This study accessed the efficiency and health risks of drinking water from different sources treated by filtration, boiling, chlorination, flocculation, and solar disinfection. The microbial quality of 45 treated water samples from boreholes, wells, and pipe-borne water was analyzed to determine treatment effectiveness and to quantify risk using quantitative microbial risk assessment. The effectiveness of each treatment method was a function of sampling sources (p < 0.05) and location (p < 0.10), chlorination and boiling being the most efficient methods (100%). Shiegella in well water samples treated by filtration and flocculation had the highest daily infection risk of 69.5 × 10−1 and 67.5 × 10−1 pppd. The annual risk of infection from Salmonella, Shigella, and Staphylococcus ranged from 7.8 × 10−1 to 1.00 pppy, exceeding the U.S. EPA annual infection benchmark (≤10−4 pppy). Salmonella, Shigella, and Staphylococcus had the highest risk of illness of 4.50 × 10−1, 3.30 × 10−1, and 9.80 × 10−1, respectively. All disease burden values exceeded the WHO disease burden benchmark (≤10−6 DALYs/pppy), with Staphylococcus and Salmonella contributing the highest disease burden of 4.71 × 10−2 and 2.13 × 10−2, DALYs/pppy. Therefore, boiling and chlorination are the best disinfection methods for the pathogens tested.

  • Filtration of drinking water with ceramic filters (the most common LHWT method in Bamenda, Cameroon) is the least effective method in removing bacterial contaminants such as Salmonella, Shigella and Staphylococcus.

  • Boiling and chlorinaIon are the most effective disinfection methods for the pathogens tested.

  • Filtration, flocculation, and SODIS did not reduce pathogenic concentraIons below risk levels.

  • Bacteria contaminants in poorly-treated drinking water exceeded the WHO disease burden benchmark.

Access to safe drinking water and basic sanitation by 2030 is the core of the United Nations 2015 Sustainable Development Goal (SDG-6). However, 2.6 billion people lack access to improved water sources, and in Africa, less than 50% of the population has improved sanitation, and about 70% of those lacking improved water sources do not drink treated mineral water (WHO/UNICEF 2018). Globally many countries are experiencing water stress due to pressure on water resources, and approximately 2.3 billion people are suffering from water-related diseases (Sauer et al. 2016). Annually, about 1.8 million people die because of poor sanitation and the consumption of contaminated drinking water, with children under 5 years old constituting 90% (Ahmed et al. 2020). Drinking water is a reservoir of water-borne pathogens that may pose devastating health risks after consumption (WHO/UNICEF 2018).
Figure 1

Map of Bamenda urban area in North West Cameroon.

Figure 1

Map of Bamenda urban area in North West Cameroon.

Close modal

However, simple local household water treatment (LHWT) and good storage methods can considerably reduce health risks by improving the microbial quality of drinking water. Globally, approximately 1.8 billion people use household water treatment to improve drinking water quality and thus prevent water-borne diseases (Rosa et al. 2016). In developing countries, bio-sand filters, ceramic water filters, boiling, solar disinfection (SODIS), and chemical treatment (chlorination) are the common LHWT methods (Rosa et al. 2014). Treated drinking water can be considered safe, but it may be a source of water-borne infections if not treated properly (Saturday 2016). The microbial quality of treated water can be used to determine the effectiveness of the LHWT method and the quantitative microbial risk assessment (QMRA) due to the consumption of drinking water treated by various LHWT methods.

QMRA is a tool for estimating the risks of an adverse effect of infection, illness, and /or death due to exposure to various doses of water-borne pathogens (NRC 1983). So far, the QMRA model has been recently used to assess the microbial risk of wastewater, surface water, drinking water, food, and recreational water (Howard et al. 2006; Enger et al. 2012; Membré et al. 2015; Abia et al. 2016; Kouamé et al. 2017; Ahmed et al. 2020). Recent studies on the QMRA of drinking water treated by various LHWT methods are, however, limited. For instance, Sari et al. (2019) used the QMRA model to compare the annual infection risk of drinking water after boiling, filtration, and water-refill methods in urban–slum areas. However, their study was limited only to annual infection risks of coliforms and Escherichia coli from one water source. The prevalence of water-borne diseases is continuously rising even though LHWT methods are used by many to improve drinking water quality from various sources, especially in countries experiencing water stress such as Cameroon. This shows the need for a detailed QMRA study to determine the health risks associated with the consumption of drinking water from different sources treated by various LHWT methods.

Currently, rapid population growth and high rates of urbanization in Bamenda have increased water stress and the dependency on alternative drinking water sources such as boreholes, springs, rivers, rainwater, and locally dug wells. These water sources in Bamenda have been shown to have high concentrations of pathogenic contaminants (Abendong et al. 2019; Anyang 2021; Mufur et al. 2021). Water-borne illnesses are also common though the inhabitants use ceramic filters, boiling, chlorination, or SODIS to treat drinking water (Nde 2017). These factors show the need for assessing the microbial quality and health risks of treated drinking water to recommend the best LHWT methods for the inhabitants in Bamenda. Therefore, the current study aimed to compare the bacteriological quality and QMRA of drinking water from different sources treated by various LHWT methods.

Study area

The Bamenda Town, found in the North West region of Cameroon, lies between latitude 5°43′ to 7°10′ north and longitude 9°35′ to 11°12′ east and at an altitudinal range of 1,200 to 1,865 m above sea level (Mufur et al. 2021). The town has a surface area of about 3,125 ha (AchoChi 1998) and is characterized by very steep slopes (Ndenecho 2004). The climate is equatorial, characterized by a rainy season that lasts for 7 months (from April to October) and a short dry season that lasts for 5 months (November to March) (Mufur et al. 2021). The average annual rainfall is 2,670 mm, and the mean annual temperature is 25°C. The population of the town has been on the rise for several decades now, making it the third-largest city in Cameroon (Wirba 2020). Presently the town has a population of over 600,000 inhabitants and is divided into three subdivisions (Bamenda I, II, and III) with city councils that manage local developments (Figure 1). The water consumption rate per household is mostly below 50 l/day, with very few households using up to 200 l/day. Pipe-borne water is the main source of water (65%), boreholes 18.2%, wells 8.3%, and streams/rivers 8.3% (Wirba 2020). Because of the inconsistency and unreliability of the pipe-borne water supply, coupled with water stress, the population uses these other sources as coping strategies (Chiaga 2019).

Sample collection

Water samples were obtained from field sites by stratified random sampling. The three subdivisions of Bamenda were used as a basis for stratification into Bamenda I (Up-station), Bamenda II (Old-town), and Bamenda III (Nkwen). Forty-five water samples were collected randomly from wells, taps, and boreholes into labeled 1-l sterile bottles. The water samples were put in a cooling flask and transferred to the Science Laboratory of the Catholic University of Cameroon for bacteriological analysis. The physical properties of water were measured on-site.

Household water treatment methods

The collected water samples were independently treated with the following methods to determine the effectiveness of each treatment method.

Filtration method: A popular and commonly used ceramic water filter was used in the filtration process, after which bacteriological tests were conducted. The pore size of the ceramic filter was between 0.3 and 3 μm. For each filtered sample, 0.25 l was used for the microbiological tests to identify bacteria that were not removed by filtration (Sari et al. 2019).

Boiling method: One liter of each water sample was boiled at 100°C in a stainless-steel pot for 30 min. After cooling, the samples were tested for microbes that were not destroyed by heat treatment (Sari et al. 2019).

Chlorination: Two drops (0.1 ml) of 10% sodium hypochlorite solution (commercial liquid laundry bleach solutions; ‘Eau de javel’), NaClO were put into 1 l water samples, swirled, covered, and allowed to stand for 30 min before drinking (Lantagne et al. 2006). As it dissolved in water, hypochlorous acid (HClO) was formed, which diffused through the bacterial cell wall destroying the membrane proteins. The samples were then tested for any residual pathogen.

Flocculation method: Five gram of alum, Al2(SO4)3‧14H2O, were put into 1 l water samples and vigorously mixed to dissolve thoroughly. It was then swirled (slow mixed) and allowed to stand for 30 min to form a coagulum, and sediment negatively charged the microorganisms to collide and clump together into larger, more easily removable clots or ‘flocs’.

SODIS: Labeled sterilized transparent plastic bottles were filled with 1 l of each water sample and kept on a zinc rooftop to expose it to full sunlight for 5–6 h. That is, from 9 am to 3 pm on a sunny day (CDC 2008). This exposure allows the bottled water to receive solar radiation intensity of 500 W/m2 (equivalent to 5 h of mid-latitude sunshine in summer). The water samples were later tested for the presence of pathogenic water-borne bacteria.

Isolation and identification of bacteria in drinking water before and after treatment

Nutrient agar was prepared for bacteria culture using the manufacturer's instructions. It was then sterilized in an autoclave at 121°C for 15 min and allowed to cool before use. The cool agar was then poured into Petri dishes for inoculation of the samples. The surface viable count of bacteria colonies was done using the drop plate method by Miles and Misra (Miles et al. 1938). One milliliter of the water sample was transferred to 9 ml of normal saline to make four serial dilutions. The sterilized nutrient agar sterilized in each Petri dish was allowed to dry. The plates were divided into four sectors and labeled according to the different dilutions. Twenty milliliter of each dilution from each sample, before treatment and after treatment, was inoculated onto the corresponding sector of the agar using the spread technique. This was done in triplicate and plates were then incubated at 35°C for 18–24 h in an inverted position. The resulting observed bacteria colonies were counted and recorded for each sector. The limit of detection was determined by serial dilution to be 0–1,000 CFU/ml. The number of colonies forming units per milliliter of the original sample was determined using the following formula:
(1)

The Gram stain test was used to identify the Gram-positive bacteria from the Gram-negative bacteria. After Gram staining, the Gram-positive cocci were tested for catalase and coagulase production. This test separated pathogenic Staphylococcus aureus from other non-pathogenic Staphylococci species. The sulfide indole motility (SIM) medium test was used to identify Gram-negative bacteria. SIM differentiates Gram-negative members of Enterobacteriaceae by their ability to reduce sulfur, produce indole, or move.

Procedure to determine the effectiveness of treatment

The bacterial reduction effectiveness of the treatment method was determined using the following formula;
(2)
where is the level of contamination before treatment (influent/untreated water) and is the level of contamination after treatment (effluent/treated water).

Quantitative microbial risk assessment

The QMRA was used to determine the annual infection or disease risk resulting from exposure to pathogens in treated water samples (NRC 1983; Haas et al. 2014). The identified bacteria in water samples treated by various local household methods were considered for the QMRA. The QMRA comprises four steps: hazard identification, exposure assessment, dose–response assessment, and risk characterization (Rose & Haas 1999; Medema & Ashbolt 2006) as described below.

Hazard identification

This step identifies the microbial agent and its associated adverse health effects on a given population. In the current study, Salmonella typhi, Shigella, and S. aureus were identified and used in estimating the exposure of the inhabitants to these pathogens after consuming locally treated water. These bacteria were selected because, in Bamenda, the prevalence of typhoid fever, dysentery, pneumonia, and skin infections is high, and they are the causative infectious agents (Fonyuy 2014; Biosengazeh et al. 2020).

Exposure assessment through locally treated drinking water

The exposure assessment estimated the dose of pathogens in treated water samples following ingestion in a day and a year. The exposure dose of pathogens was determined as follows (Kouamé et al. 2017):
(3)
where D is the exposure dose expressed in pathogens day−1, Mc is the mean concentrations of the targeted pathogens (CFU/ml), and Iv is the ingested volume of treated water; based on the assumption that children and adults drink 1,000 and 2,000 ml of treated water per day (U.S. EPA 2000; WHO 2011), these values are the default rate of daily drinking water intake for children and adults used by many countries and international organizations in exposure assessment.

Dose–response assessment

The probability of water-borne illnesses P(ill) from the intake of S. aureus in treated water was estimated by the exponential dose–response model developed by Rose & Haas (1999), given as follows:
(4)
where P(inf) is the probability of being infected after daily exposure (per person per day), r is the pathogen infectivity constant for S. aureus, and d is the exposure dose calculated in Equation (3) (pathogens day−1).
For Salmonella typhi and Shigella, the beta-Poisson dose–response model was used to calculate the infection as shown below (NRC 1983; Haas et al. 2014):
(5)
where P(inf) is the probability of infection from a single exposure, d is the exposure dose of the microorganism, N50 is the median infectious dosage, and α is the distribution parameter. For multiple exposures to pathogens within a year, the individual probabilities were summed over a specified period of the year (Busgang et al. 2015). Thus, the likelihood of annual infection was calculated as follows:
(6)
where the probability of infection per day (P(inf)/day) was obtained from Equation (6) above and n is the number of exposure days in a year (365 days). The probability of illness was calculated as follows:
(7)
where P(ill/inf) is the likelihood of illness per infection, and P(inf) annual is defined in equation (6). Table 1 shows the model parameters related to all models used in the QMRA.
Table 1

Dose–response parameter for all bacteria investigated

OrganismParameterType of modelReference
Dose–response model 
Salmonella typhi α = 1.75 × 10−1
N50 = 1.11 × 106
PDi = 0.45 
beta-Poisson Abia et al. (2016
Shigella sp. α = 0.162;
N50 = 1.127 × 103
PDi = 0.35 
beta-Poisson Van Lier et al. (2016)  
S. aureus r = 7.64E-08
PDi = 1 
Exponential Rose & Haas (1999); Busgang et al. (2018)  
Disease burden 
Salmonella typhi 0.049 Havelaar et al. (2012)  
Shigella sp. 0.026 Van Lier et al. (2016)  
S. aureus 0.049 Havelaar et al. (2012)  
OrganismParameterType of modelReference
Dose–response model 
Salmonella typhi α = 1.75 × 10−1
N50 = 1.11 × 106
PDi = 0.45 
beta-Poisson Abia et al. (2016
Shigella sp. α = 0.162;
N50 = 1.127 × 103
PDi = 0.35 
beta-Poisson Van Lier et al. (2016)  
S. aureus r = 7.64E-08
PDi = 1 
Exponential Rose & Haas (1999); Busgang et al. (2018)  
Disease burden 
Salmonella typhi 0.049 Havelaar et al. (2012)  
Shigella sp. 0.026 Van Lier et al. (2016)  
S. aureus 0.049 Havelaar et al. (2012)  

Risk characterization

Results from hazard identification, exposure assessment, and dose–response assessment were used in characterizing the risk of infection/illness after consuming treated water. The annual infection risks and disease burden resulting from the consumption of water treated by each local water household treatment method was estimated based on the U.S. EPA annual probability risks benchmark (≤10−4 pppy) and the World Health Organization (WHO) disease burden benchmark (≤10−6 DALYs/pppy) (U.S. EPA 2005; World Health 2008). The disease burden was determined as follows (Hadi et al. 2019):
(8)
where DB is the disease burden expressed in DALYs per person per year (DALYs/pppy). P(inf) annual is the annual infection risk (pppy), P(ill/inf) is the probability of illness to infection ratio, B is the pathogen-specific disease burden, and Sf is the susceptible fraction of the population which is assumed to be 1 (Sf = 1) (Busgang et al. 2015). The study assumed that the whole population was susceptible to infection caused by all pathogens in treated water, there was no immunity development, and the concentration of pathogens in treated water was constant (Busgang et al. 2015). However, elderly adults, young children, and immunocompromised individuals have the highest risks of infection and severe illness (Wibuloutai et al. 2019).

Data analysis

Data for evaluating the efficiency of the various LHWT methods were analyzed using the IBM Statistics SPSS version 21.0 (SPSS, Chicago, USA). A one-way analysis of variance was used to compare the means at a significant level of p < 0.05. Microsoft Excel, version 2013, was used to plot graphs and to determine the efficiency of treatment methods.

Physical properties

Table 2 shows the physical properties of the 45 water samples measured during collection. The temperature ranged from 22.3 to 24.2°C. Nkwen well water had a brown color and was highly turbid.

Table 2

Physical properties of water samples at point-source

LocationSourceColorOdorMean turbidity (NTU)Mean temperature (°C)Mean pH
Up-station (Bamenda I) Pipe-borne Colorless Odorless 1.7 23.5 6.2 
Covered-well Colorless Odorless 1.5 24.1 5.5 
Borehole Colorless Odorless 0.6 23.9 6.2 
Old-town (Bamenda) Pipe-borne Colorless Odorless 2.1 22.3 6.2 
Covered-well Colorless Yes 1.5 23.5 5.1 
Borehole Colorless Odorless 0.4 23.8 5.8 
Nkwen–Bamenda III Pipe-borne Colorless Odorless 1.7 22.7 6.2 
Uncovered-well Brown Yes 2.1 23.4 5.6 
Borehole Colorless Odorless 0.5 24.2 5.5 
LocationSourceColorOdorMean turbidity (NTU)Mean temperature (°C)Mean pH
Up-station (Bamenda I) Pipe-borne Colorless Odorless 1.7 23.5 6.2 
Covered-well Colorless Odorless 1.5 24.1 5.5 
Borehole Colorless Odorless 0.6 23.9 6.2 
Old-town (Bamenda) Pipe-borne Colorless Odorless 2.1 22.3 6.2 
Covered-well Colorless Yes 1.5 23.5 5.1 
Borehole Colorless Odorless 0.4 23.8 5.8 
Nkwen–Bamenda III Pipe-borne Colorless Odorless 1.7 22.7 6.2 
Uncovered-well Brown Yes 2.1 23.4 5.6 
Borehole Colorless Odorless 0.5 24.2 5.5 

The number of colony-forming units counted

Table 3 shows the data obtained from bacteria viable plate count, where LCBT is the level of contamination before treatment and LCAT is the level of contamination after treatment. Twenty-five out of the 45 (55.56%) water samples collected for the studies before treatment were seen to be contaminated. Only 20 water samples (44.44%) were free of pathogens and suitable for drinking as they met the WHO guidelines for quality drinking water (Table 3). Water samples from wells were the most contaminated, with samples from Nkwen having the highest bacteria load (298 CFU/ml). Water samples from boreholes had the lowest level of contamination, with the highest bacterial load of 4 CFU/ml before treatment (4 CFU/ml) (Table 3). After disinfection, 86.67% of the 45 treated samples met the WHO drinking water standard, and 13.33% remained contaminated (not safe for drinking). Well water samples treated by filtration were the most highly contaminated (66.67%) (Table 3).

Table 3

Average bacterial colonies counted before and after treatment

SourceSite (5 samples per site)LCBT/(CFU/ml)LCAT/(CFU/ml)
FilteringBoilingChlorineFlocculationSODIS
Pipe-borne Up-station 
Nkwen 
Old-town 57 17 
Well Up-station 31 
Nkwen 298 145 73 
Old-town 109 13 
Borehole Up-station 
Nkwen 
Old-town 
WHO standard ≤ 0.01 CFU/ml (1 CFU/100 ml) 
SourceSite (5 samples per site)LCBT/(CFU/ml)LCAT/(CFU/ml)
FilteringBoilingChlorineFlocculationSODIS
Pipe-borne Up-station 
Nkwen 
Old-town 57 17 
Well Up-station 31 
Nkwen 298 145 73 
Old-town 109 13 
Borehole Up-station 
Nkwen 
Old-town 
WHO standard ≤ 0.01 CFU/ml (1 CFU/100 ml) 

There was a complete absence of pathogens of interest in borehole water sources following treatment with different LHWT methods. However, these methods partially disinfected water from well and pipe-borne sources (Figure 2). No bacteria growth was observed in boiled and chlorinated water samples (Figure 2). There was a significant difference (p < 0.01) between the source of water and the method of disinfection, implying that the effectiveness of each water treatment method depends on the water source. However, there was no significant difference between the effect of various treatment methods on disinfection (p > 0.05).
Figure 2

Effects of the treatment method of disinfection on water source.

Figure 2

Effects of the treatment method of disinfection on water source.

Close modal
Amongst the three locations, water sources from Nkwen had the highest level of contamination before and after treatment. Well water samples from Nkwen had the highest bacteria load, followed by well and pipe-borne water samples from Old-town and well water samples from Up-station (Figure 3). There was a significant difference (α = 0.05; p < 0.10) between the effectiveness of various treatment methods and the location of sources of water, indicating that the location of the source of water determines the effectiveness of the treatment method.
Figure 3

Effect of the method of disinfection on the location of water source.

Figure 3

Effect of the method of disinfection on the location of water source.

Close modal

Effectiveness of disinfection methods

All five household water treatment methods were 100% effective in treating borehole water (Figure 4). The most effective method of treatment for well water was boiling (100%) and chlorination (100%), followed by flocculation (98.25%), SODIS (85.96%), and filtration (62.33%), respectively (Figure 4).
Figure 4

Effectiveness of each method of treatment on different water sources.

Figure 4

Effectiveness of each method of treatment on different water sources.

Close modal

Risk of pathogenic infection

The concentration of Salmonella typhi, Shigella sp., and S. aureus was highest in well water samples from Nkwen, disinfected by filtration, flocculation, and SODIS, and maximum concentration values were 87.1, 56.2, and 76.3 CFU/100 ml, respectively (Table 4). Filtered water samples had the highest risk of infection. However, infection risk per day varied with the source of drinking water. The risk of infection per day due to Shigella sp. in well water was high with maximum values of 6.95 × 10−1 and 6.75 × 10−1 pppd in filtered and flocculated water samples, while Salmonella typhi and S. aureus had a low infection risk, with maximum values of 1.55 × 10−1 and 1.16 × 10−2 pppd, respectively (Table 4). The infection risk of pathogens per day for children was not significantly different from the infection risk of pathogens per day for adults.

Table 4

Probability of infection per child/adult per day exposed to pathogens in treated water under the dose–response model

Treatment methodWater sourceMean concentration of Bacteria CFU/100 mlQuantity of water consumed/children (ml)Quantity of water consumed/adult (ml)Exposure dose/childrenExposure dose/adultP(inf)/children (pppd)Pinf/adult (pppd)
Filtration Salmonella typhi 
Pipe-borne water Old-town 17.2 1,000 2,000 17,200 34,400 1.55 × 10−1 2.43 × 10−2 
Well water Nkwen 89.1 1,000 2,000 89,100 178,200 6.16 × 10−1 1.19 × 10−1 
Shigella sp. 
Well water Up-station 7.3 1,000 2,000 7,300 14,600 6.29 × 10−1 6.69 × 10−1 
Well water Old-town 13.2 1,000 2,000 13,200 26,400 6.64 × 10−1 6.99 × 10−1 
S. aureus 
Well water Nkwen 56.2 1,000 2,000 56,200 112,400 4.30 × 10−3 8.60 × 10−1 
Flocculation Shigella sp. 
Well water Nkwen 8.1. 1,000 2,000 8,100 16,200 6.36 × 10−1 6.75 × 10−1 
SODIS S. aureus 
Well water Nkwen 76.3 1,000 2,000 76,300 152,600 5.80 × 10−3 1.16 × 10−2 
Treatment methodWater sourceMean concentration of Bacteria CFU/100 mlQuantity of water consumed/children (ml)Quantity of water consumed/adult (ml)Exposure dose/childrenExposure dose/adultP(inf)/children (pppd)Pinf/adult (pppd)
Filtration Salmonella typhi 
Pipe-borne water Old-town 17.2 1,000 2,000 17,200 34,400 1.55 × 10−1 2.43 × 10−2 
Well water Nkwen 89.1 1,000 2,000 89,100 178,200 6.16 × 10−1 1.19 × 10−1 
Shigella sp. 
Well water Up-station 7.3 1,000 2,000 7,300 14,600 6.29 × 10−1 6.69 × 10−1 
Well water Old-town 13.2 1,000 2,000 13,200 26,400 6.64 × 10−1 6.99 × 10−1 
S. aureus 
Well water Nkwen 56.2 1,000 2,000 56,200 112,400 4.30 × 10−3 8.60 × 10−1 
Flocculation Shigella sp. 
Well water Nkwen 8.1. 1,000 2,000 8,100 16,200 6.36 × 10−1 6.75 × 10−1 
SODIS S. aureus 
Well water Nkwen 76.3 1,000 2,000 76,300 152,600 5.80 × 10−3 1.16 × 10−2 

Disease burden

The risk of infection per day due to Shigella sp. in well water was high, with maximum values of 6.96 × 10−1 and 6.75 × 10−1 pppd in filtered and flocculated water samples (Table 5). The annual risk of infection due to Salmonella typhi, Shigella sp., and S. aureus ranged from 7.80 × 10−1 to 1.00 pppy. The maximum risk of illness was 4.50 × 10−1, 3.30 × 10−1, and 9.80 × 10−1 for Salmonella typhi, Shigella sp., and S. aureus, respectively (Table 5). S. aureus and Salmonella typhi contributed the highest disease burden with a maximum value of 4.71 × 10−2 and 2.13 × 10−2 DALYs/pppy (Table 5). The annual infectious risk and disease burden of S. aureus and Salmonella typhi in water samples treated by filtration, flocculation, and SODIS exceeded the U.S. EPA annual infection benchmark (≤10−4 pppy−1) and the WHO disease burden benchmark (≤10−6 DALYs/pppy) (U.S. EPA 2005; WHO 2008) (Table 5). There was no significant difference between the annual risk of infection, risk of illness, and disease burden values for adults and children. S. aureus had the highest probability of the infection developing into a symptomatic illness (9.8 × 10−1), while Shigella sp. had the lowest probability of symptomatic infection (3.0 × 10−1) (Table 5).

Table 5

Probability of illness per child/adult per day and disease burden due to exposure to the pathogen in treated water

Treatment methodWater sourceP(inf)/childrenP(inf)/adultPa(inf)/children (pppy-1)Pa (inf)/adult (pppy)P(ill)/childrenP(ill)/adultDB/children (DALYs/pppy)DB/adults (DALYs/pppy)
Filtration Salmonella typhi 
 Pipe-borne water –Old-town 1.55 × 10−1 2.43 × 10−2 1.00 9.90 × 10−1 4.50 × 10−1 4.40 × 10−1 2.07 × 10−2 2.13 × 10−2 
 Well water Nkwen 6.16 × 10−2 1.19 × 10−1 1.00 1.00 4.50 × 10−1 4.50 × 10−1 2.07 × 10−2 2.07 × 10−2 
 Shigella sp.         
 Well water Up-station 6.29 × 10−1 6.69 × 10−1 1.00 1.00 3.50 × 10−1 3.30 × 10−1 9.10 × 10−3 8.58 × 10−3 
 Well water Old-town 6.64 × 10−1 6.99 × 10−1 1.00 1.00 3.50 × 10−1 3.00 × 10−1 9.10 × 10−3 7.80 × 10−3 
 S. aureus         
 Well water Nkwen 4.50 × 10−3 8.60 × 10−3 7.90 × 10−1 9.60 × 10−1 7.90 × 10−1 9.60 × 10−1 2.87 × 10−3 4.52 × 10−2 
Flocculation Shigella sp. 
 Well water Nkwen 6.36 × 10−1 6.75 × 10−1 1.00 1.00 3.50 × 10−1 3.30 × 10−1 9.10 × 10−3 8.58 × 10−3 
SODIS S. aureus 
 Well water Nkwen 5.80 × 10−3 1.16 × 10−2 8.8 × 10−1 9.8 × 10−1 8.8 × 10−1 9.8 × 10−1 3.65 × 10−2 4.71 × 10−2 
Treatment methodWater sourceP(inf)/childrenP(inf)/adultPa(inf)/children (pppy-1)Pa (inf)/adult (pppy)P(ill)/childrenP(ill)/adultDB/children (DALYs/pppy)DB/adults (DALYs/pppy)
Filtration Salmonella typhi 
 Pipe-borne water –Old-town 1.55 × 10−1 2.43 × 10−2 1.00 9.90 × 10−1 4.50 × 10−1 4.40 × 10−1 2.07 × 10−2 2.13 × 10−2 
 Well water Nkwen 6.16 × 10−2 1.19 × 10−1 1.00 1.00 4.50 × 10−1 4.50 × 10−1 2.07 × 10−2 2.07 × 10−2 
 Shigella sp.         
 Well water Up-station 6.29 × 10−1 6.69 × 10−1 1.00 1.00 3.50 × 10−1 3.30 × 10−1 9.10 × 10−3 8.58 × 10−3 
 Well water Old-town 6.64 × 10−1 6.99 × 10−1 1.00 1.00 3.50 × 10−1 3.00 × 10−1 9.10 × 10−3 7.80 × 10−3 
 S. aureus         
 Well water Nkwen 4.50 × 10−3 8.60 × 10−3 7.90 × 10−1 9.60 × 10−1 7.90 × 10−1 9.60 × 10−1 2.87 × 10−3 4.52 × 10−2 
Flocculation Shigella sp. 
 Well water Nkwen 6.36 × 10−1 6.75 × 10−1 1.00 1.00 3.50 × 10−1 3.30 × 10−1 9.10 × 10−3 8.58 × 10−3 
SODIS S. aureus 
 Well water Nkwen 5.80 × 10−3 1.16 × 10−2 8.8 × 10−1 9.8 × 10−1 8.8 × 10−1 9.8 × 10−1 3.65 × 10−2 4.71 × 10−2 

where P (inf) - probability of infection, Pa (inf) - probability of annual infection, P (ill) - probability of illness, and DB - disease burden

Bacteriological quality of water treated by LHWT methods

According to the WHO guidelines, safe drinking water should contain no pathogens to avoid water-borne diseases, and pathogenic safety levels should be less than 1 CFU/100 ml of water (WHO 2006). In this study, less than half of the drinking water samples met these criteria (Table 3), underpinning the need for a household treatment of drinking water in Bamenda. However, the choice of an appropriate LHWT method depends on the source of water since the bacteria load in the current study varied with treatment method and water source (Table 3). The significant difference (p ≤ 0.05) between the effectiveness of LHWT methods in the current study was probably due to the different levels of exposure of water sources to microbes (Rosa et al. 2014). Pipe-borne water may not be considered safe for consumption since samples from Old-town had low microbial concentrations (Table 3), probably due to the presence of contaminants in storage tanks or pipes during transportation (Katukiza et al. 2014). Borehole sources had the lowest contamination levels (Table 3), confirming findings by Njunda and colleagues at the Cameroon Development Corporation (Njunda 2013). Contrarily, locally dug well water sources were the most contaminated (Table 3), aligning with the results of Magha (2021) in Bamenda. The high contamination of exposed well water sources may be attributed to possible wash-off from fecal contaminants from neighboring toilets or sewage from poor sewer systems. Treatment methods differ in their effectiveness in preventing water-borne pathogens (Figure 4). For instance, chemical treatment with sodium hypochlorite and heat treatment by boiling showed no bacteria prevalence for tested pathogens, indicating their high effectiveness (100%) for use in treating water unless resources are unavailable. This result corroborates findings established by Okwadha & Ahmed (2017). However, chlorination may be more cost-effective for poor households. According to Mohamed et al. (2016), decisions related to scaling up LHWT practices are affected by factors such as effectiveness, adherence (correct, consistent, and sustained use), and the cost of achieving the desired aim of reducing pathogens.

SODIS was ineffective in disinfecting well water samples from Nkwen (Table 3), probably due to high bacteria load, high turbidity, and high concentration of dissolved organic matter. Normally, for solar treatment to be efficient, turbid water must be made clear by filtration or sedimentation. Little to no sunlight on rainy days also reduces the effectiveness of SODIS and prolongs disinfection time, making disinfection by solarization difficult for consistent use in the rainy season. Moreover, ultraviolet radiation catalysis the degradation of plastics to Bisphenol-A (BPA), which can diffuse into treated water in plastic bottles to pose a carcinogenic risk (Seachrist et al. 2016). Furthermore, recent research findings have proven the occurrence of carcinogenic microplastics, phthalates, and alkylphenol in plastic bottled water (Amiridou & Voutsa 2011; Gambino et al. 2022). These factors greatly limit or discourage the use of solarization method of disinfection. In the current study, filtration had the lowest effectivity (Figure 4), though the effectiveness of filtration may depend on the filtering apparatus. The current study used ceramic filters with pore sizes of 0.3–3 μm, commonly and frequently used in treating water locally by most households in Bamenda.

Health risk assessment of treated water

The approach of QMRA is a comparatively new approach in identifying health risks associated with pathogens in drinking water (Bentham & Whiley 2018). In the current study, the overall health risks (Tables 4 and 5) due to Salmonella typhi, Shigella sp., and S. aureus in locally treated water may be ascribed to the low infectious dose of pathogens (Katukiza et al. 2014), since all pathogenic concentrations above the WHO benchmark (≤1 CFU/100 ml) posed a high annual risk of infection (Table 4). The overall high risks of illness due to S. aureus, Shigella sp., and Salmonella in the current study are congruent with other studies on drinking water (Busgang et al. 2015; Ahmed et al. 2020). However, these findings were based on untreated water samples. Water samples treated by filtration contributed the highest overall risks due to the presence of all three pathogens, indicating that health risks may also vary with local household treatment methods.

Disease burden of pathogens in treated water

Diseases related to inadequate water, sanitation, and hygiene contribute to a huge burden of disease in low-income countries (Ahmed et al. 2020). The disease burden of all three pathogens in treated water samples exceeded the WHO disease burden benchmark. Katukiza et al. (2014) also found a high burden of water-borne diseases in a low-income population who use surface water as the main source of drinking. Filtered well water samples contributed to the highest disease burden (Table 5), probably due to the high prevalence of all three pathogens, putting the proportion of the population that depends on well water as a drinking source and the use of artificial filters at risk of water-borne disease. Thus, to minimize the overall risk of illness from infection by tested pathogens, users should treat well water by boiling or chlorination. However, filtration, solarization, and flocculation can be effective if combined with other treatment methods. Salmonella typhi was present in both well and pipe-borne water sources and contributed to the highest proportion of disease burden (Table 5). Globally, about 212 million cases of typhoid, with 129,000 deaths, are reported yearly, with children and young adults being the vulnerable groups (Steele et al. 2016). Several studies in Cameroon have reported a high prevalence of typhoid fever due to Salmonella sp infection (Njoya et al. 2021; Ndip et al. 2022). The global burden of Shigella and other water-borne illnesses is mostly unknown because most of the cases go unreported (Zahid 2018).

The current study is, however, limited because it did not consider the uncertainty of risk estimates as in the case of Monte Carlo risk analysis. Also, the prevalence of Salmonella sp. in treated water samples suggests fecal contamination and, therefore, the possibility of contamination by other fecal pathogens such as Vibro cholerae, E. coli, Campylobacter, and Cryptosporidium bacteria with time and availability. However, based on available resources, the current study investigated only three water-borne pathogens. Also, previous studies by Fonyuy (2014) and Biosengazeh et al. (2020) show high prevalence of illnesses such as typhoid fever, dysentery, pneumonia, and skin infections caused by Salmonella typhi, Shigella sp., and S. aureus in Bamenda probably due to contaminated water, since most households used local water treatment methods to reduce disease risk especially during water stress, when water supply is unreliable or scarce.

For the pathogen tested, chlorination and boiling were the most efficient disinfection methods, regardless of the water source. The overall health risks for both children and adults were not significantly different. Drinking water from well sources treated by filtration with a ceramic filter of pore size 0.3–3 μm contributed to the highest health risks. Salmonella typhi and Shigella sp. had approximately 1.00 annual probability risk of infection. S. aureus had the highest probability risk of developing into illness (9.80 × 10−1), while Salmonella typhi had the lowest (4.50 × 10−1). The overall probability risk of infection and disease burden exceeded the U.S. EPA annual infection and the WHO disease burden benchmark. The filtration, flocculation, and SODIS did not reduce pathogenic concentrations below risk levels. Although the boiling and chlorination methods did not pose risks for the pathogens tested, the study did not address a wide range of pathogens.

The authors wish to acknowledge Dr Polycarp Ndikvu Chia, the Head of the Department of Biochemistry of the Catholic University of Cameroon, Bamenda, for laboratory assistance. The authors also acknowledge the travel support received within the Global-SDG-Campus project funded by the DAAD.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.

Abendong
A. A.
,
Emmanuel
E.
,
Fongoh
E. J.
&
Akoachere
R. A.
2019
Assessment of groundwater quality in Bamenda–Cameroon for suitable applications
.
S.N. Applied Sciences
1
(
11
),
1389
.
doi:10.1007/s42452-019-1351-1
.
Abia
A. L. K.
,
Ubomba-Jaswa
E.
,
Genthe
B.
&
Momba
M. D. B.
2016
Quantitative microbial risk assessment (QMRA) shows increased public health risk associated with exposure to river water under conditions of riverbed sediment resuspension
.
Science of the Total Environment
566
(
567
),
1143
1151
.
doi:10.1016/j.scitotenv.2016.05.155
.
Ahmed
J.
,
Wong
L. P.
,
Chua
Y. P.
,
Channa
N.
,
Mahar
R. B.
,
Yasmin
A.
,
VanDerslice
J. A.
&
Garn
J. V.
2020
Quantitative microbial risk assessment of drinking water quality to predict the risk of water-borne diseases in primary-school children
.
International Journal of Environmental Research and Public Health
17
(
8
).
doi:10.3390/ijerph17082774
.
Amiridou
D.
&
Voutsa
D.
2011
Alkylphenols and phthalates in bottled waters
.
Journal of Hazardous Materials
185
(
1
),
281
286
.
doi:10.1016/j.jhazmat.2010.09.031
.
Anyang
N.
2021
Evaluation of the variability of drinking water quality within Bamenda Metropolis, North West Region, Cameroon
.
Current Journal of Applied Science and Technology
40
(
27
),
54
70
.
doi:10.9734/cjast/2021/v40i2731527
.
Bentham
R.
&
Whiley
H.
2018
Quantitative microbial risk assessment and opportunist water-borne infections – Are there too many gaps to fill?
International Journal of Environmental Research and Public Health
15
(
6
).
doi:10.3390/ijerph15061150
.
Biosengazeh
N.
,
Mofor
N.
,
Tamungang
N.
&
Mvondo-Ze
A.
2020
Assessment of ground water quality in Baba I Village, North-West Cameroon
.
Journal of Geoscience and Environment Protection
8
,
87
104
.
doi:10.4236/gep.2020.84007
.
Busgang
A.
,
Friedler
E.
,
Ovadia
O.
&
Gross
A.
2015
Epidemiological study for the assessment of health risks associated with greywater reuse for irrigation in arid regions
.
Science of the Total Environment
538
,
230
239
.
doi:10.1016/j.scitotenv.2015.08.009
.
Busgang
A.
,
Friedler
E.
,
Gilboa
Y.
&
Gross
A.
2018
Quantitative microbial risk analysis for various bacterial exposure scenarios involving greywater reuse for irrigaIon
.
Water-Sui
10
(
4
),
413
.
CDC
2008
Household Water Treatment Otions in Developing Countries: Solar Disinfection (SODIS). pdf presentation. Center for Disease Control and Prevention, New York. https://www.cdc.gov/safewater/publications_pages/options-sodis.pdf [Accessed: 15.03.2010]
.
Chiaga
N. F.
2019
Catchment management and the sustainability of urban water supply. Evidence from Bamenda, Cameroon
.
Canadian Journal of Tropical Geography/Revue Canadienne de Géographie Tropicale
6
(
2
),
1
8
.
Enger
K. S.
,
Nelson
K. L.
,
Clasen
T.
,
Rose
J. B.
&
Eisenberg
J. N.
2012
Linking quantitative microbial risk assessment and epidemiological data: Informing safe drinking water trials in developing countries
.
Environmental Science & Technology
46
(
9
),
5160
5167
.
doi:10.1021/es204381e
.
Fonyuy
B. E.
2014
Prevalence of water borne diseases within households in the Bamendankwe Municipality-North West Cameroon
.
Journal of Biosafety and Health Education
2
,
Article No. 122. doi:10.4172/2332-0893.1000122
.
Gambino
I.
,
Bagordo
F.
,
Grassi
T.
,
Panico
A.
&
De Donno
A.
2022
Occurrence of microplastics in tap and bottled water: Current knowledge
.
International Journal of Environmental Research and Public Health
19
(
9
).
doi:10.3390/ijerph19095283
.
Haas
C. N.
,
Rose
J. B.
&
Gerba
C. P.
2014
Quantitative Microbial Risk Assessment, 2nd edn. John Wiley & Sons, New York
.
Hadi
M.
,
Mesdaghinia
A.
,
Yunesian
M.
,
Nasseri
S.
,
Nodehi
R. N.
,
Smeets
P. W.
,
Schijven
J.
,
Tashauoei
H.
&
Jalilzadeh
E.
2019
Optimizing the performance of conventional water treatment system using quantitative microbial risk assessment, Tehran, Iran
.
Water Research
1
(
162
),
394
408
.
Havelaar
A. H.
,
Haagsma
J. A.
,
Mangen
M. J.
,
Kemmeren
J. M.
,
Verhoef
L. P. B.
,
Vijgen
S. M. C.
,
Wilson
M.
,
Friesema
I. H. M.
,
Kortbeek
L. M.
,
van Duynhoven
Y. T. H. P.
&
van Pelt
W.
2012
Disease burden of foodborne pathogens in the Netherlands
.
International Journal of Food Microbiology
156
(
3
),
231
238
.
Katukiza
A. Y.
,
Ronteltap
M.
,
van der Steen
P.
,
Foppen
J. W.
&
Lens
P. N.
2014
Quantification of microbial risks to human health caused by water-borne viruses and bacteria in an urban slum
.
Journal of Applied Microbiology
116
(
2
),
447
463
.
doi:10.1111/jam.12368
.
Kouamé
P. K.
,
Nguyen-Viet
H.
,
Dongo
K.
,
Zurbrügg
C.
,
Biémi
J.
&
Bonfoh
B.
2017
Microbiological risk infection assessment using QMRA in agriculture systems in Côte d'Ivoire, West Africa
.
Environmental Monitoring and Assessment
189
(
11
),
587
.
doi:10.1007/s10661-017-6279-6
.
Lantagne
D.
,
Quick
R.
&
Mintz
E.
2006
Household Water Treatment and Safe Storage Options in Developing Countries: A Review of Current Implementation Practices
.
Wilson Quarterly, Woodrow Wilson International Center for Scholars Environmental Change and Security Program
.
Magha
A.
2021
Assessment of groundwater quality for domestic and irrigation purposes in Northern Bamenda (Cameroon)
.
Journal of Water Resource and Protection
13
,
1
19
.
doi:10.4236/jwarp.2021.131001
.
Medema
G.
&
Ashbolt
N.
2006
QMRA: Its Value for Risk Management
.
WHO, Geneva
.
Membré
J. M.
,
Diao
M.
,
Thorin
C.
,
Cordier
G.
,
Zuber
F.
&
André
S.
2015
Risk assessment of proteolytic Clostridium botulinum in canned foie gras
.
International Journal of Food Microbiology
210
,
62
72
.
doi:10.1016/j.ijfoodmicro.2015.06.002
.
Miles
A. A.
,
Misra
S. S.
&
Irwin
J. O.
1938
The estimation of the bactericidal power of the blood
.
The Journal of Hygiene
38
(
6
),
732
749
.
doi:10.1017/s002217240001158x. PMC 2199673. PMID 20475467
.
Mohamed
H.
,
Clasen
T.
,
Njee
R. M.
,
Malebo
H. M.
,
Mbuligwe
S.
&
Brown
J.
2016
Microbiological effectiveness of household water treatment technologies under field use conditions in rural Tanzania
.
Tropical Medicine & International Health
21
(
1
),
33
40
.
doi:10.1111/tmi.12628
.
Mufur
A. M.
,
Awah
M. T.
,
Nono
G. D. K.
,
Tamfuh
P. A.
,
Wotchoko
P.
&
Beyala
K. K. V.
2021
Physico-chemical and bacteriological characterization of surface water in Bamenda (North West Cameroon)
.
Applied Water Science
11
(
12
),
185
.
doi:10.1007/s13201-021-01512-3
.
Nde
C. J.
2017
The economic burden of water related infections in the Bamenda Health District: The case of diarrhoea
.
Universal Journal of Public Health
5
(
4
),
176
182
.
doi:10.13189/ujph.2017.050407
.
Ndenecho
E.
2004
Geomorphic and anthropogenic factors influencing landslides in the Bamenda highlands, N.W Province, Cameroon
.
Journal of Applied Social Sciences
1
,
15
26
.
Ndip
R. A.
,
Awah
L. A.
,
Ghogomu
S.
,
Cho-Ngwa
F.
&
Ngemenya
M.
2022
Isolation and molecular identification of Salmonella with high multidrug resistance to first-line typhoid antibiotics in Southwest Cameroon
.
Microbes and Infectious Diseases
.
doi:10.21608/mid.2022.122411.1249
.
Njoya
H. F.
,
Awolu
M. M.
,
Christopher
T. B.
,
Duclerc
J. F.
,
Ateudjieu
J.
,
Wirsiy
F. S.
,
Atuhaire
C.
&
Cumber
S. N.
2021
Prevalence and awareness of mode of transmission of typhoid fever in patients diagnosed with Salmonella typhi and paratyphi infections at the Saint Elisabeth General Hospital Shisong, Bui Division, Cameroon
.
The Pan African Medical Journal
40
,
83
.
doi:10.11604/pamj.2021.40.83.16893
.
Njunda
L. A.
2013
Bacteriological analysis of drinking water in the Cameroon Development Corporation (CDC) Penda-Mboko
.
African Journal of Integrated Health
2
(
1
),
26
31
.
NRC
.
1983
Risk Assessment in the Federal Government: Managing the Process
.
National Research Council
,
Washington, DC
.
Okwadha
G. D. O.
&
Ahmed
A. A.
2017
Determination of effectiveness of traditional drinking water treatment methods
.
International Journal of Advanced Engineering Research and Applications (IJA-ERA)
2
(
10
),
592
604
.
Rose
J. B.
&
Haas
C. N.
1999
A risk assessment framework for the evaluation of skin infections and the potential impact of antibacterial soap washing
.
American Journal of Infection Control
27
(
6
),
S26
S33
.
doi:10.1016/s0196-6553(99)70039-8
.
Sari
S. Y. I.
,
Alfian
A. R.
,
Respati
T.
,
Agustian
D.
&
Raksanagara
A. S.
2019
Comparison of drinking water quality following boiling, household filtration and water-refill in urban-slum area
.
Journal of International Dental and Medical Research
12
(
2
),
791
796
.
Saturday
A.
2016
Performance of household water treatment methods for microbial removal under household conditions in Kabale District, Uganda
.
Journal of Environmental Health Sciences
2
(
2
),
1
9
.
doi:10.15436/2378-6841.16.1013
.
Seachrist
D. D.
,
Bonk
K. W.
,
Ho
S. M.
,
Prins
G. S.
,
Soto
A. M.
&
Keri
R. A.
2016
A review of the carcinogenic potential of bisphenol A
.
Reproductive Toxicology
59
,
167
182
.
doi:10.1016/j.reprotox.2015.09.006
.
Steele
A. D.
,
Hay Burgess
D. C.
,
Diaz
Z.
,
Carey
M. E.
&
Zaidi
A. K. M.
2016
Challenges and opportunities for typhoid fever control: A call for coordinated action
.
Clinical Infectious Diseases
62
(
suppl_1
),
S4
S8
.
doi:10.1093/cid/civ976
.
U.S. EPA.
2000
Methodology for Deriving Ambient Water Quality Criteria for the Protection of Human Health
.
Office of Water Washington, DC, EPA/822-00/004
,
Washington, DC
.
U.S. EPA
.
2005
Occurrence and Exposure Assessment for the Final Long-Term 2 Enhanced Surface Water Treatment Rule
.
Washington, DC
.
Van Lier
A.
,
McDonald
S. A.
,
Bouwknegt
M.
,
Kretzschmar
M. E.
,
Havelaar
A. H.
,
Mangen
M.-J. J.
,
Wallinga
J.
&
de Melker
H. E.
2016
Disease burden of 32 infectious diseases in the Netherlands, 2007–2011
.
PLoS ONE
11
,
e0153106
.
WHO
.
2006
Protecting Groundwater for Health: Managing the Quality of Drinking-Water Sources
(
Schmoll
O.
,
Howard
G.
,
Chilton
J.
&
ChorusI.
M.
ed.)
IWA Publishing
,
London
.
WHO/UNICEF
.
2018
Joint Monitoring Programme (JMP) for Water Supply and Sanitation: Progress on Drinking Water, Sanitation, and Hygiene
.
Wibuloutai
J.
,
Thanomsangad
P.
,
Benjawanit
K.
&
Mahaweerawat
U.
2019
Microbial risk assessment of drinking water filtration dispenser toll machines (DFTMs) in Mahasarakham province of Thailand
.
Water Supply
19
(
5
),
1438
1445
.
doi:10.2166/ws.2019.016
.
Wirba
L. N.
2020
Exploring water management practices and sustainability implications in the Bamenda Metropolis of Cameroon
.
International Journal of Global Sustainability
4
(
1
).
doi:10.5296/ijgs.v4i1.17649
.
World Health Organization
.
2008
Guidelines for Drinking-Water Quality [Electronic Resource]: Incorporating 1st and 2nd Addenda, Vol. 1, Recommendations
, 3rd edn.
World Health Organization
.
World Health Organization (WHO)
.
2011
Guidelines for Drinking Water Quality
, 4th edn.
World Health Organization
,
Geneva
,
Switzerland
.
Zahid
J.
2018
Impact of Clean Drinking Water and Sanitation on Water Borne Diseases in Pakistan
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).