Equity in access to safely managed sanitation and prevalence of diarrheal diseases in Bangladesh: a national and sub-national analysis

Bangladesh, with a population of nearly 149.8 million in 2011, is one of the most densely populated countries in the world (1015 people per sq km) [16]. The country has eight administrative divisions: south (Barishal), southeast (Chittagong), central (Dhaka), west (Khulna), mid-western corner (Rajshahi), central (Mymensingh), northwest (Rangpur), and east (Sylhet) [17]. These regions exhibit different geographical, demographic, environmental, gastronomic, and economic characteristics [18]. Other areas lag behind the socio-economic status of the southern part of Bangladesh [16, 19, 20]. Poverty, hunger, and other socio-economic indicators disproportionately impact people in the northwestern part of Bangladesh [16, 19, 20].

Data for this study were extracted from the BDHS 2017–2018, a nationally representative cross-sectional study. This survey was conducted by the National Institute of Population Research and Training of the Ministry of Health and Family Welfare conducted between October 2017 and March 2018. It was based on a two-stage stratified sampling of households. In the first stage of sampling, a fixed number of primary sampling units (PSUs) were selected, with probability proportional to their size. In the second stage, households were selected within each primary sampling unit using equal probability systematic sampling. The BDHS is a representative probability sample of men and women based on a two-stage cluster sampling of households, stratified by rural and urban areas and eight administrative regions of the country. In the first stage of sampling, 675 PSUs were selected with a probability proportional to the PSU’s size. On an average, each PSU contained 120 households. In the second stage of sampling, 20,250 households were selected using a systematic random sampling method. Interviews were successfully completed in 19,584 households, of which 19,457 were eligible. The overall response rate was 99.4%. Figure 1 illustrates sample selection framework. The detailed research protocol, methods, and structured questionnaires are available on the DHS website [20].

The outcome variables of this study were SMS and diarrheal diseases. According to the sanitation ladders of the WHO/UNICEF Joint Monitoring Program (JMP) [21, 22], SMS was defined as use of improved facilities that are not shared with other households and where excreta are safely disposed of in situ or removed and treated offsite. Moreover, an improved sanitation facility is designed to hygienically separate excreta from human contact and include the use of a flush toilet connected to piped sewer system, septic tank, ventilated improved pit latrine, pit latrine with a slab, and composting toilet. Conversely, an unimproved sanitation facility was defined as use of pit latrines without a slab or platform, hanging latrines or bucket latrines. According to the available BDHS data, improved sanitation facilities include flush or pour-flush to the piped sewer system, septic tank, or pit latrine, ventilated improved pit latrines, composting toilets, or pit latrines with slabs. Diarrhea was defined as the passage of three or more loose or liquid stools per day (or more frequent passage than is normal for the individual) [23] and the proportion of children under the age of five who had diarrhea in the 2 weeks preceding the survey were included.

We used individual, household, and community-level characteristics to assess the prevalence of diarrheal diseases. The variables were the household head’s age (≤ 24 years, 25–34 years, and ≥ 35 years), sex (male or female), and education (no education, primary, secondary education, and higher education), household’s wealth quintile (poorest, poorer, middle, rich, and richest), area of residence (urban or rural), and region of residence (Barisal, Chittagong, Dhaka, Khulna, Mymensingh, Rajshahi, Rangpur, and Sylhet).

The BDHS 2017–2018 was conducted by National Institute of Population Research and Training (NIPORT) and approved by the Institutional Review Boards (IRBs) at the ICF and the Bangladesh Medical Research Council (BMRC).

The background characteristics of the study population are presented in Table 1. In a total sample of 19,457 households, most individuals (77.5%) were ≥ 35 years old, and the majority of the household heads were male (84.6%). Of the total respondents, 40.7% fell into the poor category and most of them lived in rural areas (63.5%). Moreover, 27.8% had no formal education, and the Dhaka region accommodated the highest number of respondents (15.0%).

Table 2 presents the national and sub-national coverage of sanitation facilities in Bangladesh. The national coverage of SMS was 44.0% (95% CI 43.3–44.7), whereas it was 45.3% (95% CI 44.0–46.6) and 43.5% (95% CI 42.6–44.3) in urban and rural areas, respectively. Among the eight administrative divisions, the lowest coverage was observed in Mymensingh (32.9%) and the highest in the Chittagong division (54.1%). Access to basic sanitation facilities varied substantially across all divisions in Bangladesh, with a sanitation coverage rate of 47.8% (95% CI 44.9–50.8) in Barisal, 49.8% (95% CI 47.8–51.9) in Khulna, 40.7% (95% CI 38.8–42.5) in Rajshahi, 38.0% (95% CI 36.1–40.0) in Rangpur, and 44.8% (95% CI 41.9–47.7) in Sylhet. Figure 2 presents the residence-specific and quintile-specific coverage of safely managed sanitation at the national and sub-national levels in Bangladesh.

Greater inequalities in access to SMS facilities were observed at the national and sub-national levels (Table 2). The national level SII indicated that access to SMS in wealthy households was 60.8 percentage points higher than that of disadvantaged households. Additionally, greater wealth inequalities were observed at the sub-national level. At the sub-national level, inequality was higher in the Sylhet division and lower in the Dhaka division, which was 85.3 and 34.9 percentage points, respectively, greater in the richest household than in the disadvantaged households. Furthermore, greater inequality was observed in rural areas, which was 71.9 percentage points higher in the richest households than in the poorest households.

Table 3 presents the results of the multi-level logistic regression models for SMS. Higher age of the household heads was associated with higher odds of having access to improved sanitation facilities, by nearly 2.5-times in the age group ≥ 35 years (OR 2.36, 95% CI 1.82–3.07), followed by 25–34 years (OR 1.45, 95% CI 1.10–1.92). Gender of the household heads was associated with SMS, with female household heads having higher odds (OR 1.15, 95% CI 1.02–1.29) compared to male household heads. Similarly, household education was also an associated factor for improved sanitation facilities, with greater by nearly 2.5 times was observed with household heads having a higher education (OR 2.18, 95% CI 1.81–2.62), secondary education (OR 1.27, 95% CI 1.13–1.43), and primary education (OR 1.17, 95% CI 1.06–1.30) compared to those without education. The household wealth quintile displayed a significant association between having access to SMS, with higher odds in the richest quintile, which was more than 50-fold compared to the poorest quintile. The rural households had 2.55 times (95% CI 2.11–3.08) higher odds of having access to SMS compared to the urban households. Households from Mymensingh (OR 1.58, 95% CI 1.15–2.17), Sylhet (OR 1.91, 95% CI 1.35–2.70), Rajshahi (OR 2.06, 95% CI 1.48–2.87), Chittagong (OR 2.44, 95% CI 1.72–3.45), Khulna (OR 2.47, 95% CI 1.79–3.41), and Rangpur (OR 2.71, 95% CI 1.96–3.76) had higher odds of having access to SMS than those from Dhaka.

Table 4 presents the results of the Poisson regression model of prevalence of diarrheal diseases among 8433 children in Bangladesh. The diarrheal disease prevalence was higher in unimproved sanitation facilities, which was 4.87 (4.30–5.52) and 4.49 (3.87–5.21) in improved sanitation facilities. The prevalence of diarrheal disease was 11.0% lower in SMS users than in unimproved sanitation users. After adjustment for confounding variables, the prevalence of diarrheal disease was almost the same (14.0%).