Literature Review on the Relationship Between Discontinuation Rates and Fertility

September 2001

I. SUMMARY

• Discontinuation rates for the IUD are the lowest of modern methods.
• Hormonal methods (the pill, injectables) are most likely to be discontinued due to side effects or other health-related concerns.
• Women who have not yet reached or exceeded their ideal family size are most likely to discontinue contraception. However, discontinuation rates of hormonal methods are still much higher than IUDs – even if women have not yet reached their ideal family size.
• Discontinuation rates are high when contraception is used for spacing; therefore, contraception use and fertility may both be high when contraception is used for spacing of children and not for limiting family size.
• Service-related reasons like cost, access, and availability are rarely offered as reasons for discontinuation.
• Desire to have children young contributes to high discontinuation rates.
• Increased travel time to a clinic (31-60 minutes, greater than 5 kilometers) increases contraceptive discontinuation.
• Single-method discontinuation often occurs due to method switching; in studies should use all-method discontinuation as outcome measure of contraceptive discontinuation.
• Method failure was highest for periodic abstinence and withdrawal.

II. CONTRACEPTIVE DISCONTINUATION AN IMPORTANT INFLUENCE ON FERTILITY

• Across a study of 15 countries using DHS data, "More than half of recent unwanted fertility was due to either a contraceptive failure or a contraceptive discontinuation in all [fifteen] countries except Guatemala. The total unwanted fertility rate would be between about 0.2 and 1.1 births lower in the absence of failure and discontinuation" (Blanc, Curtis, Croft, 2001, p. 22).
• When women switch methods due to method preferences, then there is a reduced risk of pregnancy; when women switch methods because of access or health concerns, then there is an increased risk of unintended pregnancy (Kost, 1993).

III. DETERMINANTS OF CONTRACEPTIVE DISCONTINUATION

1. Time – About one-third of couples stop contraceptive method use within 12 months; about half stop within 24 months (regardless of contraceptive prevalence and female adult literacy rate; Ali & Cleland, 1995).
   
   
2. Method Type – IUD users least likely to stop method (82-89% still using after one year; 65-80% still using after two years; Ali & Cleland, 1995).¹ Users of hormonal methods like the pill or injectables had discontinuation rates similar to those who used withdrawal or periodic abstincence (30-40% discontinuance after one year; 50-60% discontinuance after two years; Ali & Cleland, 1995).¹ Across six countries, condom use continuation rates varied substantially from 18% to 47% continuation after two years (Ali & Cleland, 1995). Women who have not experienced method-related difficulties (such as IUD expulsion) or side effects during the first several months of use are least likely to switch methods (Steele, Diamon, & Wang, 1996).
   
   
3. Quality of Services – Cost, access, and availability of methods is rarely given as a spontaneous reason for contraceptive discontinuation (Ali & Cleland, 1995; Blanc, Curtis, Croft, 2001). Yet, anecdotal reports and cumulative quality of service measures (that group together all quality of service variables) suggest that quality of services may have a slight impact. Blanc, Curtis, Croft (2001) conclude that "the relationship between the two indicators [i.e., all-method rate and quality of services] is statistically significant, but not particularly strong" (Blanc, Curtis, Croft, 2001, p. 22).
   • More method choice is related to more method switching, but not discontinuation rates.
   • Method switching does not appear to be related to quality, but to (perceived) effectiveness and health-related concerns (Blanc, Curtis, Croft, 2001).
   • Adequate counseling, especially on side effects, is significantly related to decisions to continue contraception (Cotten et al., 1992).
   • Increased travel time (31-60 minutes; greater than 5 kilometers away) appears to contribute to contraceptive discontinuation (Steele, Curtis, & Choe, 1999; Zhang, Tsui, & Suchindran, 1999).
   • Discontinuation more likely with non-governmental sources of contraception as opposed to governmental sources (e.g., clinics, hospitals), perhaps because of better and/or more counseling (Steele, Curtis, & Choe, 1999; Zhang, Tsui, & Suchindran, 1999).
     
     
4. Reason Given – Across six countries and numerous studies (see references) the three main reasons given for discontinuation are (1) desire for a child, (2) method failure, and (3) health concerns (including side effects); availability, access, and cost were rarely given as reasons for discontinuation (Ali & Cleland, 1995). Also, Kost (1993) found that "most women who discontinue for nonpregnancy-related reasons probably are discontinuing because they are switching" (italics in original, p. 114).
   
   
1. Health Concerns: Hormonal method users reported discontinuation for health-related reasons substantially more than IUD users (e.g., 5% versus 19% in Egypt) (Ali & Cleland, 1995). Similarly, in a study of 15 countries, hormonal methods had the highest discontinuation rates due to side effects and health concerns (discontinuation for the pill within one year, 11-35%; discontinuation for injectables within one year, 15-37%; Blanc, Curtis, Croft, 2001). Typical side effects given were excessive bleeding, nausea, headache, and abdominal pain. Fear of side effects (as opposed to actual side effects) also can influence women to discontinue.
   
   
2. Accidental pregnancy: Method failure occurred most often for withdrawal, periodic abstinence, and other traditional methods (Ali & Cleland, 1995; Blanc, Curtis, Croft, 2001). Pill failure varied widely across countries but occurred much less frequently than withdrawal, periodic abstinence, and traditional methods. Failure with IUDs and injectables was low across all countries (Ali & Cleland, 1995). Method failure is most likely to happen when contraception is used for spacing children, as opposed to limiting family size (Moreno, 1993).
   
   
3. Desire for another child: Demand for another child and having one’s children early leads to discontinuation of contraception. However, across six countries, discontinuation after one year because of desire for another child was still relatively low (5-10%, except for Thailand at 14%), but was much higher after two years (11-24%). Women with IUD’s were especially unlikely to discontinue because they wanted another child regardless of length of time (Ali & Cleland, 1995).
   
   
   1. Motivation to Continue Contraception – Determined by examining if respondents’ desired family size was more, less, or equal to what they currently had. Discontinuation occurred more frequently among those who had not reached their desired family size, as compared to those who had already reached or exceeded their desired family size (Ali & Cleland, 1995).¹ IUD users much more likely to continue as compared to pill users (probably because "discontinuation requires a conscious decision and a clinical procedure," Ali & Cleland, 1995, p. 96).
   2. Miscellaneous Reasons: Other reasons occasionally given for discontinuation across the literature include: (a) to avoid partner disapproval, (b) to use another and/or (perceived) more effective method, (c) travel by either partner, (d) forgetfulness, and (e) lack of sexual activity.
      
      
5. Method Switching – "Between 29 and 59% of women who stopped using a modern, reversible method for a method-related reason switched to a different modern method within three months. Few women returned to the method they had discontinued. In contrast, women who had experienced a contraceptive failure and resumed using contraception after giving birth were most likely to return to the same method" (Blanc, Curtis, Croft, 2001, p. 21).
   
   
6. Demographic Variables – No fixed pattern emerged between discontinuation rates and education, residence, program effort, and adult female literacy in a study across six countries (Ali & Cleland, 1995). Age appeared to influence discontinuation, because "reported method failure was almost invariably higher among young women, who had not yet reached their desired family size, than it was among older women, who frequently had already reached or exceeded their preferred number of children" (Ali & Cleland, 1995, p. 96; Steele, Diamond, & Wang, 1996). Thus, motivation to adhere to methods properly appeared to be related to having reached one’s optimal number of children desired. Also, the longer women used a method successfully, the less likely the failure rate or discontinuation, probably because of greater experience with a specific method (Steele, Diamond, & Wang, 1996). No clear patterns emerge across the literature with regard to education/rural vs. urban persons and discontinuation rates/method failure, though rural women and lower educated women may be slightly more likely to discontinue contraception and/or experience method failure (results are often just trends, not statistically significant).

IV. REASONS FOR DISCONTINUATION BY METHOD

1. IUD – IUD users appear to report pelvic pain as the main side effect (about half of all users in a study of Niger; Cotten et al., 1992).
   
   
2. Hormonal Methods – Injectable users report headaches and dizziness as side effects; pill users report abdominal pain ( all at levels of less than 20%; Cotten et al., 1992). Pill users are likely to discontinue for access-related issues (Zhang, Tsui, & Suchindran, 1999).
   
   
3. Condoms – Condom users are likely to discontinue for access-related issues (Zhang, Tsui, & Suchindran, 1999). Condom usage is significantly associated with availability, free market distribution, compatibility with traditional methods, cultural beliefs (i.e., the condom is more natural and less likely to cause imbalance in body system than hormonal methods), more educated/wealthy, and higher levels of pre-marital or extra-marital sex (Goodkind & Anh, 1997).

V. CONTRACEPTION BEHAVIOR FOLLOWING DISCONTINUATION

Following discontinuation of one modern method, women most likely to switch to a different modern method (Blanc, Curtis, & Croft, 1999). Women are unlikely to ever return to a discontinued method. In contrast, women DO return to the same method following a contraception failure (Blanc, Curtis, & Croft, 1999).

VI. MEASUREMENT ISSUES (1^st two paragraphs from R. Bessinger)

When CPR is high, then DHS will include a calendar that collects retrospective data on the following: pregnancies, live births, termination of pregnancies, and contraceptive status. These events are recorded for each month over the five years proceeding the survey. In addition, reasons for discontinuation of contraception are also collected. Reasons can be method-related (e.g.side-effects, cost), fertility-related (e.g. became pregnant, wanted to become pregnant), and other (e.g. husband disapproves, divorced. Life tables techniques are used in the analyses of these types of data with segment of use as the unit of analysis. When CPR is low, then it really doesn’t make sense to include a calendar because too few people have had the opportunity to start and then discontinue use.

In the absence of a DHS with a calendar, there are no sources of data that would allow one to look at the effect of discontinuation on fertility in Uganda. A special study to address this question quantitatively could be done, but it would be very costly due to the very large sample size that would be required. We could address the question of whether or not discontinuation itself is an issue in Uganda via quantitative and/or qualitative means and might want to consider including it in one of our upcoming studies.

It is better to use "all-method discontinuation rates" as compared to a single method discontinuation rate because the latter does not take into account the possibility that women are using contraception but simply switched methods (Blanc, Curtis, Croft, 2001).

Reasons given for contraceptive discontinuation may be subject to bias due to politeness or acceptability about reasons for discontinuing. For example, Cotten et al. (1992) suspect that "desire for another child" was the reason most often given for discontinuation because it was more socially acceptable than admitting one discontinued due to side effects or forgetfulness. Care must be taken when devising questions and interviewing to avoid biased answers.

VII. RECOMMENDATIONS

1. Promote IUDs, as these have the lowest discontinuation rates even if desired family size has yet to be achieved.
2. Offer intensive support and counseling during first few months of method use; women who have successes with their method during the first 6 months of method use are likely to stick with the method.
3. Contraception use and fertility may be high when contraception is used for spacing of children, and not for limiting family size. Demand for having large families and/or having children young appears to be a major reason for high fertility rates, despite increases in contraceptive use. Thus, a communication campaign focusing on a norm of smaller family sizes started at a later age is needed.
4. Clinics should be within 5 kilometers and/or 30 minutes of clients.

VIII. SOURCES:

Ali, M. Cleland, J. (1995). Contraceptive discontinuation in six developing countries: A cause-specific analysis. International family planning perspectives. 21(3), 92.

Anh, P.T., Thang, N.T., & Cleland, J. (undated). Fertility regulation in Vietnam: an exploration of two puzzles. http://www.lshtm.ac.uk/eph/cps/dfid/118.htm.

Blanc, A.K., Curtis, S., & Croft, T. (1999). Does contraceptive discontinuation matter? Quality of care and fertility consequences. Chapel Hill, NC: MEASURE Evaluation.

Blanc, A.K., Curtis, S., & Croft, T. (2001). Does contraceptive discontinuation matter? MEASURE evaluation bulletin, 1, 21-23.

Cotten, N., Stanback, J., Maidouka, H., Taylor-Thomas, J.T., Turk, T. (1992). Early Discontinuation of Contraceptive Use in Niger and The Gambia. International family planning perspectives, 18(4), 145.

Goodkind, D., & Anh, P.T. (1997). Reasons for rising condom use in Vietnam. International family planning perspectives, 23, 173-178.

Kost, K. (1993). The dynamics of contraceptive use in Peru. Studies in Family Planning, 24, 109+.

Izmirlian, G., Adewuyi, A.A., Suchindran, C.M. (1997). Analysis of contraceptive discontinuation in six developing countries from durations of use at survey. Social Biology, 44, 124-135.

Magnani, R.J., Sosler, S.M., & McCann, H.G. (2000). Trends in reproductive behavior among adolescents and young adults in the Dominican Republic. Washington, DC: FOCUS on Young Adults/Pathfinder International.

Moreno, L. (1993). Differences by residence and education in contraceptive failure rates in developing countries. International family planning perspectives, 19, 54-60 & 71.

Moreno, L., & Goldman, N. (1991). Contraceptive failure rates in developing countries: Evidence from the Demographic and Health Surveys. International family planning perspectives, 17, 44-49.

Steele, F., Curtis, S.L., & Choe, M. (1999). The impact of family planning service provision on contraceptive-use dynamics in Morocco. Studies in Family Planning, 30, 28-42.

Steele, F., Diamond, I., Wang, D. (1996). The determinants of the duration of contraceptive use in China: a multilevel multinomial discrete-hazards modeling approach. Demography, 33, 12-23.

Zhang, F., Tsui, A.O., & Suchindran, C.M. (1999). The determinants of contraceptive discontinuation in Northern India: A multilevel analysis of calendar data. Chapel Hill, NC: MEASURE Evaluation.

Click here to view this document in Microsoft Word format