Complementary and Alternative Medicine Utilization among Pregnant Women Attending Antenatal Care Clinics in Tongaren Sub-County, Kenya: a Cross-sectional Survey

Background: In Kenya, Complementary and Alternative Medicine (CAM) has been used in almost every culture to manage intolerable symptoms of pregnancy since ancient times. However, studies in Kenya have focused on herbal medicine which is just one of CAM practices. In rural settings (such Tongaren Sub-County), no studies have ever evaluated the prevalence and determinants of CAM use during pregnancy including the common CAM modalities used, and so there is need to bridge these knowledge gaps. Purpose: This study aimed at determining the prevalence of CAM utilization and associated factors among pregnant women attending antenatal care clinics in Tongaren Sub-County, Bungoma County. Methods: A cross-sectional descriptive study design was employed in collecting data from 340 pregnant women attending antenatal clinics in Tongaren, Kenya. Quantitative data was collected through semi-structured questionnaires. Chi-square test was used to establish the associations between socio-demographic variables and CAM usage, and multiple logistic regressions were used to examine the relative effects of the various aspects of the variables which were significantly associated with CAM usage. Results: Among the 340 participants, 50.7% indicated using CAM modalities during pregnancy, indicating that CAM use prevalence in Tongaren is about 50%. 55.80% of CAM users concomitantly use CAM and conventional medicine. The high CAM use prevalence among participants was motivated by the preference of CAM over conventional medicine for certain illnesses (31.4%) and the perceived lack of response to conventional medicine (20.1%). It was noted that marital status, educational attainment of the spouse, employment status, occupation type, and household monthly income level are the major determinants of CAM usage. From the logistic regression analysis, the variables positively associated with CAM use among pregnant women were married status (Odds ratio (OR) = 2.341, p = 0.008), having a spouse with informal education (OR = 5.371, p < 0.001), selfemployed status (OR = 0.415, p = 0.028), farming occupation (OR = 2.777, p = 0.206), and household monthly income higher than 35,000 Kshs (OR = 0.093, p = 0.012). Conclusions: Due to the widespread trust in the safety and efficacy of CAM, the utilization of CAM during pregnancy is common among women in Tongaren Sub-County, which is a rural setting, and therefore CAM can be regarded as the backbone of rural health care in Kenya.

Sub-County), no studies have ever evaluated the prevalence and determinants of CAM use during pregnancy, and so there is need to bridge this knowledge gap.

OBJECTIVE
The purpose of this present study is to determine the prevalence of CAM utilization and associated factors among pregnant women attending antenatal care clinics in Tongaren Sub-County, Bungoma County.

I. Study Area
The study was undertaken in Tongaren Sub-County, located in Bungoma County, Kenya. The County borders Trans-Nzoia County to the south, Uasin Gishu County to the south-west, and Kakamega County. This region was initially a white settlement scheme before the end of the colonial period hence it is currently a multi-tribe context consisting of immigrants from other Kenyan counties. The Tongaren Sub-County is divided into six political wards namely Naitiri-Kabuyefwe, Ndalu, Mbakalo, Milima, Tongaren, and Soysambu-Mitua. The population of this area is 186,279 with 90,610 males and 95,669 females ( (KNBS), 2014). Residents in this constituency have an illiteracy level of 17% which is the lowest in the whole county of Bungoma ( (KNBS), 2013) Tongaren Sub-County has one Sub-County Hospital, two public health centres, fourteen dispensaries, and numerous privately-owned clinics. The distribution of these health facilities in the Tongaren Sub-County is shown in Figure 1.

II. Study Design and Participants
A descriptive cross-sectional survey was conducted among pregnant women attending antenatal care clinics in Tongaren Sub-County, of Bungoma County, Kenya. All consenting pregnant women between the ages of 18 to 49 years were included in the study. Pregnant women who were below 18 years and did not consent to participate were excluded from the survey.

III. Sampling procedure and sample size
The formula used for determining the sample size was the Cochran's formula n= [z² p (1-p)] ⁄d², where n is the requisite sample size, z represents the standard normal deviation 1.96 at 95% confidence level, p is the Proportion of the population with the desired characteristic which is estimated to be 70% of Kenyans who use CAM, and represents 0.05 margin error. From this equation, a sample size of 340 was determined for this survey as shown: n = 1.96 2 x 0.7 (1 -0.7) = 340.
The 340 pregnant women were chosen through a systematic random sampling approach.

IV. Data Collection Instruments
The data required for this survey was primarily collected using semi-structured questionnaires. The questionnaire contained both open-ended and closed questions. To ensure the validity of the questionnaire, the questionnaire was subjected to expert scrutiny to ensure that the questionnaires were valid for the test. The questionnaire was then pretested on a sample of 35 respondents drawn from Namarambi Dispensary in Bungoma East Sub -County.

V. Data Collection Procedure
Data collection was conducted in the months of May and June 2021 in public health facilities within the study area. During the actual data collection exercise, pregnant women attending antennal care clinics were approached as they were about to leave the health facility. They were then given a detailed explanation of the purpose of the study, willingness to leave the study, and an assurance on the confidentiality of information to be collected. They were then asked to fill and sign a consent form indicating their willingness to participate in the study. The questionnaire was then administered to the respondents for data collection.

VI. Data Processing, Analysis, and presentation
Data was organized in MS Excel 2016 before it was exported to the SPSS Version 24.0. Percentages, frequencies, means, and standard deviation were employed in describing the survey variables. Chi-square test of independence was carried out to compare statistical associations between nominal variables whereby a p-value of ≤ 0.05 was considered significant. The variables whose p-value was less than 0.05 in the chisquare analysis were transferred to multinomial logistic regression analysis to examine the relative effects of selected socio-demographic variables on CAM use.

VII. Ethical considerations
Before data collection, ethical clearance was obtained from the ethical review committee of the University of Eastern Africa, Baraton. After which, a research permit was obtained from the National Commission on Science, Technology, and Innovation (NACOSTI). Written consent from the informed respondents who met the inclusion criteria was obtained before the administration of the questionnaire. The information obtained was kept confidential.

I. Response rate
A total of 340 consenting pregnant women who satisfied the inclusion criteria participated in the study by completing and returning the 340 questionnaires required; hence a 100% response rate was achieved.

II. Socio-Demographic characteristics of the study participants
The average age of the respondents was 26.9 ± 7.2. Of the majority of the respondents, 46.5% of were from Naitiri-Kabuyefwe ward, 77.9% were from the Luhyas ethnic, 49.4% had attained up to secondary education, 38.2% had spouses who were holders of college diplomas, and 95% professed the Christian faith (Table 1).  Sixty-five percent of the respondents were married or cohabited while thirty-four percent were single/widowed/divorced ( Figure 2). Many of the survey participants (n=122; 35.9%) were farmers, followed by 102 (30%) respondents who were students, 80 (23.5%) were involved in trading, and 36 (10.6%) were working as civil servants (Figure 4).

Figure 4: Respondents occupation type
Results of the income distribution indicated that majority of the survey participants (n=138; 40.6%) were from households receiving an average monthly income between 5001 and 15000 Kenyan Shillings while 129 (37.9%) were from ones earning less than 5000 Kshs. Only a few respondents (n=73; 21.5%) were from households having an average monthly income above 15000 Kshs ( Figure 5).
14.4% 37.9% 47.6% Of the majority of the respondents, 50.3% were living 5-10 km away from the nearest health facility, 54.4% had attended the ANC twice during the current pregnancy, and 94.7% had Linda Mama insurance cover offered freely by the Kenyan government for women in maternity (Tables 2). The mean age of pregnancy of the respondents was 5.3± 1.9 months. The mean number of times of the respondents being pregnant was 2.3± 1.8. For those who had been pregnant before the current one, the average number of living children was 1.3± 1.7 (Table 3).

III. Prevalence of CAM utilization
Half (50.7%) of the participants had used CAM products at least once during past or present pregnancy. Of the 172 CAM users, 23.5% were from the Naitiri-Kabuyefwe ward, 40.9% were from the Luhya ethnic group, 39.7% were married or cohabitated, 22.4% had attained up to secondary education, 27.1% had spouses having college diplomas, 47.9% professed the Christian faith, 25.3% were self-employed, 25.6% were from households receiving an average monthly income of between 5001 and 15000 Kenyan Shillings, 24.7% were living less than 5 km away from the nearest health center, 25.9% had attended the ANC more than three times, and 47.3% had Linda Mama insurance offered freely by the government for women in maternity (Tables 4,5 and 6).

IV. Factors affecting CAM utilization
The pre-disposing factors for the use of CAM among pregnant women were analyzed using a combination of chi-square test of independence (Tables 4, 5, and 6) and multiple logistic regression analyses (Table 7). It was revealed that marital status of the respondent was significantly associated with the use of CAM (χ2 = 24.610, p < 0.001). Pregnant women who were married/cohabited were found to be 2.34 times more likely to use CAM than those who were single/widowed (OR = 2.341, 95.0% CI: 1.709 -13.372, p = 0.008).
According to the analyses, there is a significant association between educational attainment of spouse of the respondent and the use of CAM (χ2 = 12.086, p = 0.034). Pregnant women whose spouses had informal education were 5.37 times more likely to use CAM than those with other educational qualifications (OR = 5.371, CI: 0.391 -15.945, p < 0.001).
Also, employment status of the respondent was found to be significantly associated with the use of CAM (χ2 = 22.473, p < 0.001). The odds of using CAM were found to be lower among self-employed participants than unemployed ones (OR = 0.415, CI: 0.161 -1.068, p = 0.028).
Further, it was established that there exists a significant association between occupation of the respondent and the use of CAM (χ2 = 27.011, p < 0.001). Pregnant women who were farmers were found to be 2.77 times more likely to use CAM than other occupations (OR = 2.777, CI: 0.57 -13.531, p = 0.206).
Finally, the average monthly income of the respondent's household and the use of CAM were found to be significantly associated (χ2 = 25.559, p < 0.001). Pregnant women from households earning more than 35,000 Kshs per month were at lower odds of using CAM than those with lower monthly income (OR = 0.093, CI: 0.015 -0.6, p = 0.012).

V. Reasons for CAM utilization
The three most commonly provided reasons for CAM utilization among pregnant women attending ANC clinics were "I preferred CAM to orthodox medicine for this ailment" (31.4%), followed by "Orthodox medicine had failed to work for me" (20.1%), and "Long distance to the healthcare facility" (9.1%) ( Table 8). Other important motivations for CAM use cited by the respondents included lack of necessary drugs at the health center, unlikeable attitude of the health clinic personnel, and inability to afford treatment at the health center. VI. Utilization of CAM together with conventional medicine 55.80% of the CAM users reportedly utilize CAM and conventional medicine separately while 44.20% use CAM and conventional medicine together ( Figure 6).

VII. Potential CAM utilization in the future
The majority of the respondents (86.6%) plan to use CAM products and practices in their future pregnancies, 10% were not sure if they will use them in the future, and 2.7% stated that they don't intend to use them in their future pregnancies (Figure 7).

DISCUSSION
The study set out to determine the prevalence of CAM utilization and associated factors among pregnant women attending ANC clinics. The study established that half of the pregnant women from Tongaren had used some CAM products at least once during past or present pregnancy. Compared to previous estimates in the nation, the 50.7% prevalence determined by this study is equal to the rate observed among pregnant women in Thika (Njoroge & Kibunga, 2007). However, it is higher than the prevalence rate of 40.5% observed among pregnant women in Nairobi (Mothupi, 2014) and 42.5% recorded among pregnant women in Kiambu County (Githinji, 2015) and lower than the 68.9% registered in Gucha District among pregnant women (Ondicho et al., 2016) and 70% observed among pregnant women in Embu (Hillary, 2013

Plan to use CAM /TM products and practices when you are pregnant
imply that CAM is still heavily relied upon in most Kenyan societies to enhance, promote and restore human health. The differences in the prevalence of CAM utilization among pregnant women from various parts of Kenya could be attributed to the socio-cultural differences that exist between the various ethnic groups in the country. For instance, from this study, it was established that almost 80% of the respondents are from the Luhya ethnic tribe and therefore their CAM utilization patterns would be expected to be different from those of women from Embu, Meru or any other ethnic groups in Kenya since they all have varied customs and traditions which usually shape their perceptions and practice. According to (Owumi, 2012), while acknowledging the cultural possession and origin of CAM practices, adds that these therapies are ancient and have simply been passed down through generations in a certain culture or region therefore the society claims ownership of them. In a nutshell, Africa is a cradle of human race with different cultures and abundant life as well as diverse therapeutic practices in different regions (Mahomoodally, 2013), thus these practices are indigenous to the people who observe them.
Regionally, the high CAM utilization seen in this study was consistent with the relatively high CAM use prevalence reported by most studies conducted across Sub-Saharan Africa. Among pregnant women from neighboring countries, a prevalence rate of 45% was registered in DRC (Mbarambara et al., 2016) 52% has been recorded among 398 women surveyed in Zimbabwe (Mureyi et al., 2012) 55% was found among 235 women surveyed in South Africa (Hillary, 2013) 55% has been observed among 400 women surveyed in Tanzania and 80% was noted among 258 pregnant women surveyed in Ethiopia. This study complements these previous pieces of literature in demonstrating that CAM use has been increasing in Kenya and the developing world. According to the WHO (2019), above 80% of the population in the developing world rely on CAM use for their health improvement and disease treatments. However, it's not easy to state if these disparities in prevalence between surveys from various nations reflect real variances in CAM use or are due to differences in study design, characteristics of research sites, data collection, and sampling methodologies (James et al., 2018).
This study established that preference of CAM over conventional medicine for certain illnesses and the perceived lack of response to conventional medicine were the two leading reasons for the high CAM use prevalence among pregnant women from Tongaren Sub-County. These findings are similar to the conclusions of previous studies. For instance, (Yekta et al., 2007) study had reported that non-response to modern medicine was one of the major reasons for CAM utilization among pregnant Iranian women. In Uganda, (Langlois-Klassen et al., 2007) found that most AID's patients use CAM because modern medicine had failed to work for them. In Iraq, about 28% of pregnant women reported using CAM because of its effectiveness for certain illnesses (Hwang, Kim, et al., 2016).
From the foregoing, it is evident that the perceived efficacy of CAM modalities is the major driver of CAM utilization among pregnant women in Tongaren as is the case in other places around the world. Besides efficacy, this study reported that long distance to the healthcare facility is also responsible for CAM utilization. This is plausible because the analysis in this study had revealed that half of the pregnant women in Tongaren were living between 5-10 km away from the nearest health facility while the other half were living less than 5km away from the nearest health center. The issue of long-distance is compounded by the lack of enough road networks in this study area, which therefore limits easy access to health facilities since pregnant women in most cases are expected to walk the whole distance to reach the hospitals. Trekking to the hospital for long distances is extremely challenging and unsafe to these women considering their fragile conditions; hence they more often resort to CAM use, which is a more readily available alternative treatment method.
The inaccessibility challenge posed by the distance to be traveled to the healthcare facilities is not unique to Tongaren because it has also been reported to be one of the major reasons for CAM use among pregnant women in Nairobi (Mothupi, 2014) which is a more urbanized area. In Ethiopia, it was established that 80% of pregnant women use traditional medicine because of the difficulties they face in accessing health centers. It can therefore be concluded that the distance to be traveled to the healthcare facility is one of the major drivers of CAM use in the country.
In determining factors that affect CAM use, this research established found several predictors of CAM. Marital status was significantly found to be associated with CAM use. This is contrary to the earlier studies (Addo, 2007;Nzuki, 2016), which both established that the marital status of women and their CAM use were not significantly associated. According to this study, pregnant women who are married or cohabited were more likely to use CAM products compared to those who are single, a finding not consistent with the findings of (Gyasi, 2014). This finding suggests that spouses have some influence on CAM utilization among the pregnant women in this study area. This was further demonstrated by the significant association observed between the educational attainment of spouses and CAM utilization among pregnant women from Tongaren Sub-County. However, this association was not congruent with the findings of Gyasi (2014) who found no association between the educational attainment of partners and the use of CAM.
This study observed that women whose spouses had informal education were more likely to use CAM. This could be due to fact that these spouses lack enough knowledge on the effectiveness of conventional medicine and would be inclined to advise their pregnant wives to use the CAM which they are knowledgeable of and trust its efficacy since it is embedded in their socio-cultural backgrounds. In line with previous studies (Frawley, 2015;Strouss et al., 2014) this study found a significant association between a woman's education status and the utilization of CAM. Moreover, it was revealed that pregnant women who had attained up to secondary education were more likely to use CAM modalities. This could be due to fact that exposure to only secondary education might not provide sufficient health literacy to these women on the efficiency of conventional medicine on pregnancy-related symptoms. Therefore, these pregnant women use CAM which they are profoundly knowledgeable of and trust its efficacy since it is naturally embedded in their socio-cultural backgrounds. This appears to be contrary to previous studies which had established those pregnant women who use CAM are more likely to be those having tertiary education especially university degrees (Bishop et al., 2011;Steel et al., 2014;Torri & Hornosty, 2017). It is, however, consistent with Torri and Hornosty's (2005) study which stated that pregnant women with less education were more likely to use CAM products during pregnancy because most CAM users in this study had secondary education and below.
The significant association between a pregnant woman's occupational (employment) status and CAM use seen in this study was similar to the findings of previous research (Mureyi et al., 2012;Nzuki, 2016). Being self-employed was predictive of a high likelihood of CAM use among pregnant women. This was consistent with the Australian study by Frawley (2015) which found that pregnant women who utilize CAM were unlikely to be employed. It was, however, contrary to the findings of Bishop et al. (2011b) who had reported that an increased likelihood of CAM use was strongly associated with being employed.
In this study, household monthly income was significantly associated with CAM use, and this agrees with other previous studies (Biswas et al., 2020). Since being self-employed was also associated with CAM use, it was perfectly logical for higher household income to have an association with increased CAM use too because selfemployment would usually translate to better income in the household. This finding was contrary to the Addo (2007) study which stated that increased CAM use was associated with respondents coming from lower socioeconomic statuses. Moreover, it was contrary to an earlier analysis by Frawley (2015) which observed that women who were unemployed (meaning they were neither self-employed nor in any formal employment) and earn low income were more likely to use CAM.
Of the 172 CAM users in this study, half of them indicated using CAM together with modern medicine while another half stated that they use CAM separately. Most CAM users indicated that they use CAM products for certain illnesses that are not "treatable" with modern medicine, which implies that if a pregnant woman is having various symptoms at the same time, she would use CAM products for some symptoms and modern medicine for other symptoms. For, instance, when a woman is having skin problems and frequent migraines at the same time during the pregnancy, she will possibly use herbal medicine to eliminate the skin problems and modern drugs for curing headaches. Pregnant women probably find it much more convenient to be able to utilize both types of medicine to combat the risks of certain health problems during pregnancy.
The use of CAM together with conventional medicine could be driven by the widespread perception among CAM users that it would be safe if the drugs were for separate conditions. This perception has been found among CAM users in two regions in Ghana (Gyasi, 2014). Among communities, pharmacological pluralism is fast becoming a norm. Therefore, there's a need for these communities to be sufficiently informed of the hazards and issues that can arise when using CAM in conjunction with conventional medications. The other half of CAM users, who reportedly use CAM separately from conventional medicine, expressed concern about using CAM in conjunction with conventional ones for certain ailments. This suggests that they are concerned that CAM treatments may interact with certain conventional medications, causing undesirable reactions which could be fatal. Gyasi (2014) asserts that, since CAM and conventional medicine arose from varying theoretical assumptions and methodological perspectives, issues are highly likely when the two components are used at the same time.
An overwhelming majority (86.6%) of the 340 pregnant women surveyed in this study indicated that they plan to use CAM products and practices in their future pregnancies because they perceive them to be beneficial to the mother and the fetus. This implies that many women from Tongaren feel that complementary and alternative medicine (CAM) is just as safe as modern medicine in treating pregnancy-related ailments. Moreover, CAM use is probably perceived by this population as safer or effective than modern medicine in treating pregnancy-related illnesses. This notion had previously been established among pregnant Hispanic women in the US by Bercaw et al. (2010) (Bercaw et al., 2010). Research has proven that utilization of complementary and alternative medicine (CAM) during pregnancy seems to be mainly mediated, at least to some extent, by a quest for a non-toxic and efficacious natural treatment (Holst et al., 2009). Among, most CAM users, there is some comfort in knowing that herbal therapy has been used for hundreds of years and has not been "interfered with" (Torri & Hornosty, 2017).

CONCLUSION
Due to the widespread trust in the safety and efficacy of CAM, the utilization of CAM during pregnancy is highly common among women in Tongaren Sub-County, and therefore CAM seems to form the backbone of rural health care in Kenya. The rural women use CAM during pregnancy to complement conventional medicine rather than as an alternative, and this usage is highly influenced by their marital status, educational level, and employment status. Given the high prevalence of CAM use observed in this study, efforts should be directed towards validating the safety and efficacy of CAM modalities using accepted scientific principles. Moreover, since education was generally correlated with health seeking behaviour of pregnant women in Bungoma, more education programs such as peer group health education and community-based lectures should therefore be implemented to strengthen their health literacy which would be essential in guiding their safe utilization of CAM. Finally, the Kenya's Ministry of Health should collaborate with relevant stakeholders to develop a framework for strengthening recognition of CAM modalities as legitimate forms of healthcare and integrating them into the country's modern healthcare system.