Measuring nuance in individual contraceptive need: A case study from a cohort in Malawi

Abstract

Contraceptive counseling protocols tend to focus narrowly on pregnancy intentions, which may overlook other factors that contribute to whether an individual wants or needs contraception. In this short report, we demonstrate the potential of two measures of individual contraceptive need that could be assessed as part of contraceptive counseling: (1) a composite score constructed from pregnancy intentions, sexual frequency, and perceived fecundity and (2) a direct measure of contraceptive need (“do you feel it is necessary for you to be using contraception right now?”) We compare the two measures using data from Umoyo wa Thanzi, a cohort study in Central Malawi (N=906; 2017–18). More frequent sex, perceptions of being more fecund, and stronger desire to avoid pregnancy were associated with directly reporting contraceptive need (p<0.001). Women who directly reported contraceptive need had a higher average composite score than women who directly reported they no need (mean=7.4 vs. 6.3; p<0.01), but nearly all participants had scores indicating some risk of unintended pregnancy. Contraceptive counseling protocols should consider assessing women’s direct report of contraceptive need, along with risk factors for unintended pregnancy, such as sexual frequency, perceived fecundity, and desire to avoid pregnancy, to better counsel clients.

Background

Quality contraceptive counseling may improve people’s reproductive health, wellbeing, and autonomy by helping people reach their reproductive goals in ways that are acceptable to them. Personalized counseling tailored to individual needs and preferences is one aspect of high-quality family planning services (Brandi and Fuentes 2020; Dehlendorf, Krajewski, and Borrero 2014; Dehlendorf et al. 2016).

Several family planning counseling tools are available to healthcare providers, including One Key Question (“Would you like to become pregnant in the next year?”) (Allen et al. 2017), PATH (Parenting/Pregnancy Attitude, Timing, and How) Efficient Questions for Client-Centered Contraceptive Counseling (“How to Use PATH” n.d.), and Contraceptive Action Plan’s 5-step client centered approach (“What Is CAP?” n.d.). In all of these tools, screening for pregnancy desires and intentions is the first step in contraceptive counseling. PATH Questions center on pregnancy desires and timing as a way to introduce the topic of contraception. Contraceptive Action Plan’s 5-step approach similarly begins with assessing pregnancy intentions before moving into a discussion about contraceptive methods. All three tools are driven by the goal to provide contraception to people who indicate that they do not want to become pregnant.

Of note, additional factors, beyond pregnancy desires and intentions, have a demonstrated influence on people’s contraceptive use and non-use among those who do not wish to become pregnant (Moreira et al. 2019). Such factors include exposure to the possibility of pregnancy (e.g., sexual activity or frequency) and one’s perception of their biological risk of pregnancy (e.g., perceived fecundity) (Bell and Gemmill 2021; Bornstein et al. 2021; Gemmill and Cowan 2021; Moreira et al. 2019; Polis et al. 2020). People may choose not to use a modern method of contraception, regardless of a counseling protocol that indicates their need, and this is a choice that must be respected (Senderowicz 2019). At the same time, it is important for healthcare workers to ensure that a person’s choice not to use a modern method of contraception, despite not wanting to become pregnant, is well-informed and proactive (Manzer and Bell 2021; Senderowicz 2019).

Assessing sexual frequency and perceived fecundity during contraceptive counseling, along with pregnancy desires and intentions, may provide a more comprehensive entry point into discussions about whether or not someone wants to use contraception, as compared to screening for pregnancy desires and intentions alone. To illustrate the potential utility of these three factors in contraceptive counseling as a way to assess contraceptive need, we present a case study from the Umoyo wa Thanzi (UTHA) research program, a cohort study of women^[1] in Central Malawi. We measured and analyzed components of a composite contraceptive need score and a direct measure of contraceptive need that could be utilized during counseling to assess contraceptive need. The composite contraceptive need score is computed from three self-reported items: sexual frequency, perceived fecundity, and desire to avoid pregnancy. The direct measure simply asked women if they have a need for contraception. We then compare the two proposed measures (the composite score and the direct measure) to understand discordance between them and to assess the potential utility of both a composite and direct measure of contraceptive need in contraceptive counseling.

Methods

Data

We used data from the fourth wave of the Umoyo wa Thanzi (UTHA) cohort in Central Malawi to construct and assess our proposed measures. The cohort, which drew from the 20,000-person catchment area of a private rural hospital, was established in 2013 following a census of the area. Eleven villages were randomly selected for inclusion in the study and all reproductive aged women who resided in the villages were eligible to participate. The initial response rate was 96% (Huber-Krum et al. 2017; Rao et al. 2017). This short report is a cross-sectional analysis of the fourth wave (2017–2018). In the fourth wave, 783 members of the original cohort participated in the survey (retention rate=76%). Women who were not retained in the cohort were younger and less likely to be married than women retained in the cohort (Bornstein 2021). An additional 379 women who met the inclusion criteria were recruited into the cohort at Wave 4 via a convenience sample, for a total of 1,162 participants.

Analytic sample

The analytic sample excludes women who had been sterilized (n=128) and women who were pregnant at the time of data collection (n=101) because they were not asked about their need for contraception. The analysis was further restricted to women who responded to the primary measures of interest, reducing the sample by an additional 27 women. Thus, the analytic sample includes 906 women. We included women in the analytic sample who had no lifetime sexual partners (n=39) because women who have not had sex may identify as needing contraception, particularly if they intend to have sex.

Key variables

Composite contraceptive need score:

We used three measures to compute a composite contraceptive need score: sexual frequency in the past month, perceived fecundity, and desire to avoid pregnancy. The three measures were not highly correlated (sexual frequency and perceived fecundity r²=−0.15; sexual frequency and desire to avoid pregnancy r²=−0.13; and perceived fecundity and desire to avoid r²=−0.02), indicating that they are likely measuring independent constructs (not shown).

Sexual frequency in the past month was collected as a continuous, right censored variable using the survey item, “in the past 30 days, how many times have you had sex?” Participants could respond with a whole number from 0–8 or an option of ‘9 times or more frequently.’ We recoded the variable with three categories: 0 times, 1–8 times, and 9 times or more. The cut-off points for sexual frequency are based on research linking the probability of pregnancy and the frequency of sexual intercourse. The Practice Committee of the American Society for Reproductive Medicine (ASRM) notes that daily sexual intercourse during a cycle confers the highest probability of conception, but that sexual intercourse on alternative days is comparable. The average menstrual cycle length is 28 days, but cycles of 21–35 days are also considered normal (American Society for Reproductive Medicine 2017). ASRM recommends that couples be informed that the probability of conception increases with the frequency of sexual intercourse and is highest when intercourse occurs every 1 to 2 days per cycle (American Society for Reproductive Medicine 2017). Studies concur that there is little difference in the probability of conception when the frequency of intercourse occurs more than every two days per cycle (Stanford and Dunson 2007). This would be roughly equivalent to 9 or more days per cycle, depending on cycle length. Thus, we created a category indicating 9 or more times. If no sex occurs in a cycle, a woman has zero probability of becoming pregnant, and thus we created a category indicating 0 times. There is variability in the probability of conception for women having sexual intercourse 1–8 times in a cycle, but because we did not inquire about menstrual cycle lengths or regularity, we are unable to use this measure to compute a precise probability of conception, and instead created a third category to capture those who have neither the lowest nor the highest probabilities of becoming pregnancy in a given cycle.

Perceived fecundity was collected using the survey item, “considering your life, do you find that becoming pregnant is easy, difficult, or average compared to other women in your community?” Response options were difficult, average/same as others, easy, and do not know. For the purpose of the composite score, we combined do not know with average/same as others. The majority of participants who responded ‘do not know’ (39/47) had never had sex. Perceived fecundity represents a person’s own assessment of their likelihood of becoming pregnant if they do not use a method of contraception. Perceived fecundity has been demonstrated to influence contraceptive use across settings, including in Malawi, such that lower perceived fecundity is associated with contraceptive non-use (Bell and Gemmill 2021; Bornstein et al. 2021; Polis et al. 2020).

Desire to avoid pregnancy was collected using a survey item that read, “how much do you want to avoid pregnancy in the next year?” There were six response options: strongly do not want to avoid, somewhat do not want to avoid, slightly do not want to avoid, slightly want to avoid, somewhat want to avoid, and strongly want to avoid. We recoded the variable into three categories: strongly do not want to avoid, somewhat/slightly want to avoid/not avoid, and strongly want to avoid. We combined somewhat do not want to avoid, slightly do not want to avoid, slightly want to avoid, and somewhat want to avoid for the purpose of this composite score, because these responses likely indicate ambivalent or indifferent fertility desires (Huber-Krum et al. 2017). In a previous analysis with this sample of women in our cohort and using the same question, women who responded at the poles (i.e., strongly do not want to avoid and strongly want to avoid) were primarily categorized as pronatal or antinatal; those who responded in the middle were primarily categorized as ambivalent and indifferent (Huber-Krum et al. 2017).Thus, we created three categories to captures those whose fertility desires were at the poles and those whose fertility desires were not.

Direct measure of contraceptive need:

We asked a single item to directly assess women’s need for contraception. The survey question asked, “considering your life, do you feel it is necessary for you to be using contraception right now?” (binary yes/no). The vast majority of women in the sample who were currently using a method of contraception responded to the direct measure of contraceptive need affirmatively (98%), indicating the face validity of the measure (not shown).

Survey items, response options, and numerical coding are included in Table 1.

Table 1.

Key variables

Construct	Survey item	Response categories and numeric coding
Composite measure of contraceptive need
Sexual frequency1	In the past 30 days, how many times have you had sex?1	• 0 times (1) • 1–8 times (2) • 9 or more times (3)
Perceived fecundity	Considering your life, do you find that becoming pregnant is easy, difficult, or average compared to other women in your community?	• Difficult (1) • Average/same as others (2) • Do not know (2)2 • Easy (3)
Desire to avoid pregnancy	How much do you want to avoid pregnancy in the next year?	• Strongly do not want to avoid (1) • Somewhat or slightly do not want to avoid (2) 2 • Somewhat or slightly want to avoid (2) 2 • Strongly want to avoid (3)
Direct measure of contraceptive need
Need for contraception	Considering your life, do you feel it is necessary for you to be using contraception right now?	• No • Yes

¹Participants who had never had sex were coded as 0.

²For perceived fecundity, average/same as others and do not know were combined. For desire to avoid pregnancy, somewhat or slightly do not want to avoid and somewhat or slightly want to avoid were combined.

Computing the composite contraceptive need score

We summed the sexual frequency, perceived fecundity, and desire to avoid pregnancy variables to create a composite contraceptive need score. Response options that aligned with potentially lower contraceptive need (e.g., no sex in the past month, low perceived fecundity, and strong desire to not avoid pregnancy) were each given a score of 1 and response options that aligned with potentially higher contraceptive need (e.g., sexual frequency of 9+, high perceived fecundity, and strong desire to avoid pregnancy) were each given a score of 3 (coding shown in Table 1). Thus, the possible range of the composite score was 3 (lowest potential need for contraception) to 9 (highest potential need for contraception).

Analysis

We examined the relationship between each of the three variables that made up the composite contraceptive need score and the direct measure of contraceptive need using Chi² tests of independence. We then compared the composite contraceptive need score to the direct measure of contraceptive need using one-way ANOVA. We conducted sensitivity analyses excluding women with no lifetime sexual partners and it did not meaningfully change the findings (Appendix Tables 5 and 6).

Results

Participants ranged in age from 15–44 years (mean=26.8). The majority had some primary education (60%) and 70% had a monthly household income of less than MK 20,000, or approximately $28 USD (average exchange rate in 2017 was 1 USD to MK 725). Most women were married (77%), and on average had 2.5 children (range=0–9) (Table 2).

Table 2.

Socio-demographic characteristics of women in the UTHA Wave 4 sample (N=906)

	N	%
Age (mean=26.8; range=15–44)
15–19	137	15.1%
20–24	229	25.3%
25–29	200	22.1%
30–34	147	16.2%
35–39	92	10.2%
40–44	42	4.6%
Missing	59	6.5%
Education (mean years=5.1; range=0–12)
None	92	10.2%
Some primary	542	59.8%
Completed primary	98	10.8%
Some secondary or completed secondary	114	12.6%
Missing	60	6.6%
Monthly household income (in Malawian Kwacha)
< 5,000	330	36.4%
5,000–19,999	307	33.9%
20,000 – 39,999	93	10.3%
40,000 – 99,999	38	4.2%
> 99,999	19	2.1%
Don’t know	118	13.0%
Missing	1	0.1%
Marital status
Married/cohabiting	695	76.7%
Not married/cohabiting	211	23.3%
No. of living children (mean=2.5; range=0–9)
0	115	12.7%
1	194	21.4%
2	193	21.3%
3	165	18.2%
4 or more	237	26.2%
Missing	2	0.2%

Over a quarter (29%) of women did not have sex in the past 30 days (‘low’ sexual exposure), 36% had sex between 1–8 times (‘moderate’ sexual exposure), and 35% had sex 9 or more times (‘high’ sexual exposure) (Table 3). Most women reported that becoming pregnant was easy (68%), while 225% did not know how difficult or easy it would be for them to become pregnant. The majority strongly desired to avoid pregnancy in the next year (79%), while 13% strongly desired to not avoid pregnancy, and the remaining women’s desires fell somewhere in the middle (Table 3). In the direct measure of contraceptive need, 19% of women indicated they did not have a need for contraception.

Table 3.

Variables that make up the composite contraceptive need score by the direct measure of contraceptive need (Chi² tests; N=906)

	Total sample (n=906)		No contraceptive need (n=174)		Contraceptive need(n=732)		p
	N	%	N	%	N	%
Total	-	-	174	19.2%	732	80.8%
Frequency of sex in past month							<0.001
0	259	28.6%	115	66.1%	144	19.7%
1–8	330	36.4%	26	14.9%	304	41.5%
9 or more	317	35.0%	33	19.0%	284	38.8%
Perceived fecundity							<0.001
Difficult	201	22.2%	24	13.8%	177	24.2%
Average or same as others	45	5.0%	9	5.2%	36	4.9%
Easy	613	67.7%	109	62.6%	504	68.9%
Don’t know	47	5.2%	32	18.4%	15	2.1%
Avoid pregnancy							<0.001
Strongly do not want to avoid	118	13.0%	51	29.3%	67	9.2%
Somewhat/slightly want to avoid/not avoid	74	8.2%	17	9.8%	57	7.8%
Strongly want to avoid	714	78.8%	106	60.9%	608	83.1%

We assessed the relationship between each of the three variables in the composite contraceptive need score and the direct measure of contraceptive need. More frequent sex, the perception of being more fecund, and stronger desire to avoid pregnancy were all associated with responding ‘yes’ to the direct contraceptive need question (p<0.001) (Table 3).

We summed the sexual frequency, perceived fecundity, and desire to avoid pregnancy variables to create a composite contraceptive need score. The possible range of the composite score was 3 (lowest potential need for contraception) to 9 (highest potential need for contraception). The observed range was 4–9 (mean=7.18). Scores were not highly concentrated at the poles (Figure 1). Just 2% of women had the lowest observed composite contraceptive need score (4) and 15% had the highest composite contraceptive need score (9)./span>The majority of women (60%) scored between 7–8.

Figure 1.

Distribution of composite contraceptive need score (N=906)

In comparing the mean composite contraceptive need scores between those who directly reported that they needed contraceptive and those who directly reported that they did not need contraception, we found a significant difference. Women who directly reported that they did not need to use contraception had a lower mean composite score than women who directly reported that they needed contraception (6.33 vs. 7.38; p<0.001) (Table 4).

Table 4.

Composite contraceptive need score by the direct measure of contraceptive need (N=906) (ANOVA)¹

	Score (total sample)	No contraceptive need	Contraceptive need	p
Mean	7.18	6.33	7.38	<0.001
Std. Dev.	1.21	0.97	1.17
Range	4–9	4–9	4–9
N	906	174	732

As expected, the majority of women who directly reported a need for contraception scored highly on the composite contraceptive need measure (47% scored an 8 or 9) and an additional 34% scored a 7 on the composite measure (Figure 2). However, some participants’ direct report of contraceptive need and their composite contraceptive need score did not align. While women who directly reported that they did not need contraception were unlikely to score very high on the composite measure (5% scored an 8 and 1% scored a 9), most also did not score very low on the composite measure (5% scored a 4 and 16% scored a 5) (Figure 2).

Figure 2.

Distribution of composite contraceptive need score disaggregated by direct measure of contraceptive need

Discussion

Contraceptive counseling that includes an assessment of a range of factors allows people who are capable of pregnancy to share information with their healthcare provider on potential exposure to and risk for unintended pregnancy along with their pregnancy intentions and desires. These factors – independently and in aggregate – provide nuanced insight into what constitutes contraceptive need that can be useful to healthcare providers and community-based family planning programs. Our findings suggest that including questions about sexual frequency, perceived fecundity, and pregnancy desire may help providers understand an individual’s contraceptive need and provide important context as part of contraceptive counseling. Ultimately, we argue that people are capable of knowing if they want or need to use contraception, and such knowledge must drive contraceptive provision (Senderowicz, 2020; Skogsdal et al., 2019).

Current counseling protocols are predicated on pregnancy desires, such that if a person indicates they want to become pregnant, they are considered not to have contraceptive need. In reality, people who want to become pregnant may still choose to use contraception for a variety of reasons including economic, relationship, health and other circumstances. People who do not want to get pregnant may also weigh the costs and benefits of contraceptive use and actively choose not to use contraception (Casterline, Perez, and Biddlecom 1997; Chipeta, Chimwaza, and Kalilani-Phiri 2010; Staveteig 2017). We found that many people who directly reported no need for contraception still scored relatively high on the composite contraceptive need measure (79% of participants who directly reported no need for contraception scored 6 or higher). In such cases, including both the composite contraceptive need score and the direct measure of contraceptive need can enhance contraceptive counseling, focusing on why a person says that they do not need contraception, including asking about low sexual exposure, low perceived fecundity, wanting to become pregnant, or another reason not captured in the composite contraceptive need score.

A benefit of using a composite score is that it can identify when an individual’s exposure to pregnancy and desire to become or not become pregnant do not align with their expressed need for contraception. An additional strength may be the opportunity to identify and counsel individuals who want to become pregnant, but may have low sexual exposure or perceive themselves to have low fecundity. People appreciate when providers elicit their individual needs and provide counseling that addresses issues that are relevant to them and their circumstances (Callegari et al. 2015). Information about sexual frequency, perceived fecundity, and desire to become pregnant can guide contraceptive counseling and may also be an entry point into counseling around preconception care if the client either indicates that they want to become pregnant or that they are choosing not to use contraception, and recognize they are at risk of becoming pregnant.

Historically, assessing contraceptive need has focused on increasing use of contraception as a means of preventing an undesirable outcome – unintended pregnancy. Met contraceptive need has even been used as a proxy for reproductive autonomy (Potter et al. 2019). This approach assumes that contraceptive use is positive, even if it does not align with people’s stated desires to not use contraception (Potter et al. 2019; Senderowicz 2020). In reality, some people, particularly in highly pro-natalist settings, may prefer to err on the side of becoming pregnant, and thus decide not to use contraception even if they indicate that they do not currently intend to become pregnant (Bornstein et al. 2022). Indeed, attitudes towards a hypothetical pregnancy may be more predictive of contraceptive use, in some settings, than perceived importance of pregnancy prevention (Geist et al. 2019). This is further evidence that pregnancy “planning” is not a priority for all people, and attitudes towards pregnancy prevention/achieving pegnancy do not always align with contraceptive use.

The composite measure of contraceptive need provides information useful for counseling to facilitate or to prevent pregnancy. Together, the individual items and the composite measure can provide an overall sense of contraceptive need to facilitate counseling relevant to individual client needs. Although providers’ time and facility resources are constrained, perhaps particularly in low-resource settings like rural Malawi, inclusion of these additional measures would be relatively simple and better facilitate individualized counseling, although we do not know how feasible or acceptable these measures would be to providers and their clients. Thus, additional studies, including feasibility studies, should be done to understand if these measures may be useful. Piloting a counseling protocol that includes the items that make up the composite measure, along with the direct question regarding contraceptive need, would provide further insight into how scoring of the composite measure might be refined to better reflect levels of contraceptive need in specific settings.

Contraceptive counseling that is framed primarily by pregnancy desires and intentions may miss nuance in how people think about their need or desire to use contraception. By discussing additional factors, contraceptive counseling can be more comprehensive, allow for nuanced discussions about the factors that may contribute to one’s overall assessment of their need for contraception, and be tailored to individuals.

Data availability

More information about the data can be found at https://u.osu.edu/utha/ Please reach out to Alison Norris (Norris.570@osu.edu) for further information regarding data availability.