Strategies for communicating contraceptive effectiveness

  • Review
  • Intervention

Authors


Abstract

Background

Knowledge of contraceptive effectiveness is crucial to making an informed choice. The consumer has to comprehend the pros and cons of the contraceptive methods being considered. Choice may be influenced by understanding the likelihood of pregnancy with each method and factors that influence effectiveness.

Objectives

To review all randomized controlled trials comparing strategies for communicating to consumers the effectiveness of contraceptives in preventing pregnancy.

Search methods

Through February 2013, we searched the computerized databases of MEDLINE, POPLINE, CENTRAL, PsycINFO and CINAHL, ClinicalTrials.gov, and ICTRP. Previous searches also included EMBASE. We also examined references lists of relevant articles. For the initial review, we wrote to known investigators for information about other published or unpublished trials.

Selection criteria

We included randomized controlled trials that compared methods for communicating contraceptive effectiveness to consumers. The comparison could be usual practice or an alternative to the experimental intervention.

Outcome measures were knowledge of contraceptive effectiveness, attitude about contraception or toward any particular contraceptive, and choice or use of contraceptive method.

Data collection and analysis

For the initial review, two authors independently extracted the data. One author entered the data into RevMan, and a second author verified accuracy. For the update, an author and a research associate extracted, entered, and checked the data.

For dichotomous variables, we calculated the Mantel-Haenszel odds ratio with 95% confidence intervals (CI). For continuous variables, we computed the mean difference (MD) with 95% CI.

Main results

Seven trials met the inclusion criteria and had a total of 4526 women. Five were multi-site studies. Four trials were conducted in the USA, while Nigeria and Zambia were represented by one study each, and one trial was done in both Jamaica and India.

Two trials provided multiple sessions for participants. In one study that examined contraceptive choice, women in the expanded program were more likely to choose sterilization (OR 4.26; 95% CI 2.46 to 7.37) or use a modern contraceptive method (OR 2.35; 95% CI 1.82 to 3.03), i.e., sterilization, pills, injectable, intrauterine device or barrier method. For the other study, the groups received educational interventions with differing format and intensity. Both groups reportedly had increases in contraceptive use, but they did not differ significantly by six months in consistent use of an effective contraceptive, i.e., sterilization, IUD, injectable, implant, and consistent use of oral contraceptives, diaphragm, or male condoms.

Five trials provided one session and focused on testing educational material or media. In one study, knowledge gain favored a slide-and-sound presentation versus a physician's oral presentation (MD -19.00; 95% CI -27.52 to -10.48). In another trial, a table with contraceptive effectiveness categories led to more correct answers than a table based on pregnancy numbers [ORs were 2.42 (95% CI 1.43 to 4.12) and 2.19 (95% CI 1.21 to 3.97)] or a table with effectiveness categories and pregnancy numbers [ORs were 2.58 (95% CI 1.5 to 4.42) and 2.03 (95% CI 1.13 to 3.64)]. Still another trial provided structured counseling with a flipchart on contraceptive methods. The intervention and usual-care groups did not differ significantly in choice of contraceptive method (by effectiveness category) or in continuation of the chosen method at three months. Lastly, a study with couples used videos to communicate contraceptive information (control, motivational, contraceptive methods, and both motivational and methods videos). The analyses showed no significant difference between the groups in the types of contraceptives chosen.

Authors' conclusions

These trials varied greatly in the types of participants and interventions to communicate contraceptive effectiveness. Therefore, we cannot say overall what would help consumers choose an appropriate contraceptive method. For presenting pregnancy risk data, one trial showed that effectiveness categories were better than pregnancy numbers. In another trial, audiovisual aids worked better than the usual oral presentation. Strategies should be tested in clinical settings and measured for their effect on contraceptive choice. More detailed reporting of intervention content would help in interpreting results. Reports could also include whether the instruments used to assess knowledge or attitudes were tested for validity or reliability. Follow-up should be incorporated to assess retention of knowledge over time. The overall quality of evidence was considered to be low for this review, given that five of the seven studies provided low or very low quality evidence.

Plain language summary

Informing consumers about how well birth control works

To make a good choice for family planning, people have to know how well different methods work. The pros and cons of the methods are important. People may choose birth control based on how well the method prevents pregnancy. Consumers also need to know what affects the usefulness of the birth control method.

Through February 2013, we did computer searches for randomized trials of ways to inform people about how well family planning methods prevent pregnancy. We wrote to researchers to find other trials. The new program could be compared to the usual practice or to another program or means of informing people.

We found seven trials with a total of 4526 women. Two had several sessions for participants. One of those looked at the choice of birth control method. Women in the test program more often chose to be sterilized or to use modern birth control than women with the usual counseling. In the other study, the groups had different sessions on family planning. Both groups increased their birth control use. However, the groups were similar at six months in using methods that work well to prevent pregnancy. Five trials had a single session for each group. In one, women learned more from a slide-and-sound format than from having a doctor talk to them. Another trial found that effectiveness categories were better than pregnancy numbers for comparing the methods. Still another study provided structured counseling using a flipchart on family planning methods. The groups were similar in choice of birth control and in numbers who still used their chosen method at three months. The last study used videos to inform couples about family planning. The groups were mostly similar in birth control use after the videos. But those who watched videos on motivation and on family planning did not choose pills or an injectable method as often as those who watched only the family planning video.

The studies had different types of participants and programs. We cannot say overall what would help consumers choose their method of birth control. Ways to inform women about family planning options should be tested in clinics. Trials should look at the choice of birth control method, along with how much consumers remember later.

Background

Knowledge of contraceptive effectiveness is crucial to making an informed choice. The consumer has to comprehend the pros and cons of the contraceptive methods being considered. Choice may be influenced by understanding the likelihood of pregnancy with each method and factors that influence effectiveness, as well as the program's emphasis. A study promoting long-acting methods showed that 56% of the women chose intrauterine contraception (Secura 2010). The next most frequent choices were oral contraceptives (12%), the vaginal ring (12%) and a subdermal implant (11%). Intervention studies have found that most women cited method effectiveness as the most important reason for choosing a contraceptive (Steiner 2003; Steiner 2006). A study of hormonal method preferences found that women chose the vaginal ring most often, and the main reasons were convenience and the low likelihood of error (Lete 2007). The pill was second in frequency and was generally chosen for its effectiveness. In earlier studies of preferences for contraceptive characteristics, women identified effectiveness as the most important factor in choosing a contraceptive method (Garcia 1997; Grady 1999). Researchers often take the concept of risk for granted, while health care consumers and providers may not understand enough about probability theory to interpret the available information, as it is often presented numerically (ARHP 2006; Mahajan 2007). Strategies for communicating information about contraceptive effectiveness and risks should be developed and tested formally. For example, the assumption that people understand fractions better than rates impedes communication about risk of chromosomal abnormalities (Van Vliet 2001).

Theories and models of education, communication, or behavioral change can help explain how an intervention works. Many health education interventions have no identified theoretical premise (Bellg 2004; Borrelli 2005). To many professionals, the lack of any theory or principles guiding an educational or psychological intervention is similar to not having a physiologic basis for a medical intervention. O'Connor 2003 reviewed decision aids for making health care choices. The authors discussed the need for learning what types of aids work better with certain groups of people (O'Connor 2003), but they did not address any theoretical basis. DiCenso 2002 abstracted the theoretical basis for interventions to reduce unintended pregnancies among adolescents. Not all the strategies identified included contraceptive effectiveness since they were focused on adolescents.

Since published reports often have insufficient information on the intervention to assess its relevance, Mayo-Wilson 2007 suggested that CONSORT guidelines be expanded to include more intervention information. Even if an intervention has a reasonable premise, it must be implemented with fidelity to the design to be useful. Borrelli 2011 developed a tool to assess the fidelity of health behavior interventions in clinical trials. The framework was intended for assessing current trials and can be useful in reviewing educational interventions. Domains of treatment fidelity include having a curriculum or treatment manual, specifying training of providers, assessing delivery of intervention, and assessing participants' receipt of treatment and ability to use the treatment skills.

This review focuses on randomized controlled trials of ways to communicate contraceptive effectiveness or pregnancy risk to consumers or potential users of the contraceptive methods. Lopez 2011 examined interventions with an explicit theoretical basis for improving contraceptive use. Outcomes for the current review include attitudes toward a particular contraceptive method and knowledge about its effectiveness as well as contraceptive use. Others reviews focused more on contraceptive use than choice. Halpern 2011 studied interventions to improve adherence to hormonal contraceptive regimens. Lopez 2012 reviewed randomized controlled trials (RCTs) of educational interventions to increase contraceptive use among postpartum women.

Objectives

To review all randomized controlled trials comparing strategies for communicating to consumers the effectiveness of contraceptives in preventing pregnancy.

Methods

Criteria for considering studies for this review

Types of studies

We included randomized controlled trials in any language that compared methods for communicating contraceptive effectiveness for contraception to consumers. The comparison could be usual practice or an alternative to the experimental intervention. Studies that focused on sexually transmitted diseases were included if they also addressed contraceptive effectiveness.

Types of participants

The participants in the trials were consumers or potential users of the contraceptive methods. They could be females or males.

Types of interventions

Interventions could include educational programs or materials and counseling sessions focused on individuals or groups. The content of the program, sessions, or material had to include the effectiveness of more than one type of contraception. Contraceptive methods could be hormonal or non-hormonal.

Trials were excluded if the intervention did not address the relative effectiveness of methods or the report did not provide sufficient information to determine whether effectiveness or pregnancy risk was addressed.

Types of outcome measures

  • Knowledge of contraceptive effectiveness

  • Attitude about contraception or toward any particular contraceptive

  • Choice or use of contraceptive method.

Search methods for identification of studies

Electronic searches

Through February 2013, we searched the computerized databases MEDLINE, POPLINE, Cochrane Central Register of Controlled Trials (CENTRAL), PsycINFO and CINAHL for studies of communicating contraceptive effectiveness. We also searched for current trials via ClinicalTrials.gov and ICTRP. The search strategy is given in Appendix 1. The previous search strategy also included EMBASE, and is shown in Appendix 2.

Searching other resources

We also examined references lists of relevant articles and wrote to known investigators for information about other published or unpublished trials not discovered in our search.

Data collection and analysis

Selection of studies

We assessed for inclusion all titles and abstracts identified during the literature searches with no language limitations. One author reviewed the search results and identified reports for inclusion or exclusion. A second author also examined the reports identified for appropriate categorization. Specifics for the original review and the updates can be found in Acknowledgements and in Contributions of authors.

Data extraction and management

One author abstracted the data and entered the information into RevMan. This included the Characteristics of included studies and Data and analyses. Another author or research associate conducted a second data abstraction and verified correct data entry. No discrepancies occurred. Specifics for the original review and the updates can be found in Acknowledgements and in Contributions of authors.

In addition, we summarized the characteristics of these educational interventions. The framework in Borrelli 2011 helped guide this synthesis. We included basics about the participants and interventions, the outcome measures (whether they assessed participants' understanding and skills regarding the intervention), and the timeframe for outcome assessment.

For the initial review, we looked for a theoretical base for the interventions. We had extracted data on constructs and identified implied theory or models. However, none of the original five trials had an explicit theoretical base. The studies employed a variety of guiding principles or concepts, which may have their basis in behavioral theories or models. Since they were not explicit, we removed this table for the 2012 update. Theory-based interventions for improving contraceptive use were examined in Lopez 2011.

Assessment of risk of bias in included studies

Studies were examined for methodological quality, according to recommended principles (Higgins 2011). Factors considered were study design, randomization method, allocation concealment, blinding, and losses to follow-up and early discontinuation. Adequate methods for allocation concealment include a centralized telephone system and the use of sequentially-numbered, opaque, sealed envelopes (Schulz 1995; Schulz 2002). High losses to follow-up threaten validity (Strauss 2011). Limitations in design were presented in Risk of bias in included studies, and were considered in interpreting the results.

Measures of treatment effect

For continuous variables, the mean difference (MD) was computed with 95% confidence interval (CI) using a fixed-effect model. RevMan uses the inverse variance approach (Higgins 2011). The measures could include mean scores for the assessment of knowledge or change in the mean scores. For dichotomous outcomes, the Mantel-Haenszel odds ratio (OR) with 95% CI was calculated using a fixed-effect model. An example would be the proportion of women who chose a specific contraceptive method. Fixed effect and random effects give the same result if no heterogeneity exists, as when a comparison includes only one study.

Assessment of heterogeneity

Given the diversity of interventions, we did not conduct meta-analysis for pooled estimates. We assessed sources of heterogeneity without pooling the data and noted heterogeneity due to differences in interventions, study design, and populations.

Data synthesis

Following GRADE principles, we assessed the quality of evidence (Balshem 2011). When a meta-analysis is not viable due to varied interventions, a summary of findings table is not feasible. Therefore, we did not conduct a formal GRADE assessment, i.e., an evidence profile and summary of findings table (Guyatt 2011).

Given the differences across studies, we first summarized characteristics of the interventions and the outcome assessments. We assessed the quality of evidence for this review based on the Risk of bias in included studies and the characteristics summary; we then summarized the evidence quality results. Quality could be high, moderate, low, or very low. RCTs were considered to be high quality then downgraded for each of the following:

  • allocation: no information on randomization sequence generation or allocation concealment, or no concealment;

  • outcome measure: no assessment of contraceptive method choice or use, or did not have independent assessment or record linkage;

  • follow-up length: outcome measured during or immediately after intervention

  • loss to follow-up greater than 20%.

We did not downgrade on the basis of blinding. Due to the differences between interventions, blinding of investigators and participants to assignment was difficult, although blinding of outcome assessors was more feasible.

Results

Description of studies

See: Characteristics of included studies; Characteristics of excluded studies; Characteristics of ongoing studies.

Results of the search

The 2012 to 2013 search produced 238 references. This includes 175 unduplicated citations from the electronic databases, 1 from a reference list, and 63 found in ClinicalTrials.gov and ICTRP. Two new trials were included (Langston 2010; Stephenson 2011). Four trials (with five reports) were excluded. One ongoing trial was added (Madden 2011); details can be found in Characteristics of ongoing studies.

Included studies

Seven trials met the inclusion criteria and had a total of 4526 women. Five trials were multi-site studies. Four trials were conducted in the USA, Nigeria and Zambia were represented by one study each, and one trial was done in both Jamaica and India. Four studies took place in clinical settings (Marshall 1984; Omu 1989; Langston 2010; Stephenson 2011), while three provided the intervention in the community (Kraft 2007; Steiner 2003; Steiner 2006).

Most of the studies provided one-session interventions (Marshall 1984; Steiner 2003; Steiner 2006; Langston 2010; Stephenson 2011). Marshall 1984 compared five different media for providing the same contraception information. Steiner 2003 and Steiner 2006 each tested three educational tools with different formats for similar contraception information. Langston 2010 provided structured counseling using a flipchart developed by the World Health Organization (WHO) or usual care. Omu 1989 was focused on sterilization acceptance but also provided education regarding "all methods of contraception." The experimental group had four sessions while the control group received standard counseling in one session. Two trials had couples-based interventions (Kraft 2007; Stephenson 2011). Kraft 2007 reported on the contraception outcomes from an HIV and STD risk reduction trial. Three intervention sessions were provided for the experimental group while the comparison group had one information session. Stephenson 2011 compared different video presentations provided to HIV serodiscordant and concordant couples. More detail on the interventions can be found in Characteristics of included studies.

Outcomes of interest from the included trials were as follows:

  • Marshall 1984 - contraceptive knowledge gained from pretest to post-test; satisfaction with the educational medium.

  • Omu 1989 - choice of contraceptive method; attitude toward sterilization.

  • Steiner 2003 and Steiner 2006 - responses to two questions about contraceptive effectiveness asked while viewing the educational material; women's perceptions of the information (amount and whether easy to understand).

  • Kraft 2007 - use of effective contraceptives; psychosocial factors affecting contraceptive use. Assessment was six months after the intervention.

  • Langston 2010 - choice of very effective contraceptive method and method initiation at enrollment visit and after the intervention; continuation at three months. Very effective methods included those with one-year pregnancy rates less than 1% for typical use.

  • Stephenson 2011 - uptake of modern contraception at randomization visit and after the intervention.

Excluded studies

For the 2012 update, we excluded four additional trials (Akman 2010; Carneiro 2011; Garbers 2012; Vogt 2012). See Characteristics of excluded studies for reasons.

Risk of bias in included studies

Two trials were published in the 1980s and five from 2003 to 2011.

Allocation

Three trials had adequate concealment of allocation (Steiner 2003; Steiner 2006; Langston 2010). Three trials did not provide information on how the randomization sequence was generated or whether allocation was concealed before assignment (Marshall 1984; Omu 1989; Stephenson 2011). In Kraft 2007, participants drew slips of paper with their assignments on them.

Blinding

In two trials, the physician-providers were reportedly blinded to the study arms (Marshall 1984; Langston 2010). However, one group received the additional intervention prior to seeing the physician, so the blinding may have been incomplete. Kraft 2007 was open. Stephenson 2011 was single blind, but no details were provided. One trial did not mention any blinding of participants or assessors (Omu 1989). The randomization sequence was reportedly concealed from staff managing the study in Steiner 2003 and Steiner 2006.

Incomplete outcome data

Losses to follow-up were only relevant for the three trials with multiple sessions or follow-up data (Omu 1989; Kraft 2007; Langston 2010). Omu 1989 lost 8% by the six-week postpartum visit. By the last session, Kraft 2007 reportedly lost less than 10% of the intervention group and none from the control group that had only one session. However, Kraft 2007 lost 22% overall by the six-month follow-up. By three months, Langston 2010 lost 17% of the intervention group and 16% of the usual care group. In addition, Langston 2010 excluded 28 women after randomization because of second trimester pregnancy, completed spontaneous abortion or ectopic pregnancy.

Three trials provided one session with assessment at that visit (Marshall 1984; Steiner 2003; Steiner 2006). Stephenson 2011 also had one session for each group, but has data on follow-up to be reported later.

Effects of interventions

No trials could be combined in a meta-analysis due to differences in the interventions. Furthermore, measurements were taken in differing time frames so the results are not comparable across studies. Although the study designs were similar in Steiner 2003 and Steiner 2006, different educational tools were tested.

Multiple educational sessions

Two trials provided several sessions for the intervention group (Omu 1989; Kraft 2007).

Contraceptive method choice, use, and continuation

Omu 1989 measured contraceptive 'choice' as an outcome. Several differences were found between women in the expanded program, which emphasized voluntary sterilization, and those in the standard information group. By the end of the intervention, women in the expanded program were more likely to be sterilized (OR 4.26; 95% CI 2.46 to 7.37) (Analysis 1.1). By six weeks postpartum, those in the expanded program were more likely to use a 'clinical' or 'modern' contraceptive (OR 2.35; 95% CI 1.82 to 3.03) (Analysis 1.2), i.e., sterilization, pills, injectable, intrauterine device or barrier method. They were less likely to use no contraceptive method (OR 0.44; 95% CI 0.32 to 0.61) (Analysis 1.3).

Kraft 2007 was a couples-based intervention for preventing HIV and STDs with a three-session intervention group and a one-session information group. Reportedly, by the six-month follow-up, the two groups did not differ significantly in consistent use of an effective contraceptive, defined as sterilization, IUD, injectable (Depo Provera), implant (Norplant), as well as consistent use of OCs, diaphragms, or male condoms. Use among both groups reportedly increased over time.

Attitudes or perceptions about contraception or particular method

Psychosocial issues were assessed in two trials that included contraceptive information (Omu 1989; Kraft 2007). Omu 1989 had several questions on attitudes or beliefs about sterilization. The women in the expanded program were more likely to agree that sterilization was safe (OR 9.15; 95% CI 6.77 to 12.36), that a woman would still be strong after sterilization (OR 9.67; 95% CI 7.14 to 13.10), that sex drive would not change (OR 11.02; 95% CI 8.08 to 15.03), and that a woman's status would not change (OR 9.19; 95% CI 6.86 to 12.31) (Analysis 1.4). In Kraft 2007, the study groups were not significantly different for the psychosocial variables and relationship factors at six months. These included perceived pregnancy risk (Analysis 2.1), importance of not becoming pregnant (Analysis 2.2), expectations for partner's support for contraception (Analysis 2.3), and participation in contraceptive decision making (Analysis 2.4).

Single session testing media

Five trials provided one session for each study group and were mainly focused on testing educational material or media (Marshall 1984; Steiner 2003; Steiner 2006; Langston 2010; Stephenson 2011).

Contraceptive method choice, use, and continuation

Stephenson 2011 focused on HIV serodiscordant or concordant couples. Videos were shown to the four study groups (contraceptive methods, motivational, and control videos). The study groups were not significantly differences in the types of contraceptives chosen after the video and offer of free contraceptives (Analysis 3.1 to Analysis 3.3). However, within groups, choices shifted from nearly 80% using condoms to more use of oral contraceptives (35% to 40%) and injectables (37% to 41%), as well as implants (10% to 13%).

In Langston 2010, structured counseling was provided using a contraceptive flipchart. The intervention and comparison groups did not differ significantly in choice of contraceptive method (very effective, effective, or less effective method) (Analysis 4.1). The investigators used WHO definitions of method effectiveness: very effective had one-year typical use pregnancy rates less than 1% (sterilization, IUDs, implants); effective had typical pregnancy rates of 1% to 9% (pills, rings, patches, injections). They defined less effective methods as having a pregnancy rate of 10% or higher (condoms, withdrawal, periodic abstinence, no method). The study groups did not differ significantly in continuation of the chosen method at three months (Analysis 4.2).

Understanding of effectiveness or preference for educational method

Comprehension of contraceptive effectiveness was addressed in three trials that focused on providing contraceptive information (Marshall 1984; Steiner 2003; Steiner 2006). These studies tested educational media or tools and also examined preferences for each educational method. Marshall 1984 compared different media for communicating contraception information. Knowledge gain was less for an oral presentation by the physician-provider compared to a 'slide-and-sound presentation with unfamiliar voice' (MD -19.00; 95% CI -27.52 to -10.48) (Analysis 5.1). The other groups did not differ significantly in knowledge gain. For satisfaction with the educational method, the study groups in Marshall 1984 did not differ significantly (Analysis 5.2).

Steiner 2003 compared three tables of contraceptive effectiveness information: 1) numeric risk of pregnancy; 2) categories of method effectiveness (more effective, effective for most users, and less effective for most users); and 3) both categories and numbers (see Characteristics of included studies). Use of the categories table showed a greater change in women who understood that hormone injections were more effective than pills compared to the numeric table (OR 2.42; 95% CI 1.43 to 4.12) (Analysis 6.1) and compared to the table with categories and numbers (OR 2.58; 95% CI 1.50 to 4.42) (Analysis 6.2). The categories table also showed a greater increase in those who understood that combined pills were more effective than condoms compared to the other tables: OR 2.19 (95% CI 1.21 to 3.97) (Analysis 6.1) and OR 2.03 (95% CI 1.13 to 3.64) (Analysis 6.2), respectively. Fewer women with the categories table thought the tool was hard to understand compared to women with the numbers table (OR 0.29; 95% CI 0.13 to 0.63) (Analysis 6.4) or to women with the categories and numbers table (OR 0.38; 95% CI 0.17 to 0.85) (Analysis 6.6).

In Steiner 2006, three charts with contraceptive effectiveness information were tested: 1) three categories of effectiveness stratified by typical or consistent users; 2) four categories of effectiveness with numbers of pregnancies; and 3) a continuum of effectiveness by numbers of pregnancies (see Characteristics of included studies). The groups with each chart did not differ significantly in change in comprehension of relative contraceptive effectiveness (Analysis 7.1; Analysis 7.2). Steiner 2006 also examined preferences for each educational chart. Women with the categories chart were more likely to feel the chart provided enough information to choose a method compared to those with the chart of stratified categories (OR 1.97; 95% CI 1.13 to 3.44) (Analysis 7.3) and that the chart was easy to understand (OR 1.47; 95% CI 1.03 to 2.10) (Analysis 7.3). The categories-chart group did not differ significantly from the continuum-chart group for these perception items (Analysis 7.4).

Discussion

Summary of main results

Within studies, some methods of communicating contraceptive effectiveness appeared to work better than others. However, since the interventions differed across trials, we cannot determine the most useful approach. Two trials had multiple sessions for participants. In Omu 1989, women in the enhanced intervention group were more likely to choose a more effective contraceptive method than women in the standard counseling group (Omu 1989). For Kraft 2007, the groups received educational interventions of differing intensity. Both reportedly had increased contraceptive use, but use did not differ significantly between the groups.

The dominant delivery mode was a single information session. Five studies tested approaches to communicating contraceptive information in one session. Three trials had no follow-up, so we do not know if the knowledge was retained or whether the intervention affected contraceptive choice. In Marshall 1984, a slide-and-sound presentation resulted in more knowledge gain than the traditional oral presentation by a physician. Satisfaction was similar regardless of the medium used for communicating. In Steiner 2003, understanding of contraceptive effectiveness was better with a table focused on effectiveness categories than one with pregnancy numbers, and more women preferred the categories table. However, in Steiner 2006, comprehension was similar regardless of the type of chart viewed. More women in the group with the effectiveness categories chart liked their chart compared to women with an effectiveness chart stratified by typical or correct use. The other two trials had some follow-up. Langston 2010 provided structured counseling with a flipchart on contraceptive methods. The intervention and usual-care groups did not differ significantly in choice of contraceptive method (by effectiveness category) or in continuation of the chosen method at three months. Stephenson 2011 used videos to communicate contraceptive information to couples. The analyses showed no significant differences in the types of contraceptives chosen. The report only had baseline data; manuscripts are reportedly in progress with pregnancy and contraceptive use data.

Many of the trials focused on testing educational tools, and none had an explicit theoretical base. However, Omu 1989 mentioned perceived benefits from the Health Belief Model (Janz 2002). The intervention in Kraft 2007 specifically addressed the constructs of positive expectations, norms, and self-efficacy (i.e., confidence in one's ability to perform a certain behavior). These concepts are elements of Bandura's Social Cognitive Theory (Baranowski 2002). In Stephenson 2011, one study arm received a motivational video. The report did not mention what motivational principles were used, e.g., the Information-Motivation-Behavior Skills Model (Fisher 1992). The limited use of theory or models in these trials contrasts with the extensive use in HIV research (Albarracín 2005; Fishbein 2000). Behavioral science has also been incorporated more formally into interventions for preventing sexually transmitted diseases (Aral 2007). More intentional use of theory may help expand the knowledge base of what works in contraceptive counseling.

Overall completeness and applicability of evidence

Given the limited number of studies, not all regions and settings were represented. Four trials were conducted in the USA, while Nigeria and Zambia were represented by one study each, and one trial was done in both Jamaica and India. Four studies took place in clinical settings, while three provided the intervention in the community.

We cannot meaningfully compare the varied types of interventions, outcome measures, and assessment timeframes (Table 1). Interventions included providing structured counseling in a post-abortion setting, showing charts of contraceptive methods to women in a shopping center, and conducting multiple risk-reduction sessions with HIV serodiscordant or concordant couples. Most of the reports did not provide specifics about how contraceptive effectiveness was presented, e.g., categories of relative effectiveness or tables of specific numbers. Outcomes ranged from understanding a chart the woman was viewing to consistent use of an effective contraceptive method. Further, some studies had follow-up of three to six months while others assessed outcomes immediately after the intervention.

Quality of the evidence

The quality of evidence is based on the evidence from the included studies. The assessment addresses the utility of the evidence for this review as well as design features. The overall quality is considered to be low (Table 2), given that five of the seven studies provided low or very low quality evidence for this review. Of the remaining two trials, the evidence was considered high quality from one and moderate quality from the other. The primary differences for the higher-quality studies included assessing contraceptive uptake or use as well as measuring the outcome during follow-up.

None of the studies mentioned whether the instruments used to assess knowledge or attitudes had been tested for validity or reliability. The interventions in four trials appeared to be primarily method-driven, as the studies tested tools or media to communicate contraceptive information (Marshall 1984; Steiner 2003; Steiner 2006; Langston 2010).

Potential biases in the review process

The evidence quality summary was developed post hoc. While it was based on information already extracted, we may have been biased in selecting the criteria for downgrading trials.

Authors' conclusions

Implications for practice

Health care providers routinely communicate contraceptive information to their patients. Due to the many types of interventions here, we cannot say overall what would best help consumers choose an appropriate contraceptive method. For presenting pregnancy risk data, one trial found that categories of method effectiveness were better, or more preferred, than numbers of pregnancies. Another trial indicated that audiovisual aids worked better than the usual oral presentation by a physician. One trial showed that an enhanced counseling program led to more women choosing sterilization, which was emphasized, or a modern method of contraception.

Implications for research

These trials varied greatly in the types of participants and interventions to communicate contraceptive effectiveness. More detailed reporting of intervention content would help in interpreting results. Reporting should include whether the instruments used to assess knowledge or attitudes were tested for validity or reliability. Strategies for communicating information should be tested in clinical settings and measured for their effect on contraceptive choice. Follow-up should be incorporated to assess retention of knowledge over time.

Acknowledgements

Carol Manion of FHI 360 helped with the literature searches in 2008 and 2010.

Data and analyses

Download statistical data

Comparison 1. Expanded counseling and education program versus standard family planning information
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Woman's contraceptive use: sterilized by 6 weeks postpartum11012Odds Ratio (M-H, Fixed, 95% CI)4.26 [2.46, 7.37]
2 Woman's contraceptive use: modern contraceptive method at 6 weeks postpartum11012Odds Ratio (M-H, Fixed, 95% CI)2.35 [1.82, 3.03]
3 Woman's contraceptive use: no contraceptive method at 6 weeks postpartum11012Odds Ratio (M-H, Fixed, 95% CI)0.44 [0.32, 0.61]
4 Woman's attitudes at 6 weeks postpartum1 Odds Ratio (M-H, Fixed, 95% CI)Subtotals only
4.1 Agreed sterilization was safe11012Odds Ratio (M-H, Fixed, 95% CI)9.15 [6.77, 12.36]
4.2 Believed woman would still be strong after sterilization11012Odds Ratio (M-H, Fixed, 95% CI)9.67 [7.14, 13.10]
4.3 Agreed sex drive would not change after sterilization11012Odds Ratio (M-H, Fixed, 95% CI)11.02 [8.08, 15.03]
4.4 Agreed women's status would not change11012Odds Ratio (M-H, Fixed, 95% CI)9.19 [6.86, 12.31]
Analysis 1.1.

Comparison 1 Expanded counseling and education program versus standard family planning information, Outcome 1 Woman's contraceptive use: sterilized by 6 weeks postpartum.

Analysis 1.2.

Comparison 1 Expanded counseling and education program versus standard family planning information, Outcome 2 Woman's contraceptive use: modern contraceptive method at 6 weeks postpartum.

Analysis 1.3.

Comparison 1 Expanded counseling and education program versus standard family planning information, Outcome 3 Woman's contraceptive use: no contraceptive method at 6 weeks postpartum.

Analysis 1.4.

Comparison 1 Expanded counseling and education program versus standard family planning information, Outcome 4 Woman's attitudes at 6 weeks postpartum.

Comparison 2. Couples intervention to reduce risk for unintended pregnancy versus standard care
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Perceived pregnancy risk (at 6 months)1221Mean Difference (IV, Fixed, 95% CI)0.0 [-0.23, 0.23]
2 Perceived importance of not becoming pregnant (at 6 months)1220Mean Difference (IV, Fixed, 95% CI)0.01 [-0.32, 0.34]
3 Positive expectations for partner's support for contraception (at 6 months)1220Mean Difference (IV, Fixed, 95% CI)0.07 [-0.25, 0.39]
4 Participation in contraceptive decision making (at 6 months)1190Mean Difference (IV, Fixed, 95% CI)0.09 [-0.17, 0.35]
Analysis 2.1.

Comparison 2 Couples intervention to reduce risk for unintended pregnancy versus standard care, Outcome 1 Perceived pregnancy risk (at 6 months).

Analysis 2.2.

Comparison 2 Couples intervention to reduce risk for unintended pregnancy versus standard care, Outcome 2 Perceived importance of not becoming pregnant (at 6 months).

Analysis 2.3.

Comparison 2 Couples intervention to reduce risk for unintended pregnancy versus standard care, Outcome 3 Positive expectations for partner's support for contraception (at 6 months).

Analysis 2.4.

Comparison 2 Couples intervention to reduce risk for unintended pregnancy versus standard care, Outcome 4 Participation in contraceptive decision making (at 6 months).

Comparison 3. Couples intervention for long-term contraceptive use (videos)
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Contraceptive chosen: methods video versus control video1 Odds Ratio (M-H, Fixed, 95% CI)Subtotals only
1.1 Condoms or none1767Odds Ratio (M-H, Fixed, 95% CI)0.99 [0.54, 1.82]
1.2 Oral contraceptives or injectables1767Odds Ratio (M-H, Fixed, 95% CI)0.79 [0.57, 1.09]
1.3 Implant, IUD, or tubal sterilization1767Odds Ratio (M-H, Fixed, 95% CI)0.91 [0.62, 1.33]
2 Contraceptive chosen: motivational video versus methods video1 Odds Ratio (M-H, Fixed, 95% CI)Subtotals only
2.1 Condoms or none1763Odds Ratio (M-H, Fixed, 95% CI)0.94 [0.52, 1.70]
2.2 Oral contraceptives or injectables1763Odds Ratio (M-H, Fixed, 95% CI)1.17 [0.83, 1.64]
2.3 Implant, IUD, or tubal sterilization1763Odds Ratio (M-H, Fixed, 95% CI)0.84 [0.57, 1.24]
3 Contraceptive chosen: motivational+methods videos versus methods video1 Odds Ratio (M-H, Fixed, 95% CI)Subtotals only
3.1 Condoms or none1735Odds Ratio (M-H, Fixed, 95% CI)1.21 [0.68, 2.14]
3.2 Oral contraceptives or injectables1735Odds Ratio (M-H, Fixed, 95% CI)0.83 [0.59, 1.15]
3.3 Implant, IUD, or tubal sterilization1735Odds Ratio (M-H, Fixed, 95% CI)1.18 [0.81, 1.71]
Analysis 3.1.

Comparison 3 Couples intervention for long-term contraceptive use (videos), Outcome 1 Contraceptive chosen: methods video versus control video.

Analysis 3.2.

Comparison 3 Couples intervention for long-term contraceptive use (videos), Outcome 2 Contraceptive chosen: motivational video versus methods video.

Analysis 3.3.

Comparison 3 Couples intervention for long-term contraceptive use (videos), Outcome 3 Contraceptive chosen: motivational+methods videos versus methods video.

Comparison 4. Structured counseling versus usual care
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Contraceptive chosen1 Odds Ratio (M-H, Fixed, 95% CI)Subtotals only
1.1 Very effective method1222Odds Ratio (M-H, Fixed, 95% CI)0.74 [0.44, 1.26]
1.2 Effective method1222Odds Ratio (M-H, Fixed, 95% CI)1.40 [0.81, 2.40]
1.3 Less effective method1222Odds Ratio (M-H, Fixed, 95% CI)0.94 [0.36, 2.47]
2 Continuation at 3 months1172Odds Ratio (M-H, Fixed, 95% CI)1.34 [0.66, 2.72]
2.1 Very effective method1100Odds Ratio (M-H, Fixed, 95% CI)1.76 [0.63, 4.92]
2.2 Effective method172Odds Ratio (M-H, Fixed, 95% CI)1.03 [0.38, 2.79]
Analysis 4.1.

Comparison 4 Structured counseling versus usual care, Outcome 1 Contraceptive chosen.

Analysis 4.2.

Comparison 4 Structured counseling versus usual care, Outcome 2 Continuation at 3 months.

Comparison 5. Personalized audiovisual education materials versus nonpersonalized materials
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Knowledge gain1 Mean Difference (IV, Fixed, 95% CI)Subtotals only
1.1 Slide and sound with familiar versus unfamiliar voice140Mean Difference (IV, Fixed, 95% CI)-2.75 [-10.82, 5.32]
1.2 Oral presentation versus slide and sound with unfamiliar voice140Mean Difference (IV, Fixed, 95% CI)-19.0 [-27.52, -10.48]
1.3 Oral presentation versus pamphlet140Mean Difference (IV, Fixed, 95% CI)-5.0 [-13.10, 3.10]
1.4 Combined methods versus slide and sound with familiar voice140Mean Difference (IV, Fixed, 95% CI)3.0 [-4.82, 10.82]
2 Satisfaction excellent1 Odds Ratio (M-H, Fixed, 95% CI)Subtotals only
2.1 Slide and sound with familiar versus unfamiliar voice140Odds Ratio (M-H, Fixed, 95% CI)2.85 [0.78, 10.47]
2.2 Oral presentation versus slide and sound with unfamiliar voice140Odds Ratio (M-H, Fixed, 95% CI)1.0 [0.29, 3.48]
2.3 Oral presentation versus pamphlet140Odds Ratio (M-H, Fixed, 95% CI)0.82 [0.24, 2.84]
2.4 Combined methods versus slide and sound with familiar voice140Odds Ratio (M-H, Fixed, 95% CI)1.29 [0.32, 5.17]
Analysis 5.1.

Comparison 5 Personalized audiovisual education materials versus nonpersonalized materials, Outcome 1 Knowledge gain.

Analysis 5.2.

Comparison 5 Personalized audiovisual education materials versus nonpersonalized materials, Outcome 2 Satisfaction excellent.

Comparison 6. Communicating pregnancy risk (three tables)
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Change in comprehension: categories versus numbers1 Odds Ratio (M-H, Fixed, 95% CI)Subtotals only
1.1 Knew hormone injections were more effective than pills1289Odds Ratio (M-H, Fixed, 95% CI)2.42 [1.43, 4.12]
1.2 Knew combined pills were more effective than condoms1289Odds Ratio (M-H, Fixed, 95% CI)2.19 [1.21, 3.97]
2 Change in comprehension: categories versus categories+numbers1 Odds Ratio (M-H, Fixed, 95% CI)Subtotals only
2.1 Knew hormone injections were more effective than pills1286Odds Ratio (M-H, Fixed, 95% CI)2.58 [1.50, 4.42]
2.2 Knew combined pills were more effective than condoms1286Odds Ratio (M-H, Fixed, 95% CI)2.03 [1.13, 3.64]
3 Perception that table provided enough information to choose: categories versus numbers1289Odds Ratio (M-H, Fixed, 95% CI)0.58 [0.32, 1.06]
4 Perception that table was difficult to understand: categories versus numbers1289Odds Ratio (M-H, Fixed, 95% CI)0.29 [0.13, 0.63]
5 Perception that table provided enough information to choose: categories versus categories+numbers1286Odds Ratio (M-H, Fixed, 95% CI)0.60 [0.33, 1.08]
6 Perception that table was difficult to understand: categories versus categories+numbers1286Odds Ratio (M-H, Fixed, 95% CI)0.38 [0.17, 0.85]
Analysis 6.1.

Comparison 6 Communicating pregnancy risk (three tables), Outcome 1 Change in comprehension: categories versus numbers.

Analysis 6.2.

Comparison 6 Communicating pregnancy risk (three tables), Outcome 2 Change in comprehension: categories versus categories+numbers.

Analysis 6.3.

Comparison 6 Communicating pregnancy risk (three tables), Outcome 3 Perception that table provided enough information to choose: categories versus numbers.

Analysis 6.4.

Comparison 6 Communicating pregnancy risk (three tables), Outcome 4 Perception that table was difficult to understand: categories versus numbers.

Analysis 6.5.

Comparison 6 Communicating pregnancy risk (three tables), Outcome 5 Perception that table provided enough information to choose: categories versus categories+numbers.

Analysis 6.6.

Comparison 6 Communicating pregnancy risk (three tables), Outcome 6 Perception that table was difficult to understand: categories versus categories+numbers.

Comparison 7. Communicating pregnancy risk (three charts)
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Change in comprehension: categories versus stratified1 Odds Ratio (M-H, Fixed, 95% CI)Subtotals only
1.1 Knew pills were more effective than condoms1598Odds Ratio (M-H, Fixed, 95% CI)1.36 [0.95, 1.93]
1.2 Knew IUDs were more effective than hormone injections1598Odds Ratio (M-H, Fixed, 95% CI)1.22 [0.85, 1.77]
2 Change in comprehension: categories versus continuum1 Odds Ratio (M-H, Fixed, 95% CI)Subtotals only
2.1 Knew pills were more effective than condoms1600Odds Ratio (M-H, Fixed, 95% CI)1.07 [0.76, 1.51]
2.2 Knew IUDs were more effective than hormone injections1600Odds Ratio (M-H, Fixed, 95% CI)1.11 [0.77, 1.59]
3 Perceptions of chart: categories versus stratified1 Odds Ratio (M-H, Fixed, 95% CI)Subtotals only
3.1 Provided enough information to choose1598Odds Ratio (M-H, Fixed, 95% CI)1.97 [1.13, 3.44]
3.2 Was easy to understand1598Odds Ratio (M-H, Fixed, 95% CI)1.47 [1.03, 2.10]
4 Perceptions of chart: categories versus continuum1 Odds Ratio (M-H, Fixed, 95% CI)Subtotals only
4.1 Provided enough information to choose1600Odds Ratio (M-H, Fixed, 95% CI)1.62 [0.91, 2.87]
4.2 Was easy to understand1600Odds Ratio (M-H, Fixed, 95% CI)0.91 [0.65, 1.27]
Analysis 7.1.

Comparison 7 Communicating pregnancy risk (three charts), Outcome 1 Change in comprehension: categories versus stratified.

Analysis 7.2.

Comparison 7 Communicating pregnancy risk (three charts), Outcome 2 Change in comprehension: categories versus continuum.

Analysis 7.3.

Comparison 7 Communicating pregnancy risk (three charts), Outcome 3 Perceptions of chart: categories versus stratified.

Analysis 7.4.

Comparison 7 Communicating pregnancy risk (three charts), Outcome 4 Perceptions of chart: categories versus continuum.

Appendices

Appendix 1. Search 2012 to 2013

MEDLINE via PubMed (01 Jan 2010 to Feb 12 Nov 2013)

("Contraception"[MeSH] OR contracept*) AND (decision OR choice OR choose OR counsel* OR educat* OR communicat* OR information disseminat*) NOT (cancer OR polycystic)
Field: Title/Abstract
Limits: Clinical Trial

CENTRAL (01 Jan 2010 to 20 Nov 2012)

contracept* in Title, Abstract or Keywords
AND decision OR choice OR choose OR counsel* OR communicat* OR educat* OR information disseminat* in Title, Abstract or Keywords
NOT cancer OR polycystic OR postmenopaus* in Title, Abstract or Keywords

POPLINE (2009 to 26 Nov 2012)

Global: (contraceptive effectiveness AND communication) NOT HIV
Filter by keywords: Interventions; Research report

PsycINFO and CINAHL (01 Jan 2010 to 22 Aug 2012)

(contraception or contraceptive or contraceptives or birth control) AND efficacy AND (communicat* or decision or choice or choose or counsel* or information disseminat*)

ClinicalTrials.gov (01 Jan 2010 to 22 Aug 2012)

Search terms: decision OR choice OR choose OR counsel OR counseling OR educate OR educating OR communicating OR communicate OR information dissemination
Study type: Interventional studies
Intervention: (contraception OR contraceptive) NOT abortion
Conditions: NOT (cancer OR polycystic OR HIV OR Turner OR postmenopausal)

ICTRP (01 Jan 2010 to 22 Aug 2012)

Title: decision OR choice OR choose OR counsel OR counseling OR educate OR educating OR communicating OR communicate OR information dissemination
Condition: contraception OR contraceptive

Appendix 2. Previous searches, 2007 and 2010

MEDLINE via PubMed (through 10 Mar 2010)

("Contraception"[MeSH] OR contracept*) AND (decision OR choice OR choose OR counsel* OR educat* OR communicat* OR information disseminat*) NOT (cancer OR polycystic)
Field: Title/Abstract
Limits: Clinical Trial, Randomized Controlled Trial

CENTRAL (through 17 Mar 2010)

contracept* in Title, Abstract or Keywords
AND decision OR choice OR choose OR counsel* OR communicat* OR educat* OR information disseminat* in Title, Abstract or Keywords
NOT cancer OR polycystic OR postmenopaus* in Title, Abstract or Keywords

POPLINE (through 06 Apr 2010)

contraceptive effectiveness & communication

EMBASE (through 05 Apr 2010)

contracept? AND efficacy AND (decision OR choice OR choos? OR counsel? OR communicat? OR information()disseminat?)

PsycINFO and CINAHL (through 05 Apr 2010)

(contraception or contraceptive or contraceptives or birth control) AND efficacy AND (communicat* or decision or choice or choose or counsel* or information disseminat*)

ClinicalTrials.gov (through 25 Mar 2010)

Search terms: decision OR choice OR choose OR counsel OR counseling OR educate OR educating OR communicating OR communicate OR information dissemination
Study type: Interventional studies
Intervention: (contraception OR contraceptive) NOT abortion
Conditions: NOT (cancer OR polycystic OR HIV OR Turner OR postmenopausal)

ICTRP (through 25 Mar 2010)

Title: decision OR choice OR choose OR counsel OR counseling OR educate OR educating OR communicating OR communicate OR information dissemination
Condition: contraception OR contraceptive

What's new

DateEventDescription
13 February 2013New citation required but conclusions have not changed

Added two new trials (Langston 2010; Stephenson 2011).

Added summaries of intervention characteristics (Table 1) and evidence quality (Table 2).

Removed information on implied theories or models. No trial had an explicit theory base.

12 February 2013New search has been performedSearches updated. One ongoing trial added (Madden 2011). Two trials excluded (Akman 2010; Vogt 2012).
Table 1. Intervention summary
StudyParticipantsInterventionsOutcome:
understanding,
beliefs, attitudes
Outcome:
contraceptive
choice or use
Follow-up length
Marshall 1984Seeking contraception at primary care sitePamphlet;
audiovisual;
oral communication

Score, 20-item knowledge test;

satisfaction with education (1 item)

---Immediately after
Omu 1989Pregnant; >= 4 prior deliveries4 sessions or
usual counseling
Attitudes and beliefs about sterilizationPercent sterilized; method used6 weeks postpartum
Steiner 2003At shopping malls (convenience)3 educational tables

2 items, relative effectiveness;

2 items, perception of tool information

---None (while viewing table)
Steiner 2006In urban areas (convenience)3 educational charts

2 items, relative effectiveness;

2 items, perception of tool information

---None (while viewing chart)
Kraft 2007Couples at risk for STD/HIV1 or 3 group sessionsIndividual items: motivation, relationshipConsistent use of effective method6 months
Langston 2010Seeking abortionStructured (flipchart) or usual counseling---

Choice and continuation (very effective method);

initiation

Immediately after and

3 months

Stephenson 2011Couples, HIV serodiscordant or concordant2 educational videos (+ control video)---Uptake of methodImmediately after
Table 2. Evidence quality summary
  1. 1Assessment based on information in Risk of bias in included studies and Table 1.

    2Quality could be high, moderate, low, or very low. We considered RCTs to be high quality then downgraded for the following: a) no information on randomization sequence generation or allocation concealment, or no concealment; b) outcome assessment (did not include contraceptive method choice or use, or did not have independent assessment or record linkage) c) follow-up length: outcome measured during or immediately after intervention; d) follow-up adequacy: loss > 20% (NA = not applicable due to no follow-up).

    3Not applicable due to no follow-up.

Study1AllocationOutcome
assessment
Follow-up
length
Follow-up
adequacy
Evidence quality2
Marshall 1984-1-1-1NA3Very low
Omu 1989-1Moderate
Steiner 2003-1-1NALow
Steiner 2006-1-1NALow
Kraft 2007-1-1Low
Langston 2010High
Stephenson 2011-1-1NALow

History

Protocol first published: Issue 1, 2008
Review first published: Issue 2, 2008

DateEventDescription
14 April 2010New search has been performedSearches were updated. No new trials were included; several new ones were excluded. A trial in progress was added (Westhoff 2010).
15 April 2008AmendedConverted to new review format.
1 December 2007New citation required and conclusions have changedSubstantive amendment

Contributions of authors

L Lopez conducted the primary data abstraction, drafted the review, updated the searches into 2013, and incorporated the new studies. D Grimes developed the idea, did the secondary data abstraction for the initial review, and edited the manuscript. M Steiner provided content expertise and reviewed the manuscript. In 2012, D Hilgenberg reviewed search results and did the secondary data abstraction. K Schulz provided statistical expertise and reviewed the manuscript.

Declarations of interest

M Steiner of FHI 360 is the lead author of two trials included in this review (Steiner 2003; Steiner 2006). At the time of the initial review, all authors were employed by FHI 360 where the Steiner trials were conducted.

Omu 1989 also involved personnel at FHI 360 but did not include any of the authors of this review.

Sources of support

Internal sources

  • No sources of support supplied

External sources

  • National Institute of Child Health and Human Development, USA.

    Funding to FHI 360 for conducting the review (LML, DAG, MS, KFS)

  • U.S. Agency for International Development, USA.

    Funding to FHI 360 for conducting the review (LML, DAG, MS, KFS)

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Kraft 2007

MethodsRandomized controlled trial conducted at 2 sites in the USA (Los Angeles and Oklahoma City). Sample size calculation was provided.
Participants301 heterosexual couples recruited through the women in community and clinic settings. Inclusion criteria: women were 18 to 25 years old, had male sex partner >= 18 years old, had sex without condom at least once in past 3 months, and had >= 1 of the following: 1) engaged in risky behavior (sex or drugs), 2) thought partner was at risk (e.g., had STD), 3) thought they or partner would have sex with someone else within year while they were still together. Women in Los Angeles had to self-identify as Latina or Hispanic. Women in Oklahoma City could be any ethnicity or race.
Exclusion criteria: pregnant, intended to become pregnant within a year, or reported being HIV positive.
Interventions

Primarily HIV/STD risk reduction trial that also addressed contraception.
1) One-session information condition: 1 group session of 1.5 to 2 hours (education about HIV, STDs, and contraception) included sample contraceptive methods, method use and effectiveness (no information on how effectiveness was presented) for preventing pregnancy and disease, and question and answer session.

2) Three-session intervention group: 3 group sessions of 2.5 hours each; information provided to 1-session group plus activities and discussion addressing perceived risk, expectations, norms, skills, and self-efficacy regarding prevention.

Outcomes

Consistent use of an effective contraceptive. ['Effective' based on pregnancy < 10% in 1 year of consistent use; this included sterilization, IUD, injectable (Depo Provera), implant (Norplant), as well as consistent use of OCs, diaphragms, or male condoms. 'Other' methods included withdrawal, rhythm, inconsistent use of abstinence, spermicides, or other methods or nonuse.]

Psychosocial factors (focused on motivation to use contraception) and relationship factors relevant to contraception; single questionnaire items.

Assessments were conducted at baseline and 6 months after the intervention.

Notes

Researcher provided standard errors for main outcomes and group sizes on request; used to calculate standard deviations.

Data not provided for contraceptive use but described in text.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)High riskRandomization involved slips of paper placed in a jar; participating couples chose a slip that indicated the group assignment.
Allocation concealment (selection bias)High risknone used
BlindingUnclear riskOpen
Incomplete outcome dataHigh riskLosses: for educational sessions, reportedly zero lost from comparison group and < 10% of intervention group.
Lost to follow-up by 6 months: 67/301 = 22% overall (no group data).
Researchers excluded 11 women from the analysis due to no sex for prior 3 months (N=8) and no data for primary outcome (N=3).

Langston 2010

Methods

RCT conducted at one site in New York City (USA) from Dec 2008 to Jul 2009.

A priori sample size calculation based on 80% power to detect increase in women requesting very effective method in intervention group from 30% to 50%; 125 needed in each arm.

Participants250 females recruited from a family planning referral clinic with predominately Hispanic population. Study population consisted of women seeking first trimester procedure for a spontaneous or induced abortion. Inclusion criteria: age 18 years or older, no desire to become pregnant right away, Spanish or English speaking, and access to a telephone.
Interventions1) Experimental: structured counseling using contraceptive flipchart, i.e., modified WHO Decision-Making Tool for Family Planning Clients and Providers (2005). (WHO tool presented method effectiveness by category and with pictures; reference under Langston 2010.) Included audio description of contraceptive methods delivered by trained research coordinator; standardized by presenting same information each time presented. Questions not answerable from flipchart were referred to physician. Flipchart modified for methods available in USA.
2) Usual care: one physician performed the clinical procedures. Contraceptive counseling routinely offered by the physician during the visit. Content and duration of contraceptive counseling performed by provider was left to discretion.
Outcomes

Primary: proportion of participants choosing very effective contraceptive method (on enrollment day; after counseling)

Secondary: method initiation on day of procedure and method continuation of very effective or effective method at 3 months.

Used WHO definitions of method effectiveness: very effective had 1-year typical use pregnancy rates < 1% (sterilization, IUDs, implants); effective had typical pregnancy rates of 1% to 9% (pills, rings, patches, injections).

Defined less effective as pregnancy rate >= 10% (condoms, withdrawal, periodic abstinence, no method).

Follow-up by phone at 3 months.

NotesFirst 101 participants were followed to 6 months; reportedly no significant difference noted in method use, so researchers focused on maximizing 3-month follow-ups.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskUsing random-number table, determined sequence for 1:1 allocation constrained by blocks of 10
Allocation concealment (selection bias)Low riskRandomization assignments were sealed inside numbered, opaque envelopes. Coordinator opened the next sequentially numbered envelope after completing informed consent
BlindingUnclear risk

Physician-providers did not know the participant’s allocation group, did not discuss the study with patients, and were asked not to change their counseling.

No blinding of participants or coordinators was feasible due to the nature of the intervention.

Incomplete outcome dataLow risk

28 women excluded after randomization because of second trimester pregnancy, completed spontaneous abortion or ectopic pregnancy.

Lost to follow-up: structured counseling 18/114 (16%); usual care 18/108 (17%).

Marshall 1984

MethodsRandomized controlled trial conducted at two sites in the USA.
Participants100 women requesting contraception at 2 primary care sites (clinic and private physician's office).
Interventions

Information on process of conception and 6 methods of reversible birth control (advantages and disadvantages; effectiveness rates) conveyed via 5 media:

1) Pamphlet to read before exam (effectiveness rates in table form);

2) Audiovisual presentation (slides and sound) with unfamiliar voice;

3) Audiovisual presentation with narration by own physician (and informed it was own doctor);

4) Personal communication by own physician of standard information contained in audiovisual presentation;

5) Combination: audiovisual as per group 3, pamphlet, and oral communication from own physician.

Outcomes

Contraceptive knowledge gain from pretest to post-test (20 items); satisfaction with information (including perceived learning) from 6 items; patient's assessment of knowledge gain; physician's assessment of time spent with patient and time spent discussing contraceptives.

Assessments were conducted prior to the intervention (pretest) and after the medical examination (post-test).

NotesUnable to reach authors for further information.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNo information
Allocation concealment (selection bias)Unclear riskNo information
BlindingUnclear riskPhysicians were reportedly blinded to the 2 study arms for which they directly communicated the same intervention information. One of those groups received additional communication prior to seeing the physician.
Incomplete outcome dataLow risk

Losses are not relevant since the intervention was one session.

For satisfaction, data presented for 1 of 6 items.

Omu 1989

MethodsRandomized controlled trial conducted at a teaching hospital in Benin, Nigeria from Sep 1984 to Mar 1986.
Participants1012 women with >= 4 previous deliveries who attended the prenatal clinic. Exclusion criteria: women for whom physician recommended sterilization due to medical reasons.
Interventions

1) Treatment group received information and education 4 times: admission to study (gestation week 29 to 35), prenatal visit after week 36 of pregnancy, hospitalization for delivery, and 6 weeks postpartum. Health effects of high parity and benefits of family planning were addressed in sessions 1, 3, and 4; all methods of contraception were covered in sessions 2 and 3 (voluntary surgical sterilization was covered in detail). Those interested in sterilization had more in-depth counseling during sessions 2, 3, and 4.
2) Control group received standard family planning counseling at the clinic. This included contraception methods available at the clinic but not risks of high parity.

No information on how method effectiveness was presented.

OutcomesPercent sterilized, attitudes toward sterilization, and use of specific contraceptive method at 6 weeks postpartum.
Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNo information
Allocation concealment (selection bias)Unclear riskNo information
BlindingUnclear riskNo information
Incomplete outcome dataLow riskLosses: 79 did not have data for 6-week follow-up; 79/1012 = 8%

Steiner 2003

Methods

Randomized controlled trial conducted in 5 locations in USA from Aug to Sep 2001.

Sample size calculation provided.

Participants461 women aged 18 to 44 years recruited at 5 shopping malls in Illinois, Florida, Arizona, Texas, and Virginia.
Interventions

Looking at 1 of 3 tables presenting contraceptive effectiveness information:

1) US FDA package labeling table: 2 columns of numbers (numeric annual risk of pregnancy for typical use and lowest expected rate);

2) WHO table: same columns as #1 but methods were grouped according to 3 categories of effectiveness (more effective, effective, less effective)a;

3) Table developed by researchers: 3 categories of effectiveness along with limited information about STD protection.

Outcomes

Responses to 2 questions about relative contraceptive effectiveness; also asked 2 questions on perceptions of table information.

Questions were part of a self-administered questionnaire. Asked before randomization and while looking at the assigned table.

Notes

Comprehension was assessed immediately; retention of knowledge and actual choice of contraceptive were not assessed.

aMore effective methods: sterilization, implants, injectable (Depo Provera), IUD.

Effective for most users: oral contraceptives.

Less effective for most users: condoms (male or female), diaphragm, cervical cap, spermicides, withdrawal, natural family planning.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskPermuted-block randomization used random block sizes of 6 and 9.
Allocation concealment (selection bias)Low riskGroup assignment was concealed in sequentially-numbered, sealed, opaque envelopes.
BlindingLow riskRandomization sequence was concealed from staff managing the study.
Incomplete outcome dataLow riskLosses are not relevant since the intervention was one session. Researchers excluded 28 questionnaires due to missing data for age or participant identifier.

Steiner 2006

Methods

Randomized controlled trial conducted in Kingston, Jamaica and in Bangalore, India.

Sample size calculation provided.

Participants900 women with basic English literacy were recruited.
Interventions

Looking at 1 of 3 charts representing contraceptive methods by effectiveness categories:

1) WHO: methods shown by 3 categories of effectiveness (more effective, effective, less effective) with numbers of pregnancies per 100 women in one year; stratified by average users versus correct and consistent users (from WHO);

2) WHO: methods shown by 4 categories of effectivenessa with numbers of pregnancies per 100 women in one year;

3) Methods shown on continuum from most effective (<= 1 pregnancy/100 women/one year) to least effective (about 30 pregnancies/women/one year).

Outcomes

Responses to 2 questions about relative contraceptive effectiveness; also asked 2 questions on perceptions of chart information.

Questions were part of an interviewer-administered questionnaire. Asked pre-intervention and while participants looked at their assigned chart.

Notes

Comprehension was assessed immediately; retention of knowledge and actual choice of contraceptive were not assessed.

aMost effective: implants, sterilization, IUD

Very effective: injectables, pills, lactational amenorrhea

Effective: condoms (male or female), diaphragms, fertility-awareness based

Least effective: spermicides

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskPermuted-block randomization method used random block sizes.
Allocation concealment (selection bias)Low riskGroup assignment was concealed in sequentially-numbered, sealed, opaque envelopes.
BlindingLow riskRandomization sequence was concealed from staff managing the study.
Incomplete outcome dataLow riskLosses are not relevant since the intervention was one session.

Stephenson 2011

  1. a

    FDA = Food and Drug Administration
    STD = sexually transmitted disease
    WHO = World Health Organization

MethodsRandomized controlled trial conducted at research site in Lusaka, Zambia
Participants

Cohort of 1502 HIV serodiscordant and concordant positive couples.

Inclusion criteria: cohabiting at least 12 months, planning to stay in Lusaka for at least 1 year from enrollment, age 18 to 45 years for women and 18 to 65 years for men, no evidence of infertility and no medical contraindication to contraception.

Interventions

1) Control group - video about nutrition, hand washing and bed net use; had access to usual family planning education.

2) Methods group - video on contraceptive methods emphasized IUDs and implants.

3) Motivational group - video modeled desirable future planning behaviors (will preparation, financial planning, and pregnancy prevention).

4) Methods and motivational group - videos #2 and #3.

Video contraceptive content was reportedly based on known efficacy of methods, starting with most effective long-acting methods (IUD, implant, tubal sterilization) followed by injectables, OCs, and then condoms. Aimed to promote long-acting methods. No information on how method efficacy was presented.

After video, counselor was available for questions; couples were offered choice of contraceptives free of charge.

Outcomes

Uptake of modern contraceptive at randomization visit (after intervention). Methods included condoms for contraception, pills, injectables, implant, IUD, and tubal sterilization.

Data presented for entire sample (N=1502) and for baseline contraceptive users (N=324).

NotesJul 2012: Researcher communicated they have manuscripts in progress on pregnancy incidence and on long-acting reversible methods.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNo information on how sequence was generated. Randomization occurred one to two weeks after enrollment.
Allocation concealment (selection bias)Unclear riskno information
BlindingUnclear riskSingle blind; no information on whom
Incomplete outcome dataUnclear riskNot applicable; report only included data on uptake at randomization (see Notes above).

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
  1. a

    RCT = randomized controlled trial
    IUD = intrauterine device

Akman 2010Not RCT; assigned according to order attended clinic. Intervention: no mention of relative contraceptive effectiveness.
Barnet 2009Intervention information did not mention relative effectiveness of contraceptive methods. Outcomes did not include those identified for this review.
Bashour 2008Intervention information did not mention relative effectiveness of contraceptive methods, nor was knowledge of effectiveness assessed.
Bolam 1998Intervention information did not mention relative effectiveness of contraceptive methods, nor was knowledge of effectiveness assessed.
Carneiro 2011Intervention information did not mention relative effectiveness of contraceptive methods, nor was knowledge of effectiveness assessed.
Chewning 1999Not an RCT; methods section indicates alternate assignment was used
Danielson 1990Not an RCT; methods section indicates alternate assignment was used
Floyd 2007Intervention focused on motivating to use contraception. Report did not mention discussing relative effectiveness of contraceptive methods.
Garbers 2012Intervention information did not mention communicating relative effectiveness to participants. The algorithm for tailoring method recommendations included effectiveness.
Harvey 2004Intervention information did not mention relative effectiveness of contraceptive methods. Emphasis on condom use.
Lazcano Ponce 2000Primary outcome was assessed differently between groups: intervention group - woman's choice of IUD; control group - physician's recommendation for IUD. Researchers claimed they did not want to interfere in routine care for control group.
Lee 2007Coin flip determined which rooms were assigned to program. Even-numbered rooms were the experimental group and odd-numbered were the control group. Cluster assignment (by room) not addressed in the analysis.
Lei 1996Not an RCT. Methods section indicates participants were enrolled sequentially.
Marcy 1983Intervention information did not mention relative effectiveness of contraceptive methods. Outcome was 'effective' contraceptive use.
McBride 2000Seven projects. Not RCTs, as assignment to treatment or control was generally by birth date.
Melnick 2008Intervention included providing contraceptive of choice. Report did not mention discussing relative effectiveness of methods.
Nobili 2007Not an RCT. Methods section states 'alternative order' although abstract mentions 'randomly divided.'
Peipert 2008Outcomes did not include those identified for this review.
Petersen 2007Outcomes were not pertinent to this review.
Saeed 2008Intervention information did not mention relative effectiveness of contraceptive methods.
Schunmann 2006Intervention focused on providing contraceptive of choice (and length of interview) rather than comparative effectiveness of methods.
Shlay 2003Outcomes were not pertinent to this review ('effective' contraceptive use)
Smith 2002Assignment to groups appears to be systematic not random - by clinic day or week. Intervention content was not specified.
Vogt 2012Intervention focused on oral contraceptive benefits and risks and 'alternative methods'; no mention of relative method effectiveness. Two brochures used for intervention.
Wight 2002Intervention information did not mention relative effectiveness of contraceptive methods
Zhu 2009Intervention information did not mention relative effectiveness of methods; recommended "most suitable" methods. Outcomes did not include specific contraceptive method.

Characteristics of ongoing studies [ordered by study ID]

Madden 2011

Trial name or titleComputerized Decision Aid (CDM RCT)
MethodsRCT, open label, conducted at Washington University (St Louis, MO (USA)). to one of two arms to receive computer survey concerning contraceptive choices
Participants

201 healthy women.

Inclusion criteria: 18 to 45 years old, at risk for unintended pregnancy and seeking reversible contraception at one of two clinics, English speaking.

Exclusion criteria: had a hysterectomy, bilateral oophorectomy, or have undergone female sterilization; unable to give informed consent due to language barrier or cognitive limitation.

Interventions

1) online CDM tool and tailored printout of contraceptive options based on survey answers;

2) similar computer tool without decisive factors and printout

Outcomes

Primary: percent of women using IUD, Implanon, or injectable contraception. Secondary: average decisional conflict score; self-reported satisfaction with contraceptive counseling; self-reported continuation and satisfaction with contraceptive method.

Measures at 6 months

Starting dateJan 2012; estimated completion Jul 2013
Contact information

Tessa Madden, MD, MPH; 314-747-6495; maddent@wudosis.wustl.edu

Ragini Maddipati, MSW; 314-747-6418; maddipatir@wudosis.wustl.edu

Notes 

Ancillary