choose the pill, whereas the longer a woman has been married, the more sons she has and the older she is, the less likely she will be to opt for the pill over other modern methods. A woman's education is not significantly associated with her choice of the pill, but if her husband has at least a complete primary education, she has an increased likelihood of choosing this method. Assessments of contraceptives and the accessibility of methods are important factors in a woman's method choice: Women who are concerned about side effects of other methods, find the pill convenient to use and consider it easy to obtain have an elevated likelihood of choosing the pill. The pill is distributed at no cost in Bangladesh; this factor also increases the likelihood of choosing the pill. Rural women and those who have had a recent visit from a family planning worker are more likely to choose the pill than other methods.

Results for long-term reversible methods are somewhat similar to those for the pill. An increasing number of living children, concerns about other methods' side effects and convenience of use are associated with an increased likelihood that women will choose these methods over other modern contraceptives, and women become less likely to adopt long-term reversible contraception as they get older. Having a husband with a primary school or higher education, however, reduces the likelihood of choosing these methods, as does being non-Muslim.

The determinants of choosing condoms vary considerably from those for the more effective methods. The choice of this method is predicted by a lower number of living children, shorter duration of marriage and older age. Both a woman's and her husband's level of schooling influence condom use; the likelihood of choosing this method is elevated if either spouse has a primary education or more. Concern about possible side effects of other methods is a significant factor in the choice to use condoms, and the effect is larger than that for the pill or other reversible methods. Non-Muslim and rural women have a reduced likelihood of using condoms.

Sterilization is the only method whose choice is affected by the experience of having lost a child; women who have suffered such a loss have a reduced likelihood of choosing permanent contraception. The choice of this method also becomes less likely as a woman's number of children declines, and it becomes more likely as her age and duration of marriage rise; if either partner has at least a primary education, sterilization becomes a less-likely choice. In contrast to the results for other methods, the findings for sterilization reveal a negative effect of concern about other methods' side effects and of the convenience, cost and availability of the procedure. Non-Muslim women are more likely than Muslims to opt for sterilization rather than another modern method.

While acknowledging that a lack of appropriate data prohibits exploration of several attitudinal and behavioral factors that may affect women's contraceptive method choice, the analyst contends that by including attributes of methods as possible factors, this study broadens the understanding of Bangladeshi women's contraceptive decision-making. He urges the inclusion in future surveys of a wider range of variables to help elucidate women's method choice, and stresses the importance of programs' offering safe, effective contraceptive methods that are suitable to clients' attitudes, behavior and reproductive goals. --D. Hollander

REFERENCE

1. Mannan HR, Factors in contraceptive method choice in Bangladesh: goals, competence, evaluation and access, Contraception, 2002, 65(5):357-364.