Determinants of the uptake of intermittent preventive treatment of malaria in pregnancy with sulphadoxine pyrimethamine in Sabatia Sub County, Western Kenya

Background Annually, 125.2 million pregnant women worldwide risk contracting malaria, including 30.3 million and 1.5 million in Sub-Saharan Africa and Kenya respectively. At least three doses of sulphadoxine pyrimethamine for intermittent preventive treatment of malaria in pregnancy (IPTp-SP) is recommended for optimal benefit. Kenya recorded low IPTp-SP optimal uptake in 2015. This study investigated the prevalence of and factors influencing IPTp-SP optimal uptake in Sabatia Sub County, Western Kenya. Methods A cross-sectional study was conducted in Sabatia Sub County from April to October 2020. Using a validated semi-structured questionnaire, data were obtained from 372 randomly sampled post-delivery women aged 15–49 years with live birth within one year preceding the study. Women on cotrimoxazole prophylaxis during pregnancy were excluded. Pearson Chi-square and Fisher’s Exact test were measures of association used. Binary logistic regression analysed predictors of optimal IPTp-SP uptake. Results Optimal IPTp-SP uptake was 79.6%, 95% CI 75.5%–83.7%. Predictors of IPTp-SP optimization were gestational age at first antenatal care (ANC) visit (P = 0.04), frequency of ANC visits (P < 0.001), maternal knowledge of IPTp-SP benefits (P < 0.001), maternal knowledge of optimal sulphadoxine pyrimethamine (SP) dose (P = 0.03) and SP administration at ANC clinic (P = 0.03). Late ANC initiators were less likely to receive optimal IPTp-SP (aOR = 0.4, 95% CI 0.2–0.9). Odds of optimizing IPTp-SP increased among women with ≥ 4 ANC visits (aOR = 16.7, 95% CI 7.9–35.3), good knowledge of IPTp-SP benefits (aOR = 2.4, 95% CI 1.3–4.5) and good knowledge of optimal SP dose (aOR = 1.9, 95% CI 1.1–3.4). Women who never missed being administered SP were highly likely to receive optimal IPTp-SP (aOR = 2.9, 95% CI 1.1–7.2) Conclusions This study has found high IPTp-SP optimal uptake in the study area. Efforts should be directed towards early and more frequent ANC visits. Intensive and targeted health education is required. It’s fundamental to adequately stock and consistently administer SP. Future studies considering larger samples and health workers’ perspectives of the health system delivery factors are recommended. Graphical abstract Supplementary Information The online version contains supplementary material available at 10.1186/s40249-021-00887-4.

In malaria endemic areas, MiP risk is 50% higher among pregnant women [2] due to pregnancy induced lowered immunity [3]. MiP causes anaemia, which in turn raises maternal and neonatal mortality risk [2]. Other negative consequences of MiP include haemorrhage, intrauterine growth retardation, still birth, preterm delivery, placental malaria, miscarriage, low birth weight (LBW) and congenital infections [2,4].
Globally, 125.2 million pregnant women risk contracting malaria yearly, including 30.3 million in stable malaria transmission zones of Sub-Saharan Africa (SSA) [5]. Approximately 40% of all pregnancies in SSA would experience Plasmodium falciparum placental infection in the absence of MiP preventive interventions causing 900 000 LBW deliveries annually [6]. In Africa, about 10 000 women and 75 000 to 200 000 infants die yearly due to MiP while LBW from P. falciparum infections during pregnancy causes nearly 100 000 neonatal deaths [2]. In Kenya, about 1.5 million pregnancies occur yearly, with 44% in moderate to intense malaria endemic areas [3]. The country's MiP prevalence was 4918 cases per 100 000 persons in 2018 [7]. Intermittent preventive treatment of malaria in pregnancy (IPTp) with sulphadoxine pyrimethamine (SP) is recommended by World Health Organization (WHO) for controlling MiP [8].
The WHO recommends provision of IPTp-SP to all pregnant women in moderate to high malaria transmission areas starting earliest in the second trimester at each scheduled antenatal care (ANC) visit until delivery, observing at least one month interval between the SP doses [8]. Kenya's Ministry of Health recommends at least three IPTp-SP doses (IPTp-SP3+) to pregnant women in malaria endemic areas [9]. The drug is safe and can be taken with or without food. However, WHO does not recommend SP to Human Immunodeficiency Virus (HIV) infected pregnant women on cotrimoxazole prophylaxis due to the drugs' redundant mechanisms of action and synergistic worsening of adverse drug reaction [8]. SP remains cost effective and efficacious for IPTp [10,11]. IPTp-SP has been linked to reduced MiP cases, lower placental malaria risk and reduced neonatal deaths [12]. The intervention was associated with increases in maternal haemoglobin and birth weight even in areas that experienced SP resistance in SSA [13].
Socio demographic and obstetric characteristics have been found to influence the uptake of IPTp-SP by pregnant women [16][17][18]22]. Knowledge related factors such as maternal knowledge of MiP and IPTp-SP have also been demonstrated to predict IPTp-SP optimal uptake [23,24]. In SSA, poor IPTp-SP coverage levels and many missed opportunities for IPTp-SP delivery have been attributed to health care system inadequacies [25][26][27]. Despite this, determinants of IPTp-SP3+ uptake in Sabatia Sub County were unclear and unknown. This study sought to investigate the prevalence of and factors influencing IPTp-SP3+ uptake in the Sub County. Understanding the prevalence and determinants of IPTp-SP3+ uptake is core for making decisions and setting priorities towards improved access to and optimization of the intervention.

Study site and study design
The study site was Sabatia Sub County, Vihiga County in Western Kenya. It is a malaria endemic zone. In 2018, its estimated pregnant women's population was 5117 [7]. It's mainly rural and most residents are in agriculture and rural development sector. A household-based cross-sectional study using a quantitative approach was conducted from April to October 2020. Cross-sectional study design was used due to its relevance in estimating prevalence and measuring both outcome and exposure variables simultaneously.

Target and study population
The target population was pregnant women aged 15-49 years. However, the study population included post-delivery women aged 15-49 years with live birth within one year preceding the study, were residents of Sabatia Sub County during their last pregnancy for ≥ 9 months before the study and consented to participate. The study measured IPTp-SP uptake at pregnancy end, hence the use of post-delivery women. HIV-infected women on cotrimoxazole prophylaxis during their last pregnancy and those with mental disorders were excluded.

Study variables
The dependent variable was number of IPTp-SP doses received. Though a numerical variable, a dichotomy was created to have sub-optimal IPTp-SP uptake (≤ 2 doses) and optimal IPTp-SP uptake (≥ 3 doses). Table 1 shows independent variables and their definitions.

Data collection
Face to face interviews of post-delivery women by trained research assistants were conducted using validated semi-structured questionnaires (Additional file 1). Data were collected on both the dependent and independent variables. ANC clinic attendance booklets were used to verify data on SP doses received, ANC visits made and gestational age at first ANC visit after asking study participants on the same. Each participant's responses and booklet information were compared to determine their accuracy. All the women studied had the clinic attendance booklets.

Sample size and sampling
Sample size was calculated using Cochran's (1977) formula for categorical data. Further, Cochran's (1977) correction formula was applied since the initial obtained sample size exceeded 5% of the population [28,29]. The sample size was determined based on optimal IPTp-SP prevalence of 37.5% as per the 2015 national survey [21], Health worker-client relationship On most occasions; 1. Whether a health worker greeted a pregnant woman before attending to her, 2. Whether enough time was spent attending to the woman, 3. Whether the woman was counselled/ educated on IPTp-SP benefits, 4. Whether the woman was given a chance to ask questions, 5. Whether the woman felt comfortable asking questions, 6. Whether her questions were answered satisfactorily, 7. Whether the health worker ever shouted at the woman. At least four positive responses to the first six questions and a negative response to the last question indicated good relationship, otherwise poor a 95% confidence level, a 5% precision and a 10% nonresponse rate. The final sample size was 372. Simple random sampling was applied. The process began by identifying, listing and informing all potential participants about the study. After seeking permission from relevant authorities, Community Health Volunteers' registers of pregnant women were used to develop a sampling frame of 3091. Next, a consecutive identification number from 1 to 3091 was assigned to each recruited participant. Microsoft Excel (Microsoft Corporation, Redmond, Washington, US) version 2016 "RANDBE-TWEEN()" function was used to randomly sample 372 participants from the sampling frame.

Statistical analysis of data
Descriptive analyses of continuous and categorical variables were done by calculating means and proportions respectively. Pearson Chi-square test and Fisher's Exact test where appropriate were used to compare differences in various predictors of interest. Binary logistic regression models were fitted to determine the relationship between IPTp-SP uptake and the predictors. Predictors with P < 0.15 from step-wise regression were included in the multivariable logistic regression model. Crude odds ratios (cOR) and adjusted odds ratios (aOR) were reported and all predictors with P < 0.05 were considered to be independently associated with IPTp-SP optimal uptake. Statistical analyses were performed using Stata version 14.0 (Stata Corp., College Station, TX).

Ethical considerations
The study was approved by Jaramogi Oginga Odinga University of Science and Technology Ethics Review Committee (Approval Number: 7/17/ERC/11/3/20-21) and the National Commission for Science, Technology and Innovation in Kenya (License Number: NACOSTI/P/20/5052). Informed consent was sought from all respondents using an approved consent form. Privacy and confidentiality of the study participants and all raw data were strictly observed. Sabatia Sub County Medical Officer of Health permitted the study (Additional file 1).

Characteristics of the respondents
A total of 372 participants were enrolled in the study.   Table 2).

Socio demographic, knowledge related and obstetric factors associated with IPTp-SP uptake
This study compared characteristics of women who received optimal IPTp-SP doses to those who received sub-optimal IPTp-SP. Table 3 shows the distribution of IPTp-SP uptake against the socio-demographic, obstetric and knowledge characteristics of women during pregnancy. Pearson Chi-square and Fisher's Exact text results show that IPTp-SP uptake was significantly associated with marital status (P = 0.04), gestation age at first ANC visit (P < 0.001), frequency of ANC visits (P < 0.001), knowledge of IPTp-SP benefits (P < 0.001) and knowledge of optimal SP dose (P = 0.01) ( Table 3).

Health service delivery factors influencing the uptake of optimal IPTp-SP
This study compared the health service characteristics with the outcome of IPTp-SP uptake.

Optimal uptake of IPTp-SP
IPTp-SP optimal uptake in the study area approaches the national target of 80% though still far from universality. This high optimal IPTp-SP uptake is consistent with studies in Ghana and Sierra Leone [19,20,23]. However, studies in Uganda, Tanzania and Malawi that investigated IPTp-SP optimization found low uptake levels [17,18,30]. The high IPTp-SP uptake prevalence could be attributable to sustained efforts by the Kenyan government and development partners towards eliminating MiP and continued investment in maternal health. For instance, maternal health care services including IPTp-SP are free. There is continuous capacity building of health workers on MiP management and prevention. Also, MiP prevention messaging is being implemented. Optimal IPTp-SP uptake in this study is higher compared to 37.5% reported in the 2015 national malaria survey [21]. Scale up of MiP preventive strategies could account for this. Besides, this may indicate geographical variations in the intervention uptake across the country. Different study methodologies could also explain the difference. For example, the survey sampled respondents from the entire country while this study drew its sample from one Sub County.

Socio demographic, knowledge related and obstetric determinants of optimal IPTp-SP uptake
In this study, late ANC initiators were less likely to receive optimal IPTp-SP. Similarly, in Tanzania first ANC booking before 17 weeks gestation age increased the a Statistically significant result at 5% significance level. ANC: Antenatal care; IPTp-SP: Intermittent preventive treatment of malaria in pregnancy with sulphadoxine pyrimethamine; SD: Standard Deviation; SP: Sulphadoxine pyrimethamine  odds of receiving IPTp-SP3+ [30]. Consistently, two studies in Zambia and Ghana reported ANC start date as a significant predictor of optimal IPT-SP uptake [24,31]. However, a different study in Tanzania found no relationship between ANC initiation and IPTp-SP dosage [32] though it used a smaller sample size of 138. In line with other studies in Uganda, Tanzania, Malawi, Sierra Leone and Ghana, this study found that ≥ 4 ANC visits predicts optimal IPTp-SP uptake [17-20, 23, 24, 30]. The WHO calls for integration of IPTp-SP with initiatives for promoting focused ANC services [8]. Therefore, early ANC initiators are likely to achieve adequate visits, maximizing their contacts with health workers hence increased health education and SP administration. Evidently, pregnant women should be urged to initiate ANC visits earliest in the first trimester and adhere to all scheduled visits. Apparently, adequate ANC visits do not necessarily guarantee optimal IPTp-SP uptake. One-third of women with ≥ 4 visits in this study received sub-optimal IPTp-SP, possibly due to health service delivery deficiencies such as drug shortage. IPTp-SP optimization could be realized when pregnant women are adequately and properly informed of the intervention. In Tanzania, Cameroon, Zambia and Ghana, maternal knowledge of IPTp-SP positively influenced the intervention's maximum uptake [30,31,33,34]. Similarly, this study found that women who understood IPTp-SP benefits and knew the recommended SP doses were more likely to receive optimal dosage. Good maternal knowledge of IPTp-SP empowers women to develop positive attitudes and perceptions towards adequate intervention uptake.
Marital status was significant before adjusting for other covariates. The findings from this study show as it does in Ghana that married women have higher odds of taking IPTp-SP3+ [35]. Though it did not consider IPTp-SP optimization, a different study in Kenya had similar findings [36]. Possibly, married women get financial and psychosocial support from their spouses towards ANC attendance. As evidenced by a study in Bungoma East district of Kenya, women who received support from their partners were more likely to receive > 1 SP dose [37]. It could also be that unmarried women get stigmatized for getting pregnant out of matrimony hence a need for their social protection.
Though no significant association, younger women, those with low education, the unemployed, rural residents and primiparous women had higher proportions of sub-optimal IPTp-SP uptake. As the unmarried, these are likely to be the 20.4% who received sub-optimal intervention. These women may lack adequate access to information and communication channels used for IPTp-SP promotion. Thus, mapping out these vulnerable women and using trained community health promoters to visit them, offer comprehensive MiP health a Statistically significant result at 5% significance level. ANC: Antenatal care; CI: Confidence interval; cOR: Crude odds ratio; IPTp-SP: Intermittent preventive treatment of malaria in pregnancy with sulphadoxine pyrimethamine; MiP: Malaria in pregnancy; SP: Sulphadoxine pyrimethamine

Health service delivery predictors of optimal IPTp-SP uptake
Health service delivery dynamics are critical to appropriate IPTp-SP uptake [26]. Previous studies have argued that sub-optimal IPTp-SP dosage was mainly due to health system gaps [25][26][27]. In this study, SP administration at ANC clinic determined IPTp-SP3+ uptake. Health workers administer SP to eligible pregnant women and offer health education during ANC attendance. Ever missing to administer SP can create a sense of mistrust among women concerning the drug's continued availability. Thus, subsequent ANC visits may be missed leading to suboptimal IPTp-SP uptake. Besides, late ANC initiators who qualify for only three SP doses may receive sub-optimal dosage if they miss any dose. Inconsistent SP administration could be attributed to erratic drug supply. In Ghana, SP shortage was found to be a barrier to realizing high IPTp-SP uptake [25]. Also, health workers' inadequacies to offer intensive IPTp-SP education due to high ANC clinic workload could contribute to the inconsistency.

Study limitations
This study employed a cross sectional study design whose weakness is the impossibility to infer causation. Only women with live birth were included, missing an opportunity to study those with still birth. Responses were selfreported hence the possibility of recall bias. However, this was minimized by verification of some responses using ANC clinic attendance booklets. It was also minimized by excluding respondents with live birth beyond one year before the study. Health service delivery factors were not