With adaptation, the WHO guidelines on calcium supplementation for prevention of pre‐eclampsia are adopted by pregnant women

Abstract

The World Health Organization (WHO) recommends calcium supplementation for prevention of pre‐eclampsia, but factors affecting adoption and acceptability of the recommendations among pregnant women have not been examined. We explored adoption of the WHO guidelines in Kenya, using the trials of improved practices. We recruited 38 pregnant women and assigned participants to three regimens representing potential trade‐offs among daily dose, bioavailability, and acceptability. Participants were provided with supplements, requested to select preferred product type, counselled on how to take them, and interviewed 4 times over 6 weeks to assess their experiences. We tracked bottle opening with electronic monitors, as proxy for supplement consumption. We analysed interview transcripts thematically. All participants were willing to try the supplements. Average daily consumption ranged from 77 to 1,577 mg/d. Most participants (74%) chose the chewable product. Participants preferred its “sweet taste” and liked the ability to consume it without water. Women in the 2‐dose regimen were least likely to switch; however, women assigned to the 3‐dose regimen, or who switched to the 3‐dose regimen, consumed the most calcium per day. Difficulties with the 4‐dose regimen included afternoon doses when women were likely to forget and having to wait hours after supper for last dose. Use of an illustrated calendar, keeping supplements in conspicuous locations and requesting support from relatives were strategies that supported adherence. Pregnant women are likely to adopt Ca supplementation, with appropriate programmatic adaptations. Careful attention to product attributes, regimen complexity, and strategies for reassuring and reminding women are needed to adapt the WHO guidelines.

1. INTRODUCTION

Hypertensive disorders in pregnancy, including pre‐eclampsia, are major contributing factors to maternal and perinatal mortality (Kassebaum et al., 2014; Say et al., 2014), but incidence is reduced by calcium supplementation in populations with inadequate dietary intake (World Health Organization [WHO], 2013). Although the pathogenesis of these disorders has not been fully elucidated (Mol et al., 2015; Pennington, Schlitt, Jackson, Schulz, & Schust, 2012), populations with inadequate dietary Ca intake have been shown to be at greater risk. Systematic reviews of randomized controlled efficacy trials found Ca supplementation significantly prevents pre‐eclampsia, reducing the risk of developing the condition by half (Hofmeyr, Lawrie, Atallah, Duley, & Torloni, 2014). Based on this evidence, the WHO issued a strong recommendation, urging introduction of Ca supplementation in populations with low habitual intake of calcium, as part of routine antenatal care programmes, to prevent pre‐eclampsia (WHO, 2013). The recommended regimen in the WHO guideline involves 1.5‐ to 2.0‐g elemental calcium in divided doses not exceeding 500‐mg elemental calcium per administration to optimize calcium bioavailability. The WHO guidelines further urge separation of iron‐folic acid (IFA) administration from calcium administration by several hours to avoid potential inhibition of iron absorption by calcium supplements (WHO, 2013). This implies that prenatal micronutrient supplementation would involve at least four pill‐taking events daily.

The current challenge is to translate the global guidelines into effective national and subnational policies and programmes (WHO, 2013), but there is little empirical evidence regarding factors that influence implementation of the WHO guidelines at different levels of policy and programme implementation (Omotayo, Dickin, O'Brien, et al., 2016; Palacios & Pena‐Rosas, 2010). Understanding such factors is important to guide optimal policymaking, programme design, and implementation. Most studies examining factors affecting acceptability of micronutrient recommendations have focused on iron supplementation among end‐users in low‐income settings (Aguayo et al., 2005; Dairo & Lawoyin, 2006; Galloway et al., 2002; Galloway & McGuire, 1994; Seck & Jackson, 2008; Tessema, Jefferds, Cogswell, & Carlton, 2009; Young & Ali, 2005). Although the literature on iron supplementation programming provides relevant insight, there are important differences that warrant formative research relating specifically to calcium supplementation. For instance, preventive antenatal iron supplementation is recommended once daily for 90 days. Recommended calcium supplementation regimen is more complex and daily adherence to a medication regimen is usually affected by the number of times a day the regimen (number of pills) needs to be taken and the duration of medication. Moreover, our earlier research found that pregnant women found IFA supplements more acceptable when counselled about benefits, although they were not familiar with hypertension in pregnancy or pre‐eclampsia as a syndrome (Martin, Omotayo, et al., 2017). A study in Dhaka has examined relative acceptability of different delivery vehicles for calcium supplementation among pregnant women and concluded that conventional pills (hard tablets that need to be taken with water) were more acceptable than chewable pills in Bangladesh (Baxter et al., 2014). However, there have been no published studies comprehensively examining patient acceptability of the various features of the WHO guidelines on calcium supplementation, including attributes of delivery vehicle, daily regimen, and other potential barriers and facilitators.

The main objective of this study was to explore factors that influence adoption and acceptability of calcium supplementation for prevention of pre‐eclampsia among pregnant women in rural communities in western Kenya. Specifically, we sought to answer four questions: (a) Will pregnant women in Kenya adopt calcium supplementation? (b) How do product attributes affect acceptability? (c) How does regimen affect acceptability and adoption of supplementation? and (d) What are the challenges, barriers, facilitators, and strategies associated with adoption and acceptability of the guidelines among pregnant women? Given that prenatal calcium supplementation is likely to be introduced in the context of ongoing prenatal IFA policies, we examined our study questions in the context of regimens including IFA supplementation and provided IFA supplements alongside the calcium supplements. For this study, we adapted the definitions of Proctor and co‐authors for adoption and acceptability (Proctor et al., 2011). Thus, we define acceptability as the perception among pregnant women that calcium supplementation is agreeable or desirable. We define adoption as the intention, initial decision, or action to consume calcium supplements.

Key messages.

• Adapting product attributes, regimen complexity, and strategies for reassurance and reminders can facilitate adoption of the WHO guidelines on calcium supplementation by pregnant women.

• Lack of counseling and reassurance about side effects, in the face of discouragement from relatives, can constitute a barrier to adoption and acceptability of calcium supplementation.

• Sweet chewable tablets that were consumable without water enhanced acceptability of calcium supplementation among pregnant women in Kenya.

• A regimen with 2 doses of 500‐mg elemental Ca/day, without separating calcium from iron consumption, was preferred by women, but prescribing 3 doses per day led to more calcium intake.

2. METHODS

2.1. Study design

We employed the trials of improved practices (TIPs) approach to explore adoption and acceptability of the WHO recommendations among pregnant women. TIPs is a flexible adaptation of commercial marketing techniques, developed by the Manoff Group (Manoff Group, n.d.). It is a research approach that assesses the acceptability of recommended health behaviours through trials at the usual site of implementation, during which people are asked to try new behaviours and participate in follow‐up evaluation regarding their experiences and willingness to adopt the behaviours (Dickin, Griffiths, & Piwoz, 1997).

We selected six communities as study sites in Malava sub‐county in western Kenya. Malava sub‐county was selected as study site because of its proximity to the sites of an integrated maternal and child health programme that was being planned by study collaborators at the time of study design. The communities were purposively selected to maximize diversity based on rural/urban characteristics, proximity to healthcare facilities, and accessibility of the healthcare facility in the community. During community sensitization, a list of all pregnant women in the six communities was compiled by the field team through door‐to‐door visits throughout the selected communities, with information on parity, age, and educational status. A sampling scheme designed to maximize diversity based on parity, age, educational status, and daily time spent away from home was prepared. On the basis of the sampling scheme, we purposefully selected five to seven women from each of the communities (38 women altogether) to participate in the study. Inclusion criteria were (a) age 15 years and above, (b) gestational age from 16 to 30 weeks, (c) not having plans to relocate from the community within the next 6 weeks, and (d) inadequate daily calcium intake according to a screening tool based on self‐reported consumption of milk and dairy products in the preceding week. We also asked about medication because antacids made from calcium salts can be a significant source of calcium intake.

We used two calcium products with different organoleptic properties in the study: hard tablets and chewable tablets. We provided all women with the same type of IFA supplements. The hard Ca supplement was Ostocal calcium and vitamin D3 manufactured by Eskayef Bangladesh Limited. It contained 500‐mg elemental Ca (as calcium carbonate) + 200 IU vitamin D (as cholecalciferol) in each film‐coated tablet. The tablet was white, tasteless, and cylindrical in shape. The chewable Ca supplement was Ideos Chewable Tablet manufactured by Helsinn Birex Therapeutics Ltd. It contained 500‐mg elemental Ca (as calcium carbonate) + 400 IU vitamin D (as cholecalciferol) in each tablet. The tablet was sweet, greyish white, and cuboidal. All supplements were purchased through Madawa Pharmaceuticals, Nairobi, Kenya. The cholecalciferol content of the tablets were due to pragmatic limitations. We were not able to find two different calcium pill types with the same amount of vitamin D content. Moreover, we considered the difference in cholecalciferol content between pill types to be unlikely to affect our research questions.

We randomly assigned the participants to three different regimens. The regimens differed in the number of separate times that supplements were required to be administered daily (Table 1). They were based on consideration of potential trade‐offs among total daily prescription, bioavailability, and acceptability. Although the three pill‐taking events and four pill‐taking events regimens had higher total daily prescription, we hypothesized that the two pill‐taking events regimen will be more acceptable. Combining Ca and IFA, which is the difference between the three pill‐taking events and four pill‐taking events regimens, was considered reasonable to reduce regimen complexity and potentially improve acceptability, given the limited evidence of large effects of interaction on haematological indices over time (Lönnerdal, 2010; Omotayo, Dickin, O'Brien, et al., 2016).

Table 1.

Daily pill‐taking regimens in the trials of improved practices of calcium and iron‐folic acid supplementation

Daily pill‐taking regimens
Regimen	First administration	Second administration	Third administration	Fourth administration
Two pill‐taking events	One pill Ca (500‐mg elemental Ca as CaCO3 + 200/400 IU of vitamin D as cholecalciferol)	One pill Ca (500‐mg elemental Ca as CaCO3 + 200/400 IU of vitamin D as cholecalciferol) with one pill of iron‐folic acid (60 mg of elemental iron as FeSO4 and 400 mcg of folic acid)	N/A	N/A
Three pill‐taking events	One pill Ca (500‐mg elemental Ca as CaCO3 + 200/400 IU of vitamin D as cholecalciferol)	One pill Ca (500‐mg elemental Ca as CaCO3 + 200/400 IU of vitamin D as cholecalciferol)	One pill Ca (500‐mg elemental Ca as CaCO3 + 200/400 IU of vitamin D as cholecalciferol) with one pill of iron‐folic acid (60 mg of elemental iron as FeSO4 and 400 mcg of folic acid)	N/A
Four pill‐taking events	One pill Ca (500‐mg elemental Ca as CaCO3 + 200/400 IU of vitamin D as cholecalciferol)	One pill Ca (500‐mg elemental Ca as CaCO3 + 200/400 IU of vitamin D as cholecalciferol)	One pill Ca (500‐mg elemental Ca as CaCO3 + 200/400 IU of vitamin D as cholecalciferol)	One pill of iron‐folic acid (60‐mg elemental iron as FeSO4 and 400 mcg of folic acid)

We aimed to interview each respondent 4 times (T1–T4) at 2‐week intervals over 6 weeks to assess barriers and motivators over time. Table 2 describes the activities involved in each visit.

Table 2.

Design of household trials of calcium supplements in Kenya

	T1	T2	T3	T4
Screening and demographic survey	X
Willingness to take supplements	X	X	X	X
Assignment to regimen	X
Choice to change from assigned regimen			X	X
Choice of pill type	X
Choice to change pill type		X	X	X
Counselling	X	X	X	X
Provision of MEMS	X
Downloading of MEMS data		X	X	X
Anticipated barriers and facilitators	X
Experienced barriers and facilitators		X	X	X

Note. MEMS = medication event monitoring systems.

Research assistants conducted the interviews in study participants' households. Counselling messages and materials were based on existing behaviour change materials that were developed for iron supplementation by the ministry of health and insights from interviews conducted earlier in the same setting (Martin, Omotayo, et al., 2017). At T1 visit, the essence of the study was explained, the respondent was screened for eligibility and written consent was sought. All selected respondents granted consent. After consent was granted, the respondent was counselled about the potential benefits of calcium and iron supplementation potential side effects and strategies to enhance adherence. The assigned regimen was also explained to the respondent, without informing her about alternative regimens. We also gave calendars to the participants. The calendars had illustrations of the assigned regimen and motivating messages. These calendars also had spaces for the participant to mark as a means of tracking daily supplement consumption. The respondent was also asked about willingness to try the recommendations, requested to choose a supplement product type after close examination (i.e., observing, smelling, and tasting) of both types and provided with Ca and iron supplements and asked about anticipated barriers and facilitators. She was then counselled on how to use the medication event monitoring systems (MEMS; Aprex Corp., Fremont, Calif.). MEMS consist of pill bottles in which the calcium supplements were kept, and covers containing microprocessor monitors that recorded timing of bottle openings. MEMS provide a proxy measurement of adherence or compliance. It is not necessarily a measure of medicating because people can open the bottle without taking the pill or take more than one pill when they open the bottle.

At T2 visit, the participant was asked about the facilitators, barriers, and strategies that affected her experience as well as her experience with the regimen and product. She was asked whether she had tried the recommendations and if she was willing to continue. She was then given the opportunity to choose the other product. Participants were not presented with the option to choose a new regimen until T3 (this is in contrast to product types, which were chosen by participants throughout the study). The T3 interview was similar to the T2 interview. In addition, however, the respondent was given the opportunity to change to another regimen. The T4 interview, which is the last contact with the study participant, repeated data collected at T2 and T3. If a respondent was unavailable on the scheduled interview date, the interview was rescheduled for the earliest possible date after the initially scheduled date. The interviews were mostly conducted in Kiswahili and were audio recorded. The recordings were later transcribed and translated to English. The MEMS electronic data were downloaded from the monitors over time by the study coordinator after interviews.

2.2. Ethical aspects

This study was reviewed and approved by the Institutional Review Board at Cornell University and Kenyatta National Hospital and University of Nairobi Ethics and Research Review Committee. All respondents were given detailed information about the objectives and purpose of the study and written informed consent was obtained from each respondent before enrolment.

2.3. Data management and analysis

We analysed interview transcripts thematically, using Atlas.ti qualitative analysis software (Muhr, 2004). One participant's set of four transcripts from T1–T4 was independently coded by two coders (S. L. M. and S. O.) to generate a preliminary code list and code book. The coders then independently coded three participant transcripts from all four TIPs visits, merged the coded transcripts, compared codes, and resolved discrepancies by discussion. Both coders continued independently coding all transcripts and met frequently to discuss and refine the coding scheme, continuously revising the code book and adding emergent codes. Throughout the coding process, common themes and key quotes were noted and similarities and differences across participants and between themes were constantly explored (Lofland & Lofland, 2006). We ran queries and extracted illustrative quotes relevant to the research questions. A matrix for facilitators, challenges, and barriers to supplement consumption was developed (Lofland & Lofland, 2006).

Themes related to hindrance or enhancement of adoption and acceptability of supplement consumption emerged from the qualitative analysis. Where a hindering factor was reported to actually lead to stopping calcium consumption, it was coded as a barrier, but where the hindering factor did not lead to stopping supplement consumption, it was coded as a challenge. We also differentiated between enhancement factors that increased the desire to consume the supplements (coded as motivators) and factors or strategies that enhanced capacity of the respondents to actualize such desire, which were coded as facilitators. We noted factors that were reported in anticipation and those that were experienced. Lastly, we coded factors that seem to have resulted from the study design.

Quantitative data from the interviews were doubly entered into REDCap database software (Harris et al., 2009). MEMS data were exported from the software. The datasets were cleaned in Microsoft Excel 2013, and all subsequent analyses carried out in Stata 14 (StataCorp, 2015. Stata Statistical Software: Release 14. College Station, TX: StataCorp LP). We multiplied the total number of MEMS entries by 500 mg and divided by the number of days in the period under assessment, to calculate average daily consumption for each participant. We excluded interview days because of the likelihood of extraneous bottle openings. There were no outcome data for two women that had relocated from their communities before the first follow‐up interview, and MEMS data were unavailable for a woman that had discarded the bottle cover after delivery although the follow‐up interview was conducted.

3. RESULTS

The 38 study participants contributed 1,161 person–days of study participation. Participants in the study sample were mostly married women living in their matrimonial homes in rural communities and a suburban settlement. They were 16–37 years old, with an average age of 24 years. Parity ranged from zero to eight with an average of three. The average gestational age was 5 months, and the participants spent an average of 5 hours away from home during the day. Most respondents had completed elementary but not secondary education and had not received any antenatal care from a health facility for the index pregnancy. Table 3 shows frequency of key characteristics of study participants.

• 1Are pregnant Kenyan women willing to accept and adopt calcium supplementation?

Table 3.

Participant characteristics in household trials of calcium supplementation in Kenya

Participant characteristics
Variables	Frequency (%)a
≤19 yr old	7 (18.4%)
First pregnancy	7 (18.4%)
Did not complete primary education	14 (36.8%)
Average time spent away from home daily > 6 hr	13 (34.2%)
Attending ANC	15 (39.5%)

Note. ANC = antenatal care.

^aNumber and percentage of participants with a particular characteristic, out of all 38 participants.

All 38 study participants were willing to try the supplements at the beginning of the study. Only three participants subsequently declined to continue taking the supplements at later stages. Two of them declined to continue at T2, with one participant attributing her decision to lack of monetary incentives from the field team, whereas the other cited fear that supplements could be harmful and pressure from her mother‐in‐law. One other participant declined to continue at T3, citing advice from her relatives and fear of harm from the supplements.

In addition, 10 women dropped out of the study before T4, due to delivery or relocation but not reasons related to the recommendations. Of these 10 women, three dropped out before T2, four before T3, and three before T4. Electronic and self‐reported data show that all participants actually tried the supplements. The lowest average daily supplement consumption for a participant was 77 mg/d whereas the highest was 1,577 mg/d. Figure 1 shows the distribution of supplement intake among study participants.

• 2How do product attributes influence adoption and acceptability of calcium supplements?

Figure 1.

Average daily calcium supplement consumption (mg/d) among pregnant women in Kenya

Respondents were offered two products with different organoleptic properties at the beginning of the study. Nearly three quarters (n = 28/38) of respondents chose the chewable product type after examining (observing, smelling, and tasting) the organoleptic properties of both product types.

The supplement types differed with respect to more than one attribute. The qualitative data provide insight into which characteristics likely influenced participants' choices. Those who chose the chewable type alluded to its sweet taste and the ability to consume it without water as the reasons for their decision. Some participants who preferred chewable pills mentioned the convenience of not needing to look for water (for example, when working on the farm); hard pills got stuck in the throat if swallowed without water. Others associated swallowing pills with water with nausea. Among participants that indicated preference for the hard product type, conventionality, lack of smell and relatively small size were cited as reasons. Women that changed from the chewable to the hard product type mostly did so out of curiosity. These quotes illustrate the reasons why participants chose particular product types.

“Those to be taken with water (hard type) are bitter … they have bitterness … I fear taking them … many times if I take medicine with water I feel like vomiting”

“I don't think I should change to that one that is taken with water … it isn't as sweet as the one for chewing … the first one (chewable) is easy to remember to take …. I will just take it and eat it but for the second one I will have to search for water first before taking it”

However, a few individuals expressed dissatisfaction with the “smell” and relative large size of the chewable type as illustrated by the quote below.

“… that (chewable) one, I am not able to take it … it gives me a bad smell …, it (chewable) cannot pass here (points to throat) because it is big …. I like this (hard type) just because I can take it with water.”

• 3How does regimen influence adoption and acceptability of supplementation?

Initially, regimen was randomly allocated, but participants were allowed to choose a new regimen or to continue with the assigned regimen at T3 (this is in contrast to product types, which were chosen by participants throughout the study). The proportion of women who chose to switch regimen provides information about their comfort with the regimen assigned. Women in the 2‐dose regimen were least likely to switch regimen at T3, indicating their comfort with that regimen. However, women assigned to the 3‐dose regimen, or who switched to the 3‐dose regimen when offered the chance at T3, consumed the most calcium per day, on average (Table 4).

Table 4.

Regimen assignment, regimen switching, and average daily calcium supplement consumption by regimen among pregnant women in western Kenya

Initial assignment	4‐dose	3‐dose	2‐dose
Daily calcium prescription in mg	1,500	1,500	1,000
Number of women assigned at first visit (T1)	12	13	13
Chose to change assignment at third visit (T3)a	5 of 8	5 of 9	3 of 10
Consumption (mg/d) after third visit (T3)b	430	860	540
Consumption (mg/d) prior to third visit (T3)c	775	955	705

^aof those who remained in the study at T3;

^bAverage for all women contributing person‐time to the dosage after the 3rd visit, when participants had opportunity to change from assigned regimen.

^cAverage for all women assigned to each dosage before the third visit, when participants had opportunity to change from assigned regimen

The qualitative data explore reasons for participants' preference of the 2‐dose regimen. Difficulties with the complex and simplified regimen included afternoon doses when women were likely to forget or be away from home, as illustrated by the following quote:

“…. I decided to change because I have discovered that often times I forget to take my pill, either because I am busy or I am away from home and did not carry the tablets with me because I forgot … but it is easy to remember in the morning … it is also easy to remember in the evening because I will just be here at home”

Having to wait after supper for the last dose in the 4‐dose regimen was an additional burden that limited acceptability. The 2‐dose regimen was acceptable because it eliminated both of these concerns. This is illustrated by this quote:

“I will try the morning and the evening one (2‐dose regimen) … because during the day, it is normally hard for me … this (2‐dose regimen) will give me a little relief … I mean there is an interval between morning and evening … I see this one as a bit of relief … because I will take just one in the morning … then in the evening take iron and calcium together … (if) I take calcium and then wait for two hours before taking iron (in the evening), I would easily forget … but this one, I won't forget”

• 4What are the challenges, barriers, facilitators, and strategies associated with adoption and acceptability of the guidelines among pregnant women?

Table 5 summarizes the challenges, barriers, motivators, and facilitators of adherence and supplement consumption in the study. The most salient challenge for participants in the study was forgetting to take the pills and not having the pills around at the right time. This challenge was both anticipated and experienced. This was usually experienced when women were not in their household either because they were out during the day or travelling.

Table 5.

Factors and strategies influencing adoption, acceptability, and adherence to antenatal calcium supplements among 38 Kenyan women in a household trial

Salient themes on barriers, challenges, facilitators, and motivators
	Anticipated	Experienced	Study design‐related
Challengesa	• Forgetting to take pills	• Forgetting to take the pills	• Inability to open pill bottle
		• Side effects (nausea, bloating, burping, stomach ache, diarrhoea tiredness, and loss of appetite)	• Pill bottle too large to carry around
	• Side effects	• Discouragement from relatives and friends	• Not allowed to put pills in any other container
	• Discouragement from relatives	• Pill burden (especially when having additional medication)	• Suspicion about origin of pills (not delivered through health facilities and a member of the field team is a foreigner)
		• Food insecurity
		• Undesirable smell and taste of the pills
Barriersb	• None	• Discouragement from trusted relatives (mother‐in‐law and husband)	• Lack of monetary incentive from research team
		• Lack of reassurance from HCWs in the face of perceived side effects
Motivatorsc	• Perceived benefits (physical and cognitive development of fetus, less bleeding at delivery, and prevent high blood pressure)	• Relief from health concerns (headache, tiredness, dizziness, and swollen legs)	• Knowledge that the MEMS was monitoring pill‐taking activity
		• Reduced eating of soil
Facilitatorsd	• Use of calendars	• Use of calendars	• None
	• Reminder by spouse	• Reminder by relatives (husbands, children, parents, and in‐laws)
		• Placing pills in conspicuous locations (on the table and by the bedside)
		• Taking pills at specific times of the day (meal times, bedtime, and first thing in the morning)
		• Administering medicine to her children

Note. HCWs = health care workers; MEMS = medication event monitoring systems.

^aA hindering factor that did not lead to stopping supplement consumption was coded as a challenge.

^bA hindering factor that was reported to actually lead to stopping calcium consumption was coded as a barrier.

^cEnhancement factors that increased the desire to consume the supplements were coded as motivators.

^dFactors or strategies that enhanced capacity of the respondents to actualize desire to consume supplements were coded as facilitators.

Typical quotes from women who spent time away from home:

“… there was no problem except forgetting. I remembered when (I was already) far away that I should have carried it … the difficult thing is that I go out during the day, then I get late, and I had not carried the pills ….”

“… I didn't take (the pills), not because I did not have them … but because I forgot … but I had them in the bag … because of the issue of dignitaries visiting the school (I forgot), and by the time I remembered, I found out the time had passed ….”

Women also forgot when they were at home, when they were busy or tired.

“When you come from there you know the duties for women are many so you easily forget”

It is not clear from our data the extent to which MEMS‐related factors affected adherence. These factors include our advice that extra pills should not be taken out of the bottles, the awkwardness of the bottle because of its large size, and difficulty of manipulating the bottle cover.

However, women also reported forgetting to take their pills even when they were in the house, particularly when busy or tired. Key strategies that helped to address the challenge of forgetting to take pills when at home included the use of a calendar with illustrations and instructions for daily marking to track consumption, keeping the pill bottle in a conspicuous location, taking the pills at specific times of the day or in relation to specified daily events (e.g., meal times), and reminders from relatives, usually spouses and children. Typical quotes regarding strategies for remembering include

“…. My husband reminds me … and because I have put it (the pills) near, when I see it, I remember … he (husband) tells me to take it every time …”

“…. I would remember every time I am eating breakfast then lunchtime, also at supper and after eating when I am going to sleep … calendar was helping me besides food ….”

Some participants anticipated that unfavourable perception of the pills by relatives could be a challenge, but this did not stop any participant from accepting to try the supplements. Side effects including nausea, bloating, burping, stomach ache, diarrhoea, tiredness, and loss of appetite were also reported, but this generally did not lead to stopping of supplementation. There were several women that encountered these conditions but did not discontinue the pills. Typical quotes regarding side effects include

“… in the beginning when I started I felt headache … but eventually it came down … when I am going to sleep I feel the throat is so dry … and I feel like I cannot swallow saliva … that saliva, when I swallow, I feel like vomiting but the throat is dry … no, I did not stop taking the tablets …”

“… so after the third day I started feeling pain (in the) stomach, then (I was) urinating all the time, then feeling like vomiting all the time. It even forced me to ask another nurse. I told her, this pill what does it do? Or which pill have I taken that makes me feel this way? She told me to continue taking it until it is finished … so I continued…”

Few women discontinued supplementation. Among these women, most attributed their decision to stop taking their supplements to advice from a relative and lack of reassurance from health workers perceived side effects. This was the main barrier identified, although there were other women that continued taking their pills despite such advice.

“…. For the past two weeks, I haven't been using any pills … I only used it when you had visited me … then I felt pain … I decided to visit the doctor and explain it to him … to see how it was …. I haven't yet been given the answers.”

The key motivators include anticipated benefit regarding physical and cognitive health of the fetus, prevention of high blood pressure and “low blood” in the mother and avoiding too much bleeding at delivery. These were consistent with the messages provided in counselling and printed materials given to the women. Experienced benefits such as feeling “strong” and relief from poor health conditions including headache, tiredness, dizziness, and swollen legs as well as reduced eating of soil and help with poor diets motivated some participants to continue supplementation.

“Generally, [the pills] have made me strong. That strength they have given me and I don't feel the pain in my body or that feeling like falling.”

“I feel well because they help me, they give me strength, raise my appetite. I don't eat the soil. I don't have any problems as it was before when I was using other drugs.”

“The benefit I get, there are times I even fail to have money for milk, I lack money to buy fish, but the way I take like this since you said they have those foods, i just feel well … the body is okay, it is strong.”

4. DISCUSSION

Antenatal calcium supplementation is an important component of the comprehensive strategy to address perinatal morbidity due to pre‐eclampsia (Omotayo, Dickin, & Stolzfus, 2016). We examined factors influencing adoption and acceptability of the WHO guidelines on calcium supplementation for prevention of pre‐eclampsia in a rural Kenyan population. The TIPs approach provided a rich set of data to inform a complex set of research questions. To our knowledge, this is the first study to assess multiple features of the guidelines, which might influence adoption, acceptability, and adherence among pregnant women, in the household setting.

4.1. Adoption and acceptability of recommended regimen

The recommended regimen in the WHO guideline involves 1.5‐ to 2.0‐g elemental calcium with separate administration of IFA in a total of at least four pill‐taking events daily (WHO, 2013). Our data suggest that the “standard” regimen (four pill‐taking events) in the guidelines is less acceptable compared to alternative regimens. An alternative (two pill‐taking events) regimen with lower daily recommended dose (1 g) and fewer recommended pill‐taking events was more acceptable. The optimal amount of calcium supplement consumption for prevention of pre‐eclampsia remains unknown (Cristina & Jean‐Pierre, 2010; Omotayo et al., 2016). A meta‐analysis of low‐dose (<1 g) calcium supplementation studies found comparable efficacy to high‐dose (>1 g) studies, but the primary studies in the low‐dose analyses were of poor quality (Hofmeyr et al., 2014). If well‐conducted randomized controlled trials demonstrate efficacy of low dose calcium supplementation in preventing pre‐eclampsia, that might make the two pill‐taking events option the regimen of choice, given the likelihood of higher acceptability. However, if the studies demonstrate significantly lower efficacy, the three pill‐taking event regimen might be most appropriate. It should be noted that this regimen prescribes concomitant iron and calcium consumption. Calcium supplementation has been shown to inhibit iron absorption, however, clinical effects of concomitant consumption are likely minimal over time. Future studies should model the clinical implications of the balance between missed iron and calcium doses due to separation and reduced iron bioavailability due to concurrent consumption.

4.2. Product attributes and acceptability

Our data indicate that supplement product attributes, including being consumable without water, sweet taste, small size, and being odourless, are associated with higher acceptability of calcium supplements. This is consistent with prior studies that have shown that organoleptic properties of pills and supplements influence acceptability (Young, Blanco, Hernandez‐Cordero, Pelto, & Neufeld, 2010). We found that the chewable product type was more acceptable than the hard product type. This is in contrast to findings in the Bangladeshi study, in which conventional tablets were found to be preferable to chewable tablets (Baxter et al., 2014). There are multiple possible explanations for this contrast: (a) It is possible that the differential properties between our hard and chewable product types do not correspond to the differences between the conventional and chewable tablets in the Bangladesh study; (b) it is not clear that the conventional and chewable tablets were recommended in identical regimens in the study in Bangladesh. If regimen was not identical, the effect of regimen could have been confounded with the effect of product type in their study; (c) there are important methodological differences between both studies. Although our study relied mostly on qualitative data in a less controlled environment, the study in Bangladesh employed a modified discrete‐choice trial method; and (d) the difference might be due to sociocultural differences between the two study populations.

It is important to note that although the chewable product type in our study was more acceptable, it was approximately 4 times as expensive as the hard product type, according to the Micronutrient Initiative's unpublished cost information (personal communication from Jacqueline Kung'u). Therefore, it is likely that large‐scale programmes in developing countries will choose the hard product type for large‐scale programming. However, we are unaware of the drivers of differential cost. If such drivers are unrelated to the attributes related to acceptability, it might be possible to produce product types that balance cost with acceptability.

4.3. Barriers, challenges, motivators, and facilitators

Forgetting to take the pills was the most salient challenge reported by participants in the study. Strategies for addressing this challenge included the take‐home brochure that had illustrations on the assigned regimen, motivational messages about why and how to take calcium and IFA pills, and a calendar which assisted women in keeping track of their daily consumption of calcium and iron pills. Other strategies included placing pills in conspicuous places and reminder by relative. This is consistent with prior findings in relation to iron supplements (Galloway et al., 2002). Gastrointestinal side effects were widely reported, but they did not lead to discontinuation of supplement consumption, likely because women had been counselled about side effects. This is also in contrast with the Baxter study that found low incidence of perceived side effects (Baxter et al., 2014). This might be due to the fact that participants in our study were also consuming iron supplements alongside the calcium supplements. Other studies have reported that pregnant women are likely to continue taking their iron supplements in the face of side effects, if they have been appropriately counselled (Galloway & McGuire, 1994). Discouragement from relatives in the presence of perceived side effects constituted a barrier that led to stoppage of supplement consumption in few cases. Prior studies in this population have demonstrated acceptability of encouragement from relatives (Martin et al., 2016). Addressing anticipated side effects and encouraging involvement of relatives through comprehensive counselling will be useful in facilitating acceptability and adoption in large scale antenatal supplementation programmes (Martin, Seim, et al., 2017).

4.4. Contributions, strengths, and limitations

This trial of improved practices, for the first time, examined multiple issues related to implementation of the WHO guidelines on calcium supplementation, among pregnant women in the household setting. Using a mixed methods approach allowed us to apply multiple lines of enquiry to explore our research questions in rich detail, shedding light on questions of key importance to maternal and child health programme managers who are taking steps to integrate primary prevention of pre‐eclampsia into their programmes.

Our findings should be interpreted in the context of several limitations. The small sample size is a limitation. We did not try to rule out chance as an explanation for the findings in our analyses of regimen and product type preferences. We used qualitative analyses to provide complementary information to understand the trends noted in our quantitative analyses. However, this type of study and analyses with small samples is important to understand appropriateness of interventions before involving larger populations. Indeed, we further examined the study questions in follow‐up studies with larger samples, details of which are reported elsewhere (Omotayo et al., 2015; Omotayo, Dickin, Pelletier, Martin, et al., 2017; Omotayo, Dickin, Pelletier, Mwanga, et al., 2017). Second, there is potential for measurement error in the study. The MEMS recordings represent electronic monitoring of bottle openings. The assumption that bottle opening represents pill taking might be erroneous in some cases, potentially resulting in misleading conclusions about adoption. We encouraged study participants to remove one pill from the bottle at a time and only for immediate consumption, in order to minimize misreporting. Analysis of our qualitative data suggests that this instruction was mostly adhered to. Third, pill‐taking behaviour is a sociocultural activity, and the exact nature in which culture interacts with the features of a recommendation might differ across cultures (Young et al., 2010). Studies in other cultures are needed to determine the applicability of our findings across cultures. Fourth, this study was designed to follow up women for 6 weeks. In contrast, the WHO guidelines recommend calcium supplement consumption over the last 20 weeks of gestation. Although data over our study period suggest reasonable supplement consumption, adherence reduced over time. Hence, application of our findings to longer periods of supplementation in routine programmes should be done with caution. Finally, decisions about programme and policy design are complex. The end‐user is just one stakeholder in a complex web of interactions, albeit an important one. Appropriate design will balance findings from our type of study with several other considerations to make optimal decisions.

5. CONCLUSION

In conclusion, we have examined factors affecting adoption and acceptability of calcium supplementation recommendations among pregnant women in a household trial in western Kenya. We find that women are likely to adopt Ca supplementation in pregnancy, with appropriate programmatic adaptations. Chewable tablets that were consumable without water and tasted sweet enhanced acceptability. A regimen with 2 doses of 500‐mg elemental Ca, without separating calcium from iron consumption was preferred, although many women also found a regimen of 3 doses acceptable which led to women taking more calcium during their pregnancy. This is an important finding, given that some women in Kenya and elsewhere are attending ANC later in pregnancy. A regimen that separated IFA from Ca resulting in four pill‐taking events per day was less acceptable. Lack of appropriate counselling and reassurance about side effects, in the face of discouragement from relatives can constitute a barrier to adoption and acceptability. Careful attention to product attributes and cost, regimen complexity, and strategies for reassuring and reminding women is needed to adapt the WHO guidelines to context.

CONFLICTS OF INTEREST

The authors declare that they have no conflicts of interest.

CONTRIBUTIONS

MOO, KLD, RJS and SLM designed the research; MOO, JKK and EM supervised the field work; MOO, SLM and SO analyzed data; MOO wrote the first draft of the paper; and all the authors edited and approved the final version of the paper.

Omotayo MO, Martin SL, Stoltzfus RJ, Ortolano SE, Mwanga E, Dickin KL. With adaptation, the WHO guidelines on calcium supplementation for prevention of pre‐eclampsia are adopted by pregnant women. Matern Child Nutr. 2018;14:e12521 10.1111/mcn.12521