Categorizing FP methods

We classified contraceptives in three ways after synthesizing the categories of contraceptives provided in the DLFPS and the NFHS and FP2020 resources. ‘Modern methods’ included female and male sterilization, intrauterine contraceptive devices (IUCD), injectables, emergency contraception (ECP), oral contraceptive pills (OCP), condoms, the standard days method, and the lactational amenorrhea method (LAM). Temporary effective modern contraception includes IUCDs, injectables, ECPs, and OCPs. Condoms, an otherwise effective and crucial family planning method, were omitted from the temporary effective category because our data showed that its use tends to be haphazard and inconsistent, and therefore not effective. Traditional methods include, but are not limited to, the calendar or rhythm method and withdrawal.

Analysis approach

Mapping the drivers of behavior enables the systematic categorization of behavioral influences, which can then point towards appropriate interventions to change behavior. To structure our analysis, we holistically considered the potential drivers of behavior using the CUBES framework [25]. According to CUBES, contextual and perceptual drivers combine to act as enablers and barriers along an individual’s path from knowledge–encompassing awareness and skills–to intention (or motivation to act) and action and beyond. Layers of influencers can affect these drivers and reach an individual through various channels. Contextual drivers, include factors such as infrastructure (e.g. distance to clinics or availability of supplies), laws, systems and processes (such as referral processes between healthcare providers), demographics, and social norms (e.g. pressure to have many children) [26]. Perceptual drivers include factors such as beliefs around perceived health risks, self-efficacy (‘Will I be able to control or achieve an outcome?’), and outcome expectations (‘Will the outcome be beneficial or not?’) [27]. Emotions, personality, and–largely unconscious and innate–biases are additional perceptual drivers of behavior [28–30]. Influencers can shape a person’s drivers through various channels, whether in person or through media. Each of these drivers can make it more, or less, likely that a person takes action, and together they sum up to an ‘action tendency’. The quantitative data provided insights into the stages of change as well as contextual drivers, while the qualitative data offered added information on contextual and perceptual drivers.

Quantitative methods

We leveraged the three quantitative population-based survey data sets on FP in UP mentioned above. For our analyses we used the latest two rounds of the NFHS: NFHS-3 and NFHS-4, conducted in 2006 and 2016 respectively and the three rounds of the AHS conducted in 2011, 2012, and 2013 [20–22]. The third quantitative data set used was the 2016 District Level FP Survey (DLFPS) conducted by the Technical Support Unit (TSU) for the Government of Uttar Pradesh in 25 high priority districts of UP. Combined the NFHS and AHS, were used to analyze FP method mix in UP overtime. The DLFPS data was also used to analyze geographic variability, the knowledge-intention-action gap in FP method usage and factors associated with FP method uptake. Table 1 shows the quantitative surveys used and their sample sizes.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Quantitative survey Uttar Pradesh sample sizes.

https://doi.org/10.1371/journal.pone.0243854.t001

Qualitative methods

In order to probe deeper into the additional drivers and decision dynamics, the quantitative analysis was supplemented by de novo qualitative research on FP in UP using three methods: in-depth immersion interviews, journey mapping, and decision games. These studies around FP were conducted as part of a larger study of RMNCH behaviors. As a first step, in-depth immersion interviews were conducted to provide preliminary insights into factors influencing RMNCH behaviours, which were consequently used to inform the design of journey mapping and decision games. We conducted research on households, other household influencers, frontline workers, and facility staff involved in FP counseling and supply. For this paper, we analyzed the data from households and their influencers only. All qualitative data collection occurred in 2016.

Sampling.

The sampling process for the in-depth immersion interviews consisted of two stages: selecting the districts and blocks from which households were drawn and selecting households. First, districts were selected based on being within a distance of 30 to 130 kilometers from the state capital, Lucknow, as well as their designation as a high priority district from the government. Within each district, blocks were chosen to reflect a combination of the following characteristics: whether blocks had a government high-priority program in place or not, and whether community literacy levels, the percentage of rural dwellers, and the percentage of community members coming from Scheduled Castes/Scheduled Tribes were high or low (see S1 Appendix for block selection methodology) [22]. 10 different blocks in 4 districts were selected of which 5 were high priority and 5 were not.

For journey mapping and decision games blocks in rural Uttar Pradesh were selected based on the percentages of currently pregnant women registered, women who received 3 antennal check-ups, women who delivered in a health facility, women who received any postnatal care home visits after delivery, couples currently using any modern contraceptive method, and representing self-help group (SHG) and non-SHG. The 5 indicators were classified as good, average and poor based on cut offs derived from the distribution of the data (see S2 Appendix for details on datasets used and selection criteria applied). An overall block classification into good, average and poor was done based on the block’s overall performance across these indicators. For journey mapping, a total of 6 blocks (3 TSU, 3 non-TSU) from 6 different district were chosen and for decision games 24 blocks (18 TSU, 6 non-TSU). For journeys, 6 families- 3 that had never used a modern FP method and 3 that did use a modern method were selected. The chosen families were selected to exclude currently pregnant women or those who had recently experienced childbirth. The chosen families represented a mix of sociodemographic characteristics such as religion, sex of last-born child, caste, economic status and woman’s age. For decision games, participants were selected to ensure uniform caste, age, SEC, and education in each group. They also selected for a balance in stage in the RMNCH journey (from first trimester pregnancy to 6-month-old newborn), place of delivery, if they had a death of a child before 6 months, number of pregnancies, use of family planning, and religion. Participants were recruited across 24 blocks across 21 districts in UP. Table 2 shows the sample size for each method.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Qualitative method sample sizes.

https://doi.org/10.1371/journal.pone.0243854.t002

Decision games.

Ethnolabs are an audio-visual decision game developed by the behavior change company Final Mile. Decision games are designed to overcome the biases such as fear of value judgment, social desirability, and expectation of higher self-control in future that may make respondents in research-settings claim behaviors that may be not be true in real life [34]. The games simulate the real-world context of participants, observes their hypothetical decision-making, and relates their choices to underlying drivers of behavior. Because FP is a topic often considered private, discussing it in the hypothetical game context allowed for participants to disclose beliefs about the topic they might not in personal interviews. 5 participants with similar profiles in terms of age, gender, and income played the 90-minute game at a time. During the game, participants were given live or audio-visual descriptions of unique situations that end with a decision conundrum, with a set of 3 different scenarios (S5 Appendix) among which the participants must choose. Participants chose based on what they believe the majority in the group will select as the most likely outcome among the options, rather than having to state their own opinion, which is designed to help circumvent the typical self-report biases mentioned above. The game involved a scoring system to motivate players. Afterwards, participants engaged in discussions about their experience, decisions, and preferences in the game (see S6 Appendix for discussion guides). Each scenario could test for the involvement of several likely behavioral focus areas. Three FP scenarios were embedded in a larger game of 10 scenarios related to an array of reproductive and maternal health behavior. See Table 3 for the drivers of FP assessed in the decision games.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Drivers of FP assessed in decision games.

https://doi.org/10.1371/journal.pone.0243854.t003

Minimal change in FP usage and types of contraceptives used over time

Combined analysis of the NFHS and AHS reveals that the FP method mix in UP has changed little from 2006–2016 (Fig 1). Overall use of modern contraceptives (mCPR) in UP, however, has increased slowly since 2006 from 29.3% to 31.7% in 2016. This trend of little change in mCPR is true for India overall as well, with mCPR decreasing slightly to 47.8% in 2015–16 from 48.5% in 2005–06 [16]. Despite consistency in method mix, the DLFPS data shows regional variability among high priority districts in UP in method use (S7 Appendix). The mCPR in Mirzapur is 55%, but it is only 13% in Balrampur. This finding suggests understanding the complex range of factors that influence uptake is very important.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Little change in distribution of contraceptive use by method overtime in UP between 2006 and 2016.

Figure shows combined analysis of responses in quantitative surveys described in methods (NFHS and AHS) to the question: “Which method are you currently using to delay pregnancy?”, considering the highest/most used method. Lactational amenorrhea method (LAM) has been deducted from the “Any traditional method” indicator in AHS to match the corresponding NFHS definition for this indicator.

https://doi.org/10.1371/journal.pone.0243854.g001

Awareness-Intention-Usage gaps in FP

To analyze the reasons for the lack of change in mCPR and the types of methods used, it was first necessary to determine where along the action tendency path towards contraceptive use women tend to be. The DLFPS data showed that overall general awareness of modern contraceptives, measured as whether the woman had heard of the method, is high among women in UP. 99% of women are aware of any modern method and 94% are aware of any effective temporary method. Awareness of specific methods ranges from 97% for female sterilization to 84% for IUCD and Injectables. Awareness, however, does not measure the person’s understanding of the side effects of an individual method. The two exceptions are awareness for emergency contraception (ECP) and lactational amenorrhea method (LAM) which are lowest at 29% and 19% respectively. In addition to high awareness, there are gaps between past use and current use of family planning methods along the cascade. 76%, 49% and 15% of women have ever used any method, modern methods, and effective temporary methods respectively, while only 65%, 31%, and 4% of women are currently using methods in the three categories, as shown in Fig 2. Different factors can influence discontinuation. The three main reasons reported for discontinuation of a method are: infrequent sex (28%), method failure (23%) and wanted to become pregnant (29%). Combined, the drop off between rates of awareness and use as well as between past use and current use demonstrate discontinuity in contraceptive use that requires further investigation.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Family planning awareness-intention-use cascade.

The figure shows descriptive analysis of 2016 DLFPS (n = 13,182 married women ages 15–49). Despite high awareness of modern methods (99% of women are aware of modern contraceptives and 94% are aware of modern temporary ones), there are large gaps between awareness, intention, and use. As shown, there is a 46% drop (bracket 1) between women who want to space/limit and women who to intend to use a modern method, a 55% drop (bracket 2) between wanting to space/limit and using any modern method, and a 78% drop (bracket 3) between wanting to space/limit and intending to use any modern temporary method. “Want to space” is defined as the response “not wanting to have child now/soon”. Any modern method includes female/male sterilization, IUCD, OCP, ECP, injectable, condom, standard days method, and LAM (note that % of standard days method and LAM are minimal). Any modern temporary method includes IUCD, ECP, OCP, and injectable. Only non-users were asked about intention to use methods in the DLFPS, hence, added current users to group because we assume that current users had intent.

https://doi.org/10.1371/journal.pone.0243854.g002

Fig 2 shows the entire cascade of method use, from wanting to space or limit childbearing to use of effective temporary modern methods. Desire to space is defined as women who do not want a child now/soon and desire to limit includes those that do not ever want a child and those already sterilized. Desire to space or limit childbearing is also high at 86%. Intent to use is defined as women already using FP methods or non-users who answered they want to use FP in the future. Despite the high percentage of women who want to limit or space, only 65% intend to use any FP method, and only 53% use any method. This suggests that awareness and need alone are necessary, but often insufficient in building intention to use family planning. Moreover, only 40% of women intend to use any modern method and only 31% use any modern method. The additional drop-off between intention and action again suggests that intention may also at times be insufficient to cause uptake. Overall, the data shows a 46% drop from wanting to limit or space childbearing to intending to use modern methods to do so and an additional 11% drop in those who ultimately do. Understanding why these drop-offs exist will help to design effective interventions to drive uptake. Only 8% of women intend to use a temporary modern method and only 4% do so.

S8 Appendix shows that the cascade differs for women who want to limit pregnancy (65% of sample) versus those who want to space pregnancies (21% of sample). Intent and use are higher for women who want to limit pregnancy than for those who want to space. Among women who wanted to limit, 79% intended to use any method and 65% used any method, while among women who wanted to space pregnancies only 51% intended to use any method and only 39% used any method. These findings reveal that the awareness-intention gap surrounding modern contraceptive methods is a significant issue contributing to FP uptake in UP. Results indicate intention to use is limited by factors beyond awareness. It also indicates while raising awareness of methods choice is necessary, it is not sufficient to build intention to use any of those methods.

Influence of contextual factors varies by modern method

There are a number of other contextual factors that may drive intention to use a modern contraceptive method. Four multivariate logit regression models were run on the DLFPS survey data to determine contextual predictors of intention to use female sterilization, condoms, IUCDs, and pills. Contextual drivers included in the model were demographics such as age and caste, reproductive history, contraceptive access, and message exposure. Ultimately, the influence of contextual factors varied by method. Older women were more likely to intend to opt for female sterilization, but younger women were more likely to intend to opt for condoms. Muslims were less likely to intend to opt for permanent methods like sterilization, unlike Hindus, but more likely to intend to opt for condoms and pills. Woman’s and husband’s literacy in intention to use of different FP methods only mattered for condoms. Scheduled Caste/Tribe populations were more likely to have an intention to get sterilized but were less likely to have an intention to use IUCDs. Economic status was positively associated with the intention to use sterilization and pills. Those that were not aware of other methods beyond female sterilization and condoms were more likely to have an intention to use female sterilization and condoms, respectively. This suggests that some women’s choice of these two methods is largely driven by the lack of awareness of other available methods. Having 1 or more son was a significant positive driver of all but IUCD intent to use.

Women that perceived accessing sterilization and IUCD to be hard, respectively, were less likely to intend to use the method, whereas those that perceived accessing other modern methods to be hard, other than sterilization and IUCD, respectively, were more likely to intend to use the method. Receiving counselling on the method by frontline workers (FLWs) and facility staff increases the likelihood of IUCD, pill and condom (FLWs only) intention to use, but not for sterilization intention to use. These findings suggest that the influence of contextual factors vary by modern method type. S9 Appendix shows the complete results of the four regression models. These findings suggest that different methods are preferred by individuals varying in their demographic characteristics, awareness levels, perceptions of access and therefore careful targeting by method should be considered.

Next, to understand the intention to action segment in the FP cascade, we looked at the differences in contextual factors between women who were currently using the method (referred to as ‘users’ for simplicity hereafter) and women who had the intention to use the method but were not using it (referred to as ‘intenders’ for simplicity hereafter). Across methods, intenders had a significantly higher proportion of younger women, women that did not know whether it was easy or difficult to access the method and women that had infrequent sex, as compared to users. Intenders of all methods, except IUCDs, had significantly higher proportion of women with fewer number of sons than their corresponding users. Sterilization and pill intenders had a significantly higher proportion of poor women as compared to sterilization and pill users. IUCD intenders had a significantly higher proportion of Muslims as compared to IUCD users. S10 Appendix shows the complete results. These findings suggest that certain contextual factors need to be specifically addressed to close the intention to action gap, and that these can be different for different methods.

Household decision making varies by method

Personal relationships within households are another important driver of FP behavior and household dynamics play a role in decision-making. Decisions around family planning are a product of discussions and beliefs shared between husbands and wives. Data from interviews, journey mapping, and decision games revealed different household decision making dynamics by modern contraceptive method. For IUCDs and sterilization, there was evidence of a syncretic dynamic, where the woman and man decide together. During the decision games when a group of men was asked who made decisions about sterilization, one responded, “both [husband and wife] should decide.” Condoms, in contrast, may be decided by men, and the woman may have almost no power over this decision. One woman during the decision games explained, “we only use condom, my husband don’t want to go for any other method.” Women are more autonomous, however, when it comes to injectables. One woman preferred injections because “there is no risk of any disease in case of injection, in Copper T there is more risk therefore I find injection to be proper.” Overall, wives have little exposure to FP outside the family, and are mostly influenced by their husbands. One woman explained during an in-depth immersion interview that it was her husband alone who chose to use condoms, a method which she had no awareness of until marriage. Men take part in more discussions with friends within the village and are exposed to outside media, and MILs have a neighbor network in the village.

Risk perception, shame, and fear often drive FP decision making

In addition to household dynamics, a number of perceptual behavioral drivers can influence FP behaviors. Shame, social pressures, and fear of health risks were all found to play a role in decision-making. Fig 3 shows the interaction role of these drivers across the action pathway towards family planning uptake for different household members.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. FP journey.

Figure shows a summary of qualitative journey mapping outputs (n = 6 women, 6 husbands, 6 mother-in-law (MIL), 6 ASHA, and 6 friends/family). The combined journey mapping analysis illustrates the different beliefs that either facilitate (+) or inhibit (-) progression to the next action tendency towards contraceptive use. Health beliefs are those about health risks to the mother and/or baby. Social beliefs include community norms and structural beliefs include anything related to family structure. At each action stage (awareness, intention, action, and consolidation) health, social, and structural beliefs determine whether a woman, husband, or mother-in-law (yellow and red lines) will seek more information, intend to use FP, use FP, and form FP habits. Influencers (circles) contribute to stakeholders’ beliefs at each stage.

https://doi.org/10.1371/journal.pone.0243854.g003

Strong social norms against and shame to obtain FP are widespread

The qualitative data shows that shame, often in relation to violating social and religious norms, is a key perceptual barrier to contraceptive use in UP. Analysis of journey mapping and decision games shows that social norms influence intentions and uptake of modern methods in UP. Strong norms to demonstrate fertility, have many children, and have at least one male child were all cited in interviews as sources of reluctance to use methods. As one husband explained during the decision games, “I have four children. At the start I had one boy. But I wished to have two more boys. Due to this I now have two girls and one boy. So now I will have four children.” One woman shared her perspective on social norms: “Everyone who got married at the same time as me are having two kids and I did not even have one so people used to say everyone is having kid why are you not having a kid yet. So, I thought I should also have a kid now.” In addition to these strong norms, general feelings of shame to procure contraceptives by women and men were also mentioned as barriers to uptake. On the topic of obtaining information about contraceptives, one woman admitted, “I will ask my sister in law because I am too shy to ask anyone else.” For some individuals, religious norms against FP played a role as well.

Risk perception is high for contraceptive side-effects, but low for early or unspaced pregnancy

Another important perceptual driver is risk perception, which was found to be high for some modern contraceptive methods. Risk was perceived both around potential side effects of methods, ineffectiveness and in the case of condom use, decreased sexual pleasure. It is especially high for more intrusive and surgical methods: IUCDs and sterilization. Risk perception around infertility was notably high for temporary methods, with women suggesting that contraceptives could cause them to become infertile. “By using Copper T women lose more blood and womb is damaged.” This misconception is mentioned for many modern temporary methods and appears to be a significant barrier to use. One woman commented on the IUCD during an in-depth immersion interview: “Even if someone kills me, I will not get it done, I don’t like Copper T and all that, due to bleeding, there is a risk factor.” Risks were also noted for OCPs. A husband commented, “My wife is weak and I don’t think pills are safe it does not suit anyone.”

In contrast, risk perception around the consequences of not using any FP method appeared at times to be absent or low, such as little perceived risk around early pregnancy, multiple pregnancies, or low spacing. When asked about how long she waited to have children after marriage, a woman responded, “We did not think about it like that, when it happens it happens.” This suggests that the causal links are not made between FP and the risk of these undesired outcomes. Households often do not always recognize the potential financial or health risks to the mother and child posed by unplanned pregnancies, having children closer together than desired or more children that desired. For those that use FP, however, financial risk of unplanned pregnancy can be key rational. One woman commented on her husband’s beliefs around FP, “He said it’s good that we are going to have baby, big family is not affordable but let our first child come.”

Decision making pathway

The above sections highlight several contextual and perceptual drivers of FP behaviors that, when considered in combination, indicate a current decision pathway for families has three potential main branches (Fig 4). The pathways ultimately lead to either adopting a permanent method or giving up on FP altogether. First, families may avoid FP, stemming from a combination of strong social norms against FP, low risk perception around pregnancy, concerns about side effects, and feeling ashamed to discuss FP. Second, families may adopt temporary modern methods as they want to control family size and better health and consider finances. After adopting temporary contraceptives, some families may eventually give up on FP after experiencing difficulty and a low sense of control with temporary methods, a belief that abstinence is safe, feeling ashamed to procure contraceptives, and experiencing decreased sexual pleasure with some methods. Alternatively, families finally adopt a permanent method after having difficulty using temporary methods consistently, and because of a belief that permanent methods have no side effects.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. FP decision tree.

Analysis of qualitative findings suggest three potential pathways for FP. Interaction of social norms, risk perception, beliefs about side effects, shame, and other beliefs lead families to 1. Avoid FP altogether, or 2. Adopt a temporary method and eventually a permanent one or 3. Adopt a temporary method but additionally give on FP altogether. Additional quantitative research is needed to confirm pathways.

https://doi.org/10.1371/journal.pone.0243854.g004