|
Closing the Effectiveness Gap |
|
| June 2007 Issue No. 13 |
The INFO Project • Johns Hopkins Bloomberg School of Public Health • Center for Communication Programs • 111 Market Place, Suite 310 • Baltimore, Maryland 21202, USA • 410-659-6300 • 410-659-6266 (fax) • www.infoforhealth.org • infoproject@jhuccp.org | |
HOW TO EXPLAIN CONTRACEPTIVE EFFECTIVENESS
A family planning provider’s approach to explaining contraceptive effectiveness to their clients should be flexible, because people differ in their preferences for the kind of information and level of detail they prefer to receive (29, 47, 75). Some prefer to receive risk information as numbers (20, 133) — for example, “2 women in every 100 become pregnant.” But others understand risk information better when it is expressed in descriptive, comparative terms, such as “more effective” and “less effective” (133).
|
Three communication approaches appear to be particularly valuable for family planning providers to adopt in explaining effectiveness to their clients:
- Rank methods in order of effectiveness. When clients are considering different contraceptive methods, rank them in order of effectiveness as typically used, to show how they compare in preventing pregnancy.
- Sometimes numbers can help. Provide statistics on typical-use pregnancy rates, if the client prefers to understand the numbers behind the rank-ordering.
- Explain long-term effectiveness. Explain long-term effectiveness, in order to help clients understand the risk of pregnancy over years of use of a given contraceptive method, as well as in the first year of use.
Rank Methods in Order of Effectiveness
Comparing family planning methods based on the risk of pregnancy during typical use helps give clients a basis for deciding on the general level of risk they are prepared to accept. Providers can compare the risk of pregnancy among different methods by grouping methods in order of their effectiveness, ranked from the more effective to the less effective.
|
Providers can also compare the risk of pregnancy when using contraception with the risk of pregnancy when not using any method. In focus-group discussions among women from the United Kingdom, participants said this type of comparison was useful to them (32). The risks of becoming pregnant in the absence of contraceptive use are substantially higher, of course, than when using contraception. On average, 3 women in every 100 will become pregnant from a single act of sexual intercourse (134). Over one year, an average of 85 in every 100 sexually active women not using contraception will become pregnant (120).
When comparing methods, focus on typical-use pregnancy rates. Between 2004 and 2006 a group of contraceptive experts convened on behalf of WHO and the U.S. Agency for International Development (USAID) to develop recommendations for communicating contraceptive effectiveness (58). The group recommended that, when clients are choosing from among different methods, providers should compare the contraceptive effectiveness of the methods by discussing only pregnancy risks during typical use, rather than comparing pregnancy risks during both typical use and perfect use (58, 112). Once a client has chosen a method, the provider can explain that the method has both a perfect-use and a typical-use pregnancy rate to give the client a goal for achieving the best contraceptive protection possible (see “Discuss the Effectiveness Gap to Encourage Correct and Consistent Use”).
The expert group based its recommendation on findings from a study of data from India and Jamaica. In the study, two groups of women from each country were shown contraceptive effectiveness charts that presented only typical-use pregnancy rates, while a third group was shown a chart that presented both typical-use and perfect-use rates. The women shown only typical-use rates were significantly more likely to describe their charts as easy to understand. They were also slightly more likely to improve their knowledge of comparative effectiveness (112).
When comparing methods, providers must take care to be consistent and avoid giving perfect-use pregnancy rates for some methods and typical-use pregnancy rates for others. Some providers give perfect-use pregnancy rates for the more effective methods but give typical-use rates for the less effective methods. For example, in the United States an early study showed that family planning providers tended to give clients perfect-use pregnancy rates for pills and IUDs but to give typical-use pregnancy rates for diaphragms and spermicides (123). Mixing rates in this way can make it appear that differences in effectiveness among the methods being compared are greater than they actually are.
|
Rank methods without using descriptive terms. Most clients are able to rank methods in order of effectiveness better when given descriptive categories of effectiveness rather than pregnancy rates. In a recent study in the United States, for example, women were shown one of three charts: (1) a chart with pregnancy risk statistics only (both perfect-use and typical-use pregnancy rates), (2) a chart with both statistics (perfect-use and typical-use rates) and descriptive effectiveness terms, or (3) a chart with descriptive effectiveness terms only (as in the second chart), based on typical-use rates. Among women who were shown the third chart—descriptive effectiveness terms only—knowledge that injectable contraceptives are more effective than oral contraceptives increased by 37%. By comparison, such knowledge among women shown either of the other two charts increased much less, by 20% (for the first chart) and 19% (for the second chart) (111).
People often interpret or understand descriptive terms differently, however. As a result, they may interpret the levels of risk differently. In the U.S. study, for example, women who were given only descriptive terms of effectiveness, without also being shown statistics, were likely to overestimate the risk of pregnancy during contraceptive use.
For instance, 41% of the study participants thought “effective contraceptive methods” were methods that would allow 30 women or more in every 100 to become pregnant over a year’s time (111). In contrast, an early (and now discontinued) WHO effectiveness table used the term “effective” to describe contraceptive methods that result in pregnancies to only 2 to 9 women in every 100 each year (137). Similar overestimating of health risks, such as risks with medications or breast cancer chemotherapy, has been reported in other studies of how people interpret descriptive categories (13, 18, 37, 51, 59, 78, 79, 94, 115, 131, 133).
Some studies have shown that people understand descriptive terms better when the terms are rank-ordered according to the numerical risks associated with the terms (19, 27, 54). For example, in a study that asked participants to assign numerical values to descriptive terms, listed either in rank order or randomly, the participants were more likely to give consistent numerical values when the descriptive terms were rank-ordered, rather than randomized (54).
Because differences in interpretation of descriptive categories affect people’s decisions (13, 20, 27), the expert group convened by WHO and USAID recommended against using descriptive terms, such as “very effective” or “effective,” to characterize effectiveness of contraceptive methods. Instead, the group developed a chart that ranks methods in order of effectiveness, without descriptive terms (see Figure 2). Providers can use this chart to help explain comparative typical effectiveness to clients.
Sometimes Numbers Can Help
Some clients may want to make their own objective comparisons among the methods they are considering and thus want to know the typical-use pregnancy rates for these methods. Some tested approaches to stating medical risks can help family planning providers explain pregnancy risk statistics to their clients:
-
Use frequencies, not percentages. Clients generally interpret statistics on risk more accurately when providers express them as frequencies, rather than as percentages (4, 29, 44). For example, people often find it easier to understand that 8 women in every 100 (a frequency) will become pregnant during the first year of use of combined oral contraceptives, rather than that 8% of women will become pregnant. To some people, percentages are abstract figures because the denominator, or the group to which the percentage applies, is not stated explicitly (47).Information about long-term effectiveness helps clients look ahead and recognize the continuing risk of pregnancy during contraceptive use.
- When comparing risks, use the same denominator. Most people find it easier to compare statistics on risk when providers express the pregnancy rates using the same denominator. When given comparisons using different denominators—for example, one rate with a denominator of 100 people and another with a denominator of 1,000 people—many people mistake which is the greater risk (8, 31).
- Explain risks using both positive and negative statements. Most experts think that providers should state risks in both positive and negative terms to avoid influencing clients’ perceptions of whether a particular level of risk is worth taking (30, 44, 83). To state pregnancy risks both ways, family planning providers could point out, for example, that in typical use of oral contraceptives during the first year, 92 women in every 100 do not become pregnant and, put the other way, 8 women in every 100 do become pregnant (47).
- Use visual aids to help clients understand pregnancy risk statistics. Visual aids help put the risk of pregnancy into perspective. The Paling Palette©, for example, shows risks in both positive and negative terms at the same time and thus presents an unbiased view (83) (see Figure 3). Some studies show that people find visual aids with human figures, such as the Paling Palette, more meaningful, easier to understand, and easier to identify with than bar charts (102). Other studies find that people prefer bar charts (39). Providers can use different visual formats to match the needs of the client. Providers also should keep in mind that, while visual aids can help with understanding, they do not replace face-to-face client-provider communication (135). Many clients say they want and need providers to explain illustrations depicting data (5).
Explain Long-Term Effectiveness
|
Providers should explain long-term contraceptive effectiveness, as well as discuss first-year pregnancy rates. Information about long-term effectiveness helps clients look ahead and recognize the continuing risk of pregnancy during contraceptive use. Long-term effectiveness is particularly important to convey because many contraceptive users believe the first-year pregnancy rate is their total risk of pregnancy, no matter how long they use their method, rather than their risk for the first year and with at least some further risk each and every year thereafter, as long as they continue use (88, 104).
In general, pregnancy rates are highest in the first year of use. The risk of pregnancy for a group of people during each year thereafter generally declines. The risk declines in part because those who are more likely to become pregnant—either because they are more fertile or they are more likely to make contraceptive use errors—do so early on. The group left consists of those people who are less fertile, have sex less frequently, or make fewer contraceptive use errors (118).
Also, the risk declines because some contraceptive methods are more likely to fail in the first few months of use. For example, most IUD expulsions occur in the first year, and especially during the first three months after insertion (6, 126, 142).
Even a low annual risk of pregnancy associated with a highly effective method turns into a higher risk over several years of use (88, 98, 118, 119). For example, in typical use of the copper-bearing TCu-380A IUD, a highly effective method, 0.8 women in every 100 (that is, 8 women in every 1,000) become pregnant during the first year (120). Over a 12-year period of typical use, 2.2 women in 100 (22 women in 1,000) relying on the TCu-380A IUD would become pregnant (127).
 |
 |
