Segmentation. For a program to focus on audiences and potential clients with unmet need, rather than the population as a whole, it must find ways to distinguish the unmet need group from the rest of the population. As Jose Rimon and Mark Lediard observed in regard to the "Red Line" communication strategy for reducing unmet need in Nepal, "to achieve maximum impact and cost-effectiveness, the practice of developing general messages to a mass audience must give way to the discipline of audience segmentation" (170).

Effective segmentation helps programs chose the right responses for the needs of the high-priority subgroups. This task may be difficult because many women in the unmet need group probably face a combination of obstacles, not just one, that keep them from using contraception, and these obstacles may change. Nevertheless, secondary analysis of survey data can point to high-priority groups.

As Oleh Wolowyna and Ellen Starbird have pointed out, secondary analysis that divides the unmet need group into various segments can help reveal reasons for unmet need. For example, analysis of the 1989 Bolivia DHS found that about half of all women with unmet need did not intend to use contraception. At the same time, about 4 women in every 10 with unmet need were not aware of any contraceptive method, suggesting that many women cannot intend to use contraception because they do not know about it. These findings suggest that more communication is needed to improve awareness of contraception among women who say that they do not intend to use it (244).

In Bangladesh a segmentation study using national survey data showed that at least two-thirds of women who say that they intend to use a method within the next 12 months have a child below age three, and most have small families (13). Such data can help guide field workers to couples likely to have unmet need, as well as help determine high-priority audiences for communication.

Also, in analyzing DHS data from Egypt (see Intention to use contraception in Chapter 3.2), Stover and Heaton found that the subgroup of women with unmet need who had once used contraceptives but had discontinued use constituted a substantial share of all married women of reproductive age in the country, at 7.5%. Of this group, about one-third had discontinued because of side effects or health concerns, and one-quarter, because of method failure. "Given the size of this group," they concluded, "improvement in the family planning program to reduce side effects and enhance correct use could certainly contribute to increased prevalence and a reduction of unmet need" (208).

Software for analysis. Specialized software packages can facilitate analysis of survey data on unmet need. For example, the computer software CHAID—for Chi-squared Automatic Interaction Detector—can be used to segment unmet need subgroups. From survey data the CHAID user chooses a dependent variable (such as unmet need or intention to use contraception) and enters a variety of possible variables that characterize survey respondents. The CHAID program examines these variables to identify which characteristics of women best distinguish subgroups on the basis of the chosen dependent variable (117). It is important to select the variables carefully, based on previous studies and reasonable hypotheses about outcomes, in order to obtain useful data. Also, CHAID ranks the selected characteristics according to their predictive value, which lets the user know whether the best one exceeded the others by a clear margin or not. CHAID is available as part of the software package SPSS—Statistical Package for Social Sciences.

Certain other software programs devoted to strategic planning take unmet need into account. The Target-Cost program calculates the total fertility rate (TFR) that will result from an assumed contraceptive prevalence trend or, alternatively, the prevalence needed for a desired fertility change (209). The newest version of this program, being developed by The Futures Group International, will extend these calculations to include unmet need explicitly. The unmet need group is divided into potential spacers and limiters. The user can specify the percentage of unmet need that is to be satisfied and also different contraceptive method mixes. The program then calculates the implications for contraceptive prevalence, method mix, numbers of users, and program expenditures.

Absolute numbers versus percentages. To set priorities, programs need to distinguish between absolute numbers and percentages of women having unmet need. While most analyses report on the unmet need group as a percentage of all married women of reproductive age, or report subgroups as a percentage of the total unmet need group, estimates of absolute numbers of women with unmet need often give a different impression and may be more useful to programs (183). For example, while the percentage of women with an unmet need is much higher among women with many children, the majority of women with unmet need, in absolute terms, have small families. This is because in fast-growing developing-country populations younger women far outnumber older women, younger women have not yet had time to have many children, and many would prefer not to have as many children as older women have had.

Data from Vietnam demonstrate the importance of distinguishing absolute numbers from percentages (183) (see Figure 12). For example, about half of all women with seven or more children have an unmet need, but only about 11% of all women with an unmet need have seven or more children. In Vietnam, as elsewhere, most women with unmet need have one to three children.

Similarly, an analysis of unmet need in Sri Lanka based on DHS data found that, while the percentage with unmet need is below average among previous users of modern methods, the number in this subgroup is large. Therefore "further effort is needed to understand both why these women are no longer using a modern method and how to modify service provision to address their concerns or dissatisfaction" (50). The study also urged special efforts to reach young women with one or two children for two reasons: a substantial percentage of this group has unmet need for spacing, and such women comprise a large percentage of the population in Sri Lanka.

Reaching the high-priority groups. In designing a strategy, a basic question is how to reach population groups that contain large numbers of women with an unmet need (178). That is, after programs understand the reasons for unmet need and identify the characteristics associated with it, they must find women with such characteristics among the general population. Some may be isolated, living in remote areas far from the reach of information and services, while others may already be in contact with reproductive health services or relatively easy to reach with information through the mass media and social networks.

The fact that most women with an unmet need either are already pregnant or have an infant or young child (see Chapter 3.1 Unmet Need Levels by Women's Characteristics) helps programs both identify a large, high-priority group of women and suggests the way to reach them, since many women with unmet need may already be served by prenatal, postpartum, or other MCH services such as immunization. Knowing this, family planning programs may choose to address messages specifically to new mothers and to establish links with MCH services.(see Chapter 4.4 Linking Family Planning and Other Services)

Guidelines for field workers. Strategic program design can help to guide the activities of field staff. When field workers make contact with couples who are likely to have an unmet need, they can ask about interest in limiting or spacing births and why the couples are not currently using family planning or intending to do so. Most programs probably already have records covering some of this information. For example, in Bangladesh field workers already record, for each woman in their catchment areas, her age, number of children, and time since last birth (open birth interval). It appears feasible to add two items that would identify whether she has an unmet need for family planning—whether the woman desires another child and, if so, when (14, 179).

In addition, field workers could ask couples about their experience with or concerns about side effects and discontinuation. "If the more qualitative dimensions are then added so that field workers define unmet need not solely in terms of use of any kind of contraceptive but use of a method that responds to the clients' individual desires and needs, one has the basis for expanding services from simple contraceptive coverage to a quality of care dimension," Sinding and Fathalla have observed (200).

Anticipating the need. Programs must not concentrate exclusively on current unmet need. Ideally, to avoid unmet need, people should have access to good-quality family planning information and services beforehand, so that they will not stay long, if at all, in the unmet need group. The unmet need group is continually changing, both because individual reproductive attitudes change and because new people are continually entering the childbearing years. Thus a comprehensive strategy to meet unmet need also reaches many people who do not currently have an unmet need but soon might, without effective program action. Another benefit of anticipating people's needs is that it probably will result in program changes that help meet the needs of current family planning clients as well.