Seasonal variations and health risk assessment of microbial contaminations of groundwater in selected coastal communities of Ghana

Abstract

This study investigated seasonal variations in microbial contaminations of groundwater and associated health risks in four coastal communities (Essiama, Winneba, Accra, and Keta) in Ghana. Membrane filtration methods, sanitary risk inspection, and quantitative microbial risk assessment were employed, respectively, to (i) quantify bacteriological quality, (ii) identify risks to contamination, and (iii) assess health risks associated with Escherichia coli in groundwater. Results showed 70.00%, 53.33%, 70.37% and 90.00% of groundwater sources in Essiama, Winneba, Accra, and Keta, respectively, were at intermediate risk, whereas 3.33%, 40.00%, 14.81%, and 3.33%, respectively, were at high risk. Very high-risk levels of contamination were recorded only in Accra. The presence of animal wastes within a 10 m radius of groundwater collection point, bad drainage systems, collection of spilt water in apron area, the use of ropes and buckets when fetching groundwater, and absence of aprons and well covers put more than 60.00% of the groundwater points in two or more locations at risk of contaminations. Assessment of bacteriological quality of groundwater indicated that mean total coliforms and E. coli ranged, respectively, between 123.40-501.30 and 30.98–141.90 CFU/100 ml for the communities; the highest microbial counts for dry and wet seasons occurred in Winneba and Keta, respectively. Seasonal variations in E. coli counts in Winneba and Accra were significantly higher in the dry season than in the wet season; Essiama and Keta showed no significant seasonal variations. Exposure to E. coli O157:H7 through drinking groundwater ranged between 5 and 23 cells per day. Although exposure to E. coli O157:H7 through bathing was less than 1 cell per day in all communities, residents were exposed to one E. coli, at least, every 62, 141, 237, and 282 days in Winneba, Accra, Keta, and Essiama, respectively. The risk of infection and illness for all communities was 1 for drinking, whereas that for bathing ranged from 0.57 to 0.98. The estimated Disability-Adjusted Life Years (DALY) exceeded the WHO-acceptable DALY. These findings show that groundwater resources in the selected coastal communities were prone to microbial contaminations, and this may be a setback to Sustainable Development Goals 6. Implications of the findings are discussed.

1. Introduction

Groundwater is a crucial resource on Earth [1]; it provides over 97% of accessible freshwater [2] particularly in Sub-Saharan Africa [3] and half of the water needed for drinking worldwide [2]. The achievement of several of the United Nations Sustainable Development Goals (UN-SDGs) is closely linked to the provision of groundwater, both in adequate quantity and optimum quality. For example, improvement in access to clean water and its sustainable management is a mechanistic approach to increasing good health and well-being (SDG 3), and sustainable development in cities and communities (SDG 11). However, this critical resource is faced with various challenges such as pollution and increased anthropogenic pressure. Aquifers are faced with an increasing global threat of pollution [4].

Water-borne infections and unsafe water sources are major global issues as a result of groundwater pollution. Lack of safe drinking water leads to remarkable increases in contagious illnesses like cholera, typhoid, dysentery, hepatitis A, poliomyelitis, and diarrhoea [5]. Globally, about 1.2 million people died in 2017 due to the lack of safe drinking water [6], and more than a quarter of a million children under the age of 5 years die each year from diarrhoea [5]. The number of deaths is high in low-income countries, where sub-Saharan Africa records the highest disease burden from inadequate water, sanitation, and hygiene [7]. In 2016, diarrheal diseases caused over half a million deaths in sub-Saharan Africa, where polluted drinking water was a major risk factor [8]. The relevance of this problem underpins the fact that clean water and sanitation (SDG 6) is geared towards ensuring the availability and sustainable management of water and sanitation for all.

A wide range of pathogenic microbes is found in groundwater [9]. Bacteria, however, constitute a major group of microorganisms that cause waterborne disease outbreaks [10]. Indicators of faecal pollution are often used in the monitoring of the bacteriological quality of drinking water. Escherichia coli (E. coli) is an excellent example of an indicator of faecal pollution in drinking water. Quantitative Microbial Risk Assessment (QMRA) is an important tool for determining the risk posed by microbial contamination as it combines information regarding the nature, presence, movement, and fate of disease-causing microorganisms into a single assessment that allows for data-informed, sufficient, and comprehensive risk management [11].

Findings by [12] indicate that 60% of Ghana's surface waterbodies are contaminated, leading to increasing dependence on groundwater. Other studies have reported similar findings from other countries [13,14]. Increased cost of water treatment due to pollution in Ghana has caused several water treatment plants to shut down [12]. Additionally, many households have resorted to the use of wells and boreholes due to the (i) cost of water bills, (ii) inability of the Ghana Water Company Limited (GWCL) and the Community Water and Sanitation Agency (CWSA) to expand water supplies to new communities, and (iii) assumption that groundwater is clean and free from pollution. According to the 2021 Ghana Population Census, only 31.7% of Ghanaians rely on pipe-borne water as a source of drinking water [15]. Other sources of drinking water include water in plastic sachets, boreholes, tube wells, protected and unprotected wells, and groundwater [15]. Several studies have reported the pollution of groundwater in Ghana with faecal bacteria [[16], [17], [18], [19], [20], [21], [22]]. This notwithstanding, few studies have focused on aquifers in coastal areas where a quarter of Ghana's population live [[23], [24], [25]]. The high population in the coastal areas does not commensurate with the availability of sanitation facilities. Less than 26% of the population in each of the coastal regions have access to basic sanitation services; about 16, 17, 8 and 38% of the population in Western, Central, Greater Accra and Volta regions, respectively, practice open defecation [26]. Also, attention given to sanitation is majorly focused on solid wastes, while little focus is placed on liquid wastes, which are likely sources of faecal contaminations of groundwater [27]. The limited provision of sewerage services in Ghana has led to the dominance of onsite sanitation facilities (e.g., septic tanks, cesspools, etc.) that pose a threat to the quality of groundwater if not well constructed and maintained [27].

Ghana was unable to achieve the Millennium Development Goal on Sanitation [28] and progress towards the achievement of the related SDG is slow [29]. Sanitation and the quality of drinking water are so interwoven that they cannot be separated [30]. Cholera is endemic in Ghana especially in the coastal communities [31], in the Central, Eastern, and Greater Accra regions [32]. There is a dire need to protect the health of the teeming population of coastal dwellers in Ghana, but the long-term assessment of bacteriological groundwater quality that captures seasonal variations in the coastal communities of Ghana is scarce. This may be attributed to (i) lack of funding by governmental and non-governmental agencies for broader, comprehensive studies, (ii) few practicing experts in the field of water and sanitation, and (iii) lack of by-laws to support institutions like Community Water and Sanitation Agency (CWSA) to carry out assessment of self-supply water sources. Furthermore, studies that assessed the potential health risks associated with bacterial contaminations of groundwater in coastal communities of Ghana are rare. In a previous study, Lutterodt et al. [24] assessed the suitability of well water for drinking and showed persistence presence and contamination by bacteria in a coastal aquifer in Cape Coast, a prominent historic Metropolis in the Central Region of Ghana. In that study [24], only the dry season was considered and no empirical assessment of health risk was carried out. In other studies, a one-time sampling was carried out to assess the quality of groundwater; these studies also did not assess the health risks associated with groundwater contamination [23,25]. Literature search on quantitative microbial health risk assessment associated with groundwater sources in the coastal communities of Ghana revealed only work in some slums of Accra [33]. Thus, the scant information and dearth of knowledge in the area of water and sanitation, with a special focus on groundwater quality, suggested a need for a series of investigations to understand the current state and quality of underground water in the coastal communities in Ghana. To this end, this study, therefore, sought to assess (i) the bacteriological quality of groundwater sources in the coastal communities of Ghana, (ii) seasonal variations in microbial contamination, and (iii) the potential health risks associated with the use of groundwater. This study is significant because it fills the gap in the long-term assessment of microbial quality as well as quantitative microbial risk assessment of groundwater in the coastal communities of Ghana. The findings of this present study serve as a springboard for future studies concerning groundwater quality in Ghana and also highlight significantly the potential sources of bacterial contaminations and the consequential health risks.

2. Materials and methods

2.1. Study sites

The study was carried out in four selected coastal towns/settlements of Ghana, namely, Essiama, Winneba, Accra (Korle lagoon catchment area), and Keta (Fig. 1). The four coastal towns/settlements were selected from each of the coastal regions of Ghana because of the tendency of exposure of their groundwater to faecal contaminations within the coastal aquifers.

Fig. 1.

Map showing study areas along the coast of Ghana.

Credit: Charles A. Faseyi, PhD. Centre for Coastal Management, Africa Centre of Excellence in Coastal Resilience, University of Cape Coast, Cape Coast, Ghana.

Essiama is located in Ellembele District, in the Western Region of Ghana. The Community Water and Sanitation Agency (CWSA) of Ghana provides water for the town. However, several households in the town are yet to subscribe to the services provided by CWSA; they depend on water from wells and boreholes for domestic use.

The Korle lagoon catchment area is located in Accra, the capital of Ghana. It is home to some of Accra's notable slums including Agblobloshie, Old Fadama, and James Town, from which samples were collected. These settlements, like most typical slums, lack good sanitary facilities. Inhabitants use public bathrooms and toilets. Wastewater, including urine, is mostly drained into open drainage systems within the settlements. Solid waste and sewage disposal are also major sanitary problems in these settlements. Major sources of water for domestic use are piped water from GWCL, water in plastic sachets, wells, and boreholes. When there is a water shortage due to disruption of flow from GWCL, some members of these settlements purchase water from vendors who usually store water from the GWCL in large plastic storage tanks and retail it.

Winneba is located in the Central Region of Ghana. The major economic activities in the town are fishing and trading. The town is divided into New Winneba and Old Winneba. The New Winneba obtains its water supplies from the GWCL; there are almost no wells and boreholes in this part of the town. However, in Old Winneba town, groundwater remains an important source of water for drinking and domestic use.

Keta is located in the Volta region of Ghana and it is home to Keta Lagoon, the largest lagoon in Ghana. Fishing is the major economic activity in Keta. Several neighbouring communities around Keta are involved in farming. Generally, hand-dug wells are the major sources of water supply for most communities in the Keta basin.

2.2. Data collection

2.2.1. Sanitary risk assessment

A sanitary risk assessment was performed before sampling of growundwater to assess the exposure of wells and boreholes to possible microbial contamination. Methods detailed by [34,35] for sanitary risk assessment in hand-dug wells and boreholes were combined and adopted for this study (Table 1). The presence or absence of possible risk factors that could lead to the contamination of groundwater at each water point was documented. A total of nineteen (19) risk factors were used for the sanitary risk assessment: eleven (11) were used for both wells and boreholes; five (5) for wells only; three (3) for boreholes only (Table 1). “Yes” was checked when a risk factor was present; “No” was checked when a risk factor was absent. The “Risk score” for each water point was determined by the total number of factors present (Yes), which was then used to calculate the percentage risk score as follows:

Table 1.

Risk factors used in sanitary risk assessment.

	Risk factors	Risk affects
1	Pit latrine/septic tank soak away within a distance less than 10 m	Wells and boreholes
2	Nearest pit latrine is uphill (at higher ground than the well/borehole) within 30 m away	Wells and boreholes
3	Surface water uphill	Wells and boreholes
4	Waste dump site within a distance less than 10 m	Wells and boreholes
5	Wastewater drain within a distance less than 10 m	Wells and boreholes
6	Absence of apron/apron does not extend more than 1.5 m/apron or cement floor does not slope away from well	Wells and boreholes
7	Collection of spilt water in the apron area	Wells and boreholes
8	Faulty masonry	Wells and boreholes
9	Drainage channel is in bad condition (cracked, broken, blocked or unlined)	Wells and boreholes
10	Presence of trees around wells that can lead to well contamination	Wells and boreholes
11	Presence of animals or their waste 10 m radius	Wells and boreholes
12	Lack of well cover	Wells only
13	Use of bucket and rope	Wells only
14	Absence of wellhead or wellhead is less than 0.3 m high	Wells only
15	Lack of inner lining	Wells only
16	Abandoned materials/waste inside well	Wells only
17	Uncapped well within 10–15 m of a borehole (for boreholes)	Boreholes only
18	Unsanitary/worn-out seal of borehole pump	Boreholes only
19	Pump is operated by feet	Boreholes only

% Risk score =((Risk score)⁄(Total number of risk factors) × 100) 1

The risk of contamination was then categorized as Low (0–25%), Intermediate (26–50%), High (51–75%), and Very High (76–100%). The results on risk factors for wells and boreholes were pulled together in each community because few boreholes were sampled in most communities.

2.2.2. Sample collection

Sampling of wells and boreholes from the selected communities was carried out between April and June in 2022 for the rainy season and in December 2022 for the dry season. Sample bottles were prewashed before fieldwork, and rinsed several times with the sampled water from the water points. Boreholes were purged before collecting water samples, while hand-dug wells were sampled after the wells had been actively used by inhabitants of the communities to prevent sampling stagnant water [24]. Thirty (30) groundwater points were each sampled in Essiama, Winneba, and Keta, while 27 were sampled in Accra. The samples were collected into 500 ml sterile plastic containers as described by the American Public Health Association [36] and transported in an ice chest to the laboratory, where they were stored at 4 °C until they were analyzed.

2.3. Laboratory analyses

The membrane filtration method was used for the analysis of total coliform and E. coli as previously described by [24,37,38]. One hundred millilitres of water samples were filtered through 0.45 μm cellulose ester filter paper (diameter 47 mm), after which the filter paper was placed on a Chromocult agar plate and incubated at 37 °C between 21 and 24 h. All dark blue to purple-coloured colonies present on each plate were counted as E. coli count [38]. All salmon red-coloured colonies were counted and recorded as other coliforms present. An oxidase-based test strip was used to confirm the coliform bacteria. All analyses were in duplicate.

2.4. Quantitative microbial risk assessment

Quantitative microbial risk assessment (QMRA) was performed for E. coli O157:H7 using the following steps:

• a)Hazard identification

E. coli was chosen because it is a common water quality indicator used in assessing faecal contamination and a significant cause of water-borne disease, particularly in developing countries. Furthermore, E. coli is easy and more economical to analyze. The O157:H7 was chosen because it is the most important serotype of E. coli in terms of public health concerns.

• b)Exposure assessment

The concentration of E. coli O157:H7 was determined by multiplying the overall mean values of E. coli in a community by 8% (0.08), which represents the percentage of E. coli known to be pathogenic [39]. The exposure pathway considered were ingestion (drinking) and incidental ingestion during showering. Incidental ingestion pathway (showering) was considered in this study, because, in the Korle lagoon catchment area, groundwater is generally presumed to be polluted. Hence, groundwater in the Korle lagoon catchment area is mostly used for bathing and washing. As the groundwater in these towns is not treated before use, pathogen reduction measures were not considered.

Exposure or dose was calculated as:

$$D=V_{Consumed}\times C_{Ecoli\mathrm{O}157:\mathrm{H}7}$$ 2

Where D is the dose per day, V_consumed is the volume of water consumed per day, and C_{E. coli O157:H7} is the mean concentration of E. coli in a community multiplied by 0.08.

For direct ingestion, the volume of water drank directly (V_consumed) was based on WHO's recommended water intake of 2L per day [40]; for incidental ingestion during showering, V_consumed of 1.43 mL per day was used based on the 10 mL/week reported by Westrell [41].

• c)Dose-response assessment

In this study, the beta-Poisson model was used for the dose-response assessment. The equation for single exposure was used as shown below. Single exposure assumes that every ingested pathogen acts independently of other ingested pathogens and has the possibility of resulting in infection [42]. It is noted that this assumption may lead to an overestimation of the response as it does not take cognizance of the differences that are present between individual pathogens [43].

$$Pinfection,day=1-{\left(1+D/N_{50}\left(2^{\frac{1}{\alpha }}-1\right)\right)}^{-\alpha }3$$ [11]

Where, P_infection, _day is the possibility of an infection per day, D is the dose, N₅₀ is the number of pathogens that will infect 50% of the exposed population, i.e., the median infection dose.

• αand N₅₀ have the values 0.373 and 2.473, respectively [11,[44], [45], [46]].
• d)Risk characterization

The possibility of infection per year was calculated using the formula

$$P_{\mathrm{i}\mathrm{n}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n},\mathrm{y}\mathrm{e}\mathrm{a}\mathrm{r}}=1-{\left(1-P_{\mathrm{i}\mathrm{n}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n},\mathrm{d}\mathrm{a}\mathrm{y}}\right)}^{365}4$$ [11a]

The number of days in a year (365 days) was used as the exposure frequency because the exposure pathway considered are drinking and bathing. It was assumed that members of the population take their baths daily. This assumption is based on the fact that groundwater in the studied communities is readily accessible and free, in most cases. Irrespective of free access to groundwater, some individuals may be constrained by some economic and attitudinal factors such that they may not take their baths daily; our assumption may therefore lead to inconsequential overestimation of the number of infections per year. Although the Equation 4 may overestimate the number of infections per year, as it is based on the assumption that infections occur daily, and that individuals do not build immunity to infection, it gives a good estimate of the yearly likelihood of infection [47,48].

The probability of illness per year was determined using the formula

$$P_{illness,year}=P_{infection,year}\times P_{ill/\mathit{inf}}5$$ [47]

The probability of illness given an infection (P_ill/inf) with E. coli O157:H7 is 1 [45].

Disability Adjusted Life Years (DALY) were calculated as years lived with disability (YLD) and years of life lost due to death (YLL). YLD and YLL were computed as the product of incidence, severity, and duration of illness [47].

Durations for YLL were calculated by subtracting the age at death for an outcome from the life expectancy. Severity weighting, incidence, and duration of E. coli O157:H7 outcomes were obtained from Havelaar and Melse [49], except incidence values for death due to diarrhoea, and severity of end-stage renal disease, which were obtained from Howard et al. [50] and Canada Health [47], respectively. The calculation of years of life lost was based on the average life expectancy of 66.3 in Ghana [51], death from diarrhoea occurring at 12 months, and death from end-stage renal disease occurring at an average of 30 years based on a study in Ghana [52].

2.5. Data analysis

Statistical analysis was performed with SigmaPlot 14.0, Microsoft Excel, and Minitab Statistical Software. Risk levels were shown as percentages. Descriptive statistics were performed for total coliform and E. coli concentrations. Levene's test was used to check for equal variance. The Wilcoxon signed-rank test for paired data was used to compare the medians of the wet and dry season for Essiama, Accra, and Keta, while the Mann-Whitney test was used for the Winneba, because data from Winneba did not meet the equal variance assumption needed for the Wilcoxon signed-rank test. The overall means of the raw data were used to calculate the microbial risk.

3. Results

3.1. Sanitary risk assessment

Fig. 2 shows results of sanitary risk assessment in Essiama, Winneba, Korle Lagoon catchment area, and Keta. Sanitary risk assessment showed that 70.00% of well/boreholes in Essiama, 53.33% in Winneba, 70.37% in the Korle lagoon catchment area, and 90% in Keta were at intermediate risk. Groundwater points at high risk of contamination were 3.33%, 40.00%, 14.81%, and 3.33% in Essiama, Winneba, Korle Lagoon catchment area, and Keta, respectively. Very high-risk levels of groundwater contamination were observed in the Korle lagoon catchment area of Accra (3.70% of the wells and boreholes) as shown in Fig. 2.

Fig. 2.

Spatial variations in risk levels.

Sanitary risk assessment results (Fig. 3) also showed that common risk factors to which groundwater points in Essiama were predisposed included the presence of animals or their waste within a 10 m radius of groundwater points, lack of well cover, use of rope and buckets, the presence of trees around wells, and absence of aprons/apron not more than 1.5 m/apron not sloping away from well/borehole as they affected, respectively, 86.67%, 73.33%, 70.00%, 53.33%, and 73.33% of the groundwater points. In Winneba, similar risk factors were observed in addition to the presence of abandoned materials in wells, wastewater drain within a distance of less than 10 m, bad drainage conditions, and collection of spilt water in the apron area, affecting 60–100% of the sampled groundwater points. An exception, however, was the presence of trees around wells. The presence of animal wastes within a 10 m radius of groundwater affected 85.19% of water points, whereas collection of spilt water in the apron area, drainage in bad conditions, and wastewater drain within a distance less than 10 m affected 74.07%, 66.67%, and 62.96%, of the water points, respectively, in the Korle lagoon catchment area. In Keta, the use of ropes and buckets in fetching water, lack of well cover, apron-related risks (i.e., absence of aprons, aprons not more than 1.5 m, apron not sloping away from well/borehole), presence of animal wastes around wells and presence of trees around wells were the major risk factors that affected, respectively, 100%, 96.67%, 96.67%, 90% and 53.33% of the wells.

Fig. 3.

Spatial variation in risk factors.

3.2. Microbial contaminations of groundwater and its seasonal variability

Mean E. coli counts for Essiama, Winneba, Korle Lagoon catchment area of Accra, and Keta were 30.98 ± 6.34, 141.90 ± 29.10, 62.00 ± 26.90, and 36.92 ± 5.85 CFU/100 ml, respectively, for the entire sampling period (Table 2). The mean total coliform count was 123.40 ± 17.50, 501.30 ± 64.10, 245.40 ± 55.10, and 171.70 ± 18.20 in Essiama, Winneba, Korle Lagoon catchment area of Accra, and Keta, respectively. The mean E. coli and total coliform counts for all the study locations far exceeded the WHO standard for drinking water throughout the sampling period.

Table 2.

Summary statistics for E. coli and total coliform counts between April and December 2022.

Location	Variable	Mean	SE Mean	StDev	Minimum	Median	Maximum	Range
Essiama	E. coli	30.98	6.34	49.10	0.00	13.00	285.00	285.00
Winneba	E. coli	141.90	29.10	225.10	0.00	34.00	930.00	930.00
Accra	E. coli	62.00	26.90	197.50	0.00	9.00	1400.00	1400.00
Keta	E. coli	36.92	5.85	45.31	0.00	22.00	265.00	265.00
Essiama	Total coliform	123.40	17.50	135.50	0.00	78.80	595.00	595.00
Winneba	Total coliform	501.30	64.10	496.90	10.00	327.50	1972.50	1962.50
Accra	Total coliform	245.40	55.10	405.20	0.00	133.80	2400.00	2400.00
Keta	Total coliform	171.70	18.20	141.20	10.00	125.30	725.00	715.00

Results on seasonal variations (Table 3) showed that, during the wet season, Keta had the highest E. coli and total coliform counts (41.80 CFU/100 ml and 199.00 CFU/100 ml, respectively). For the dry season, Winneba had the highest E. coli counts and total coliform counts (243.80 and 860.80 CFU/ml, respectively). Wilcoxon signed-rank test and Mann-Whitney test results on seasonal variations in E. coli showed that the median E. coli counts for the dry season were significantly higher than that of the wet season in Winneba (p = 0.001) and Accra (p = 0.040). E. coli counts in Essiama and Keta showed no significant differences (p = 0.145 and 0.495), respectively. For total coliform, counts were also significantly higher during the dry seasons in Winneba (0.000) and Accra (0.045), except in Essiama, where a significant decrease was seen during the dry season (0.018). Total coliform in Keta showed no significant differences (p = 0.188).

Table 3.

Seasonal variations in E. coli and total coliform counts.

Location	Season	Mean	SE Mean	StDev	Min.	Median	Max.	Range	p-value
E. coli
Essiama	Wet	37.67	8.40	46.01	0.00	18.75	212.50	212.50	0.145
	Dry	23.83	9.47	51.85	0.00	8.75	285.00	285.00
Winneba	Wet	39.60	13.20	72.40	0.00	21.30	395.00	395.00	0.001
	Dry	243.80	50.40	276.20	0.00	108.80	930.00	930.00
Accra	Wet	23.98	8.08	41.99	0.00	5.00	127.50	127.50	0.040
	Dry	99.80	52.60	273.50	2.50	16.00	1400.00	1397.50
Keta	Wet	41.80	10.40	57.20	0.00	20.00	265.00	265.00	0.495
	Dry	31.53	5.30	29.03	2.50	22.00	102.50	100.00
Total coliform
Essiama	Wet	150.70	21.40	117.30	2.50	127.50	372.50	370.00	0.018
	Dry	96.10	27.10	148.60	0.00	35.00	595.00	595.00
Winneba	Wet	141.80	23.10	126.60	10.00	122.50	600.00	590.00	0.000
	Dry	860.80	85.40	467.90	220.00	781.30	1972.50	1752.50
Accra	Wet	181.70	63.50	330.10	0.00	90.00	1727.50	1727.50	0.045
	Dry	309.20	89.70	466.20	10.00	178.00	2400.00	2390.00
Keta	Wet	199.00	30.80	168.50	17.50	147.50	725.00	707.50	0.188
	Dry	144.50	18.80	103.10	10.00	123.00	392.50	382.50

3.3. Health risk assessment

Quantitative microbial risk assessment results showed that exposure to E. coli O157:H7 via drinking water from groundwater sources in the study locations ranged from 5 to 23 E. coli O157:H7 per day, with the highest exposure occurring at Winneba (Fig. 4, Fig. 5). Exposure to E. coli from the use of groundwater sources for bathing was less than 1 E. coli per day in all study locations. However, water users were exposed to at least one E. coli O157:H7 every 62, 141, 237, and 282 days in the Winneba, Korle lagoon catchment area, Keta, and Essiama, respectively, from showering (Fig. 5).

Fig. 4.

Spatial variation in exposure to E. coli from drinking.

Fig. 5.

Spatial variation in exposure to E. coli from incidental ingestion while showering.

Results for the probability of infection per day, probability of infection per year, and probability of illness per year are shown in Fig. 6, Fig. 7, where the respective probability of infection per day for drinking groundwater from the study locations is 0.57, 0.75, 0.66, and 0.60 in Essiama, Winneba, Korle lagoon catchment area, and Keta. On the one hand, the probability of infection per year and illness per year for drinking are 1 in all study locations, whereas, on the other hand, the probability of infection per day for showering with groundwater in the study locations ranged from 0.0023 to 0.0105, with Winneba having the highest risk. The probability of infection/year and illness/year for bathing with groundwater ranged from 0.57 to 0.98, and Winneba also had the highest risk.

Fig. 6.

Spatial variations in health risks associated with E. coli from drinking water.

Fig. 7.

Spatial variations in health risks associated with E. coli from bathing.

Table 4 presents the results for the Disability-Adjusted Life Years (DALY). The DALY obtained for this study was 0.49938. This value is far higher than the WHO-acceptable DALY of 10⁻⁵-10⁻⁴.

Table 4.

Incidence, severity, duration, and estimated DALY.

Years of Life Lived with Disability
E. coli O157:H7 outcomes	Incidence	Severity	Duration	DALY
Watery Diarrhoea	0.53000	0.067	3.4 days (0.009 years)	0.00032
Bloody diarrhoea	0.47000	0.390	5.6 days (0.015 years)	0.00275
Hemolytic uremic syndrome	0.01000	0.930	21.0 days (0.057years)	0.00053
End stage renal disease	0.00118	0.950	9.350 years	0.01048
Years of Life Lost
Death due to diarrhoea	0.00700	1	65.3 years	0.45710
Death due to hemolytic uremic syndrome	0.00104	1	26.2 years	0.02725
Death due to end stage renal disease	0.00003	1	36.3 years	0.00109
DALY				0.49952

4. Discussion

4.1. Sanitary risk assessment

Results from the study showed that most of the wells were at intermediate risk levels of microbial contaminations, which raise some concerns for the safe use of groundwater in the coastal communities of Ghana. Moreso, the results showed 40% of the groundwater points in Winneba were at a high-risk level, while 3.7% of groundwater points in the Korle Lagoon catchment area were at a very high-risk level of contamination. These records in the present study were higher than other findings [53], where it was reported that 15.9% of groundwater points in the largest region of Ethiopia were at a high-risk level and 1.5% were at a very high-risk level. The high-risk levels of contaminations in Winneba and the Korle Lagoon catchment area reported in this study may be due to sanitary conditions in the communities. Amongst the studied coastal communities, Winneba and the Korle Lagoon catchment area were the most populated with the least sanitary conditions. Most houses in these communities lacked toilet facilities. Also, wells were sited close to drainages. For example, 60% of wells were sited close to drainages in Winneba, as 62.96% of wells were sited close to drainages in Accra. Curiously, 66.67% of drainages in both Winneba and Accra were faulty and most of them contain stagnant wastewater with solid wastes; improper solid waste disposal in these communities is rampant. As a result, residents in these communities were at risk of consuming contaminated water.

Fig. 3 shows that activities within the study area posed some threats to groundwater in the studied communities. For example, the use of ropes and buckets in fetching water, a long-aged practice, was one of the common risk factors in the studied communities. Findings reported [54] confirm our findings, where the use of ropes and buckets was a major factor that led to the contamination of wells in some cities in Indonesia. This practice of using ropes and buckets to fetch water from wells tends to contaminate well water, especially when many people within a community handle the ropes and buckets as they draw water, and, in the process, contaminate the well water. In other circumstances, the ropes and buckets are left on the ground around the wells and are contaminated. Interestingly, another study [55] reported that the use of a rope pump instead of a rope and bucket greatly decreased the geometric mean of E. coli contamination by 64.10%.

It was observed in this present study that the presence of animal wastes within a 10 m radius of groundwater points was another common risk factor prevalent in all the communities sampled for this work. Thus, there is a need to protect groundwater sources from stray animals in the communities. This is because, E. coli, found naturally in the intestines of warm-blooded animals is spread through their stools. Other common risk factors such as (i) the absence of aprons (impermeable concrete floors built around wells and boreholes to prevent spilt water from seeping into the ground), (ii) aprons not sloping away from water points, and (iii) the absence of well covers, were related to poor design and construction of wells and boreholes. In Adamawa, Cameroun, microbial contaminations of groundwater were attributed to poor design and construction of aprons and the absence of well covers [35]. This notwithstanding, groundwater in the communities can be protected by strict observation and adherence to standards for the construction of wells and boreholes, as well as properly training artisanal masons, most of whom construct these wells, but are not well-versed in the standards of construction of wells, to be able to construct wells that are befitting of set standards.

4.2. Microbial contaminations of groundwater and its seasonal variations

Results also showed that mean E. coli and total coliform counts exceeded the WHO standards [56]. Dzodzomenyo et al. (2022), Elisante and Muzuka (2016), and Pathak et al. (2011) Similar findings from studies conducted in wells and boreholes were reported in Ghana, Tanzania, and Nepal, respectively, [[57], [58], [59]]. Results in the present study are in agreement with the findings of Close et al. [60], from which the mean E. coli of 40 MPN/100 ml and mean total coliform of 757 MPN/100 ml were reported in wells in New Zealand. The high E. coli counts reported in this study may be attributed to poor sanitary conditions around wells in the communities.

Results of this study also showed great variability in the seasonal variations of E. coli and total coliform counts across the four study locations, where no clear pattern was established. For example, while a significant (p˂0.05) increase was seen in E. coli and total coliform count at Winneba and Accra during the dry seasons, a significant (p˂0.05) decrease was observed for the total coliform count in Essiama, no significant change was reported for Keta. Such varying patterns have been reported in other studies, where no seasonal variations in faecal and total coliform of groundwater in Iran were observed [61], and significant increases during the wet season were reported from Ghana, South Africa and Kenya [59,62,63]. Conversely, higher aerobic plate counts were reported during the dry season [64]. Run-off and seepage are major causes of groundwater contaminations with faecal bacteria during the wet season and this may account for the high E. coli and total coliform counts reported for Keta during the wet season. Because Keta is characterized by loose sandy soils with larger pore sizes, its soils have a reduced ability to prevent the movement of microbes into aquifers [65,66]. Site-specific conditions may be the reason for an increase in E. coli and total coliform in Winneba and Accra in the dry season. During the sampling period, it was particularly noted that 60% of the wells in Winneba (Fig. 3) were sited close to gutters that were extremely dirty during the dry season. Similarly, the Korle Lagoon catchment area of Accra was characterized by poor sanitary conditions, which exposed groundwater in the community to contamination and its inhabitants to water-borne diseases. Site-specific factors have been reported to have more influence on groundwater quality than seasonal effects in Finland [67]. The concentration effects of microbes due to a decrease in water volumes may also be responsible for the increase in cell counts observed during the dry season [59,64].

4.3. Health risk assessment

High counts of E. coli pose a great risk to human health, as it is associated with gastrointestinal diseases. It is, therefore, not far-fetched that the daily risk of infection ranged from 0.57 to 0.75, and the annual probability of infection and probability of illness was 1 for drinking water in all the communities. The daily risk of infections observed in this study is higher than 0.325 in Afikpo, Nigeria [68], but the annual probability of infection remained the same for both studies. A much lower daily risk of infection (6.9 × 10⁻⁶) and annual risk of infection (2.5 × 10⁻³) has been reported in Ghana [18]. In another study [69], a mean daily risk of illness of 6.29 × 10⁻³ and an annual risk of illness of 0.655 for ETEC (a related serotype of E. coli O157:H7) was reported in five rural areas of Villapinzon, Columbia; just as this study, lower risk values were also reported in that study [69]. The high risk of infection obtained in this study suggests an urgent need to protect groundwater sources in the four coastal communities sampled for this study.

Curiously, on one hand, residents of Essiama, Winneba, and Keta believe strongly that the groundwater sources in their communities are safe for domestic use (drinking, cooking, bathing, washing, etc.). Because of this belief, some of them consume groundwater directly from these sources without treatments such as boiling before consumption. Interestingly, though, residents of the Korle lagoon catchment area, are fully aware that groundwater sources in the area are not safe for domestic consumption. Hence, some members of the community use the water for only bathing and washing, but rely on industrially-treated and packaged sachet water for drinking water. Irrespective of using the groundwater for bathing, the residents are still highly exposed to at least one E. coli cell every 141 days. This is higher than the acceptable risk level set by the USEPA for recreational waters [70].

DALY represents the total burden of a disease in terms of mortality and morbidity. The estimated DALY for E. coli O157:H7 in this study was 0.49952 per person per year, which exceeded the set DALY of 10⁻⁵-10⁻⁴ by the World Health Organization [71]. The burden of disease for E. coli O157:H7 alone was higher than the total acceptable burden of disease per person per year. This implies that the pathogen has serious health consequences for members of the communities. One DALY indicated the loss of one year of full health [72]. A DALY of 0.49952 per person per year implies that about 182 days of full health are lost per person annually due to infection by E. coli, indicating increased morbidity and/or mortality, as well as reduced quality of life for the inhabitants of the study communities. This finding is similar to [33], where a DALY of 0.4 for E. coli O157:H7 was reported. Uprety et al. [40] also observed DALYs higher than acceptable DALY in Nepal, particularly in the low-lying regions. There are several disease outcomes for E. coli ranging from mild diarrhoea, bloody diarrhoea, and hemolytic uremic syndrome, to end-stage renal disease. About 297,000 children under the age of five die each year from diarrhoea [5]. Data on diarrhoea due to E. coli O157:H7 are, however, scarce, as in most cases, diarrhoea patients are not screened for the exact causative agents. Children below five years have a higher risk of developing hemolytic uremic syndrome [71]. End-stage renal disease is also an important outcome of E. coli O157:H7 infection. A study in Ghana reported that 45% of renal disease death cases were due to end-stage renal disease [52].

5. Conclusion

This study assessed the bacteriological quality of groundwater sources in Essiama, Wiineba, Accra and Keta in Ghana, the seasonal variations in microbial contaminations, and the potential health risks associated with the use of these groundwater sources. The bacteriological quality assessment indicated that water from all groundwater points is not potable. The study showed that the presence of animal wastes within a 10 m radius of the water, bad drainages, absence of aprons, absence of well covers, and use of ropes and buckets are common risk factors that exposed groundwater to contamination in the studied locations. Seasonal variations in E. coli counts in Winneba and Accra were significantly higher in the dry season than in the wet season; Essiama and Keta showed no significant seasonal variations. The risk of infection and illness from drinking groundwater in the communities was 1; and the estimated DALY far exceeded the WHO-acceptable DALY, implying a high disease burden. Urgent action is needed to protect groundwater sources in these coastal communities. Public education on the importance of environmental sanitation must be intensified. Lastly, artisanal masons must be trained to construct wells that are befitting of set standards.

Funding

This research is part of the doctoral thesis of the first author which was funded by the World Bank through the Africa Centre of Excellence in Coastal Resilience (ACECoR), Centre for Coastal Management, University of Cape Coast, Ghana. [Project Credit No.: 6389- GH].

Author contribution statement

Emuobunuvie G Ayeta, Levi Yafetto, George Lutterodt, Joel F Ogbonna and Michael Miyittah - Conceived and designed the experiments; performed the experiments; analyzed and interpreted the data; and wrote the paper.

Data availability statement

Data included in article/supp. material/referenced in article.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.