Assessing the Likelihood of Staphylococcus aureus Contamination in Bottled Drinking Water Production ^†

Abstract

This study’s objectives were to evaluate the possibility of S. aureus contamination in bottled drinking water and to determine the elements that affect the level of S. aureus in raw water. In two drinking water treatment facilities, samples of raw water, soft water, reverse osmosis (R.O.) water, and finished water were collected at various stages. In addition, raw water samples and risk indicators like pH, temperature, and residual chlorine were gathered at the packaging facility during the washing process. For Factory A (small scale), the pH values for the raw water, soft water, R.O. water, and finished water samples were 7.17, 7.24, 6.69, and 5.92, respectively. For Factory B (medium scale), the pH values were 7.9, 7.44, 6.97, and 6.8. All water samples from Factory A (2 CFU/mL) and Factory B (1 CFU/mL) had S. aureus concentrations that were within the acceptable range for human consumption. All water samples from Factory A (2–26 CFU/mL) and Factory B (11–316 CFU/mL) contained total coliforms as well. Our study revealed that S. aureus contamination in water is mostly caused by pH and processing times. To prevent pathogen contamination in bottled drinking water, it is recommended that raw and finished water be kept at a pH level between 6.5 and 7.5.

1. Introduction

About 50–70% of a person’s total weight is made up of water, which is essential for maintaining healthy cells, tissues, and organs (EPA, 2018) [1]. The source of drinking water may contain a variety of radionuclides, microorganisms, and chemicals, and the consumption of these may result in gastrointestinal disorders, brain system or reproductive issues, and long-term conditions like cancer. Typhoid fever and cholera are examples of bacteria that cause waterborne disease. Diarrhea is one of the most typical negative effects associated with contaminated water, and it can be spread by drinking contaminated water [2]. As one of the most frequent causes of food poisoning, Staphylococcus aureus produces enterotoxin from an initial concentration of 100 organisms per milliliter of water. This pathogen has been found in a variety of places, including human nasal passages, skin, clothing, food, flies’ digestive tracts, dust, and moisture droplets. It has been proven that S. aureus can be discovered in rural drinking water. Residents exposed to contaminated water may become colonized by S. aureus from drinking water [3]. Individuals are exposed to avoidable health risks when water and sanitation systems are improperly managed. According to one specific study, the prevalence of Escherichia coli and Staphylococcus aureus in the drinking water from filtration dispenser toll machines in the Mahasarakham province of Thailand was 54.17% and 16.67%, respectively, with an average concentration of 1.04 log CFU/mL and 0.26 log CFU/mL [4]. The characteristics of the contaminants in the water source determine the treatment procedure for disinfecting drinking water to decrease the prevalence of waterborne diseases [1]. Reverse osmosis (R.O.), charcoal filtration, and UV light are used to treat the water. According to the authors of [5], R.O. can eliminate 99% of the germs in tap water, as well as organic and inorganic pollutants. Drinking water may become quickly re-contaminated by other factors, particularly poor process hygiene brought on by erratic maintenance [6]. This can result in the contamination of the water with hazardous chemicals and dangerous microbes. In order to establish the high-risk parameters of processing steps as part of risk assessment, probabilistic modeling and sensitivity analyses of crucial control points in food processing and food safety systems are used [7]. Risk assessors can also utilize these models to help them make management decisions that will lower the risk of contracting foodborne illnesses [8]. In order to clarify the current state of the risk parameters for bottled drinking, it is required to model the probability of S. aureus appearing in drink water via risk assessments in the field of drinking water production.

2. Materials and Methods

2.1. Water Samples

Sample Collection

Physicochemical and microbiological tests were performed on water samples from one packing house (n = 5) and bottled drinking water (n = 3 at each step) from two factories (Factory A represents a small-scale factory, and Factory B represents a medium-scale plant). The bottled drinking water processing facility took samples of raw water, soft water, reverse osmosis water, and finished water. The samples were maintained in iceboxes after sample collection and transported to the lab (Department of Medical, Regional Medical Medical Sciences Center 5, Samut Songkhram) within two hours. Each sample was taken and examined in triplicate.

2.2. Physicochemical Analysis

The official AOAC technique was used for the physicochemical evaluations of all water samples (AOAC, 2005) [9]. The packing house gathered and evaluated the water samples’ pH, residual chlorine (ppm), temperature (°C), and processing time values. Using a pH meter (Hanna Instruments, Woonsocket, RI, USA), water samples from bottled drinking water were analyzed to determine their pH at each stage.

2.3. Microbiological Analysis

All samples were collected to measure the quantification and risk of S. aureus and total coliforms by following the FDA BAM, 2019 (Chapter 12) [10] and FDA BAM [11] procedures.

Plating Media Method for S. aureus and Total Coliforms

A quantity of 25 mL of samples was measured and rinsed in a 250 mL bottle containing 225 mL of sterile 0.1% peptone water, and then the suspension was diluted from 1 × 10⁻² to 1 × 10⁻⁴. A volume of 0.1 mL of each dilution was spread on Braid Parker agar plates (BPA) (Merck, Darmstadt, Germany) for the enumeration of S. aureus. Five tubes were used for each dilution of the multiple-dilution MPN series (used to ascertain the number of total coliforms using Violet Red Bile Agar (VRBA) (Merck, Germany)). The plates were incubated at 37 °C for 24 h to ensure the isolation of bacteria (FDA BAM 2019 [10] and FDA BAM 2020 [12]). The colony-forming units of characteristic S. aureus and total coliforms were counted and reported as CFU.

2.4. Sensitivity Analysis

To explain how S. aureus became contaminated during the various processing steps, a probabilistic risk assessment (quantitative model) was developed. Using the analysis tool program in Microsoft ExcelTM, the results of the microbiological analysis were transformed into log10 units and put through a straightforward regression analysis. This often entails fitting a relationship between inputs and outputs, such as the following linear one [13]:

Yi = bo + b1X1,i + b2X2,i + … + bmXm,i + ei

(1)

Yi is the ith output data point for the ith input data point; Xj,i is the ith input data point for the jth input; bj is the jth input’s regression coefficient; and ei is the ith data point’s error. The basis function for each term in the regression model may be distinct and either linear or nonlinear. According to Frey et al. (2003) [13], the regression coefficient, bj, for a linear model can be understood as the change in the output Yi when the input Xj,i for a given value of j increases by one unit and the values of all other inputs remain constant. Rank order correlations, a non-parametric statistic for estimating the correlational link between two means, were employed in the regression sensitivity analysis based on Spearman’s rank correlation calculations The analysis involved running simulations, which involved allocating probability distributions to the inputs and determining the impact of input variance on the output distribution. Then, tornado graphs were produced. The various input factors were represented on the graph by horizontal bars, and the length of the bars indicated the degree of association with the mean numbers and the amount of S. aureus discovered in the samples (output variables). The nominal rage sensitivity analysis equation was as follows [13]:

Sensitivity = (Output at maximum input value − Output at minimum input value)/Output at nominal input value

(2)

2.5. Statistical Analysis

All measurements were carried out in triplicate. The mean and standard deviation (±SD) values were calculated by conducting data analysis in the Excel platform.

3. Results

3.1. Physicochemical and Microbiological Properties of the Water Samples

Table 1. Results derived from our physicochemical and microbiological analysis of the water samples.

Plants	Packing House	Drinking Water Factory A				Drinking Water Factory B
Samples	Water	Raw	Soft	R.O.	Finished	Raw	Soft	R.O.	Finished
Time (min)	0–20	-	-	-	-	-	-	-	-
Temperature (°C)	17.3–19.1	-	-	-	-	-	-	-	-
pH	10.33–10.81	7.17 ± 0.25	7.24 ± 0.26	6.69 ± 0.21	5.92 ± 0.32	7.60 ± 0.33	7.44 ± 0.29	6.97 ± 0.22	6.80 ± 0.21
Residual chlorine (ppm)	0.23–1.67	-	-	-	-	-	-	-	-
Total coliform (CFU)	-	189 ± 8.60	316 ± 44.22	30 ± 44.40	145 ± 7.72	26 ± 3.89	2 ± 3.13	20 ± 1.48	13 ± 1.47
S. aureus (CFU)	3.40 LogCFU	3 ± 1.00	1 ± 0.00	1 ± 0.00	1 ± 0.00	1 ± 0.00	1 ± 0.00	1 ± 0.00	1 ± 0.00

Note: The numbers are means ± S.D. from three independent replicates.

displays the conditional factors for the bottling plant (Plant B) and the packing facility (Plant A). The packing house had water with pH values between 10.33 and 10.81, residual chlorine concentrations between 0.23 and 1.67 ppm, and temperatures between 17.3 and 19.1 °C. In order to generate the simulation models, the quantity of S. aureus (3.40 LogCFU) discovered in the water samples was evaluated. Raw, soft, R.O., and finished water had pH values of 7.17, 7.24, 6.69, and 5.92, respectively, for Plant A. For Plant B, the corresponding values were 7.60, 7.44, 6.97, and 6.80. The S. aureus concentrations at Plant A and Plant B ranged from 1 to 3 CFU. Both Factory A (2–26 CFU/mL) and Factory B (11–316 CFU/mL) provided water samples that also included total coliforms.

3.2. Probabilistic Distribution

The pH of the water and the concentrations of S. aureus and total coliforms in the water samples of Plants A and B were simulated to have probability distribution values of Normal(1.5, 1.0) and Normal(1, 0) and Normal(163.09, 89.10) and Normal(13.97, 6.70), respectively (Figure 1). The water samples from Plants A and B had normal pH distributions of 6.75, 0.61, and 7.20, 0.38, respectively. The concentrations of S. aureus and total coliforms in the water samples of Factory A and B were in the range of 0–4 CFU and 0–2 CFU and 0–316 CFU and 0–30 CFU, respectively, with the highest probabilities of S. aureus and coliform contamination in the water of Plant A and Plant B being 0.42 and 0.35 and 4.47 × 10⁻³ and 0.0514, respectively. The pH values for the water of Plants A and B had the maximum probabilities of 0.6 at pH ranging from 5 to 8 and 1 at pH ranging from 6 to 8, respectively.

3.3. Modeling with Multiple Regression Equations

The sensitivity analysis of the water samples was conducted using multiple regression equations, and the equations for the multiple regression analysis of S. aureus contamination in water (Equation (3)) and S. aureus contamination in bottled drinking water (Equation (4)) are as follows:

Y = −233.9910 + (1.8471 × Time) − (9.1343 × Temperature of Water) − (5.0746 × Residual Chlorine in Water) + (40.5970 × pH of Water)

(3)

Y = −0.1260 + (0.1972 × pH of Water)

(4)

The model’s input parameters included the water’s temperature, pH, and residual chlorine content. The model was designed for the detection of S. aureus water contamination.

3.4. Sensitivity Analysis

The Excel Platform was used to carry out the sensitivity analysis. For each input variable, horizontal bars on a tornado chart were drawn, with the length of the bars indicating the strength of the association with the output variable.

The sensitivity analysis results (Figure 2) show that the temperature of the water (−0.0717) and residual chlorine concentration (−0.0109) were the conditional parameters that led to a decrease in the amount of S. aureus in the water and an increase in the concentration of S. aureus in the water were processing time (0.0626) and pH of water (0.0571). Water temperature and residual chlorine concentration were the main intervention measures that helped control the effect on the water.

4. Discussion

To assess the health risks associated with the drinking water in our study, the study of Wibuloutai and coleagues, which assessed the microbiological quality of drinking water produced by drinking water filtration dispenser toll machines (DFTMs), was compared to this study. In total, 210 samples were chosen at random from 70 DFTMs that were spaced 500 m apart at Mahasarakham University. The DFTM water had high levels of Escherichia coli and Staphylococcus aureus, with prevalence rates of 54.17% and 16.67% and average values of 1.04 log cfu/mL and 0.26 log cfu/mL, respectively. The risk assessment involved the use of the @Risk tool to calculate the likelihood of exposure, which was deemed to be 1.67 × 10⁻¹, and the odds of getting sick from S. aureus and E. coli were found to be 2.08 × 10⁻³ and 1.58, respectively. @Risk is a software tool used to generate a distribution of possible outcomes and estimate the probability of different levels of pathogen exposure [4]. The findings of our investigation indicate that the water we studied may be safe for consumption. It is suggested to keep the water’s pH between 6.5 and 7.5, as this could help to lessen the level of microbial contamination in the water [13]. According to Pratum and Khananthai (2017) [14], the quality of drinking water also depends on the tap water or water source used throughout the manufacturing process. Local governments must issue local laws for the control of such processes in accordance with the Ministry of Public Health’s guidelines on hazardous health activities [4].

5. Conclusions

This study was successful in identifying the potential risk of S. aureus contamination in water as well as the recommended pH levels for processing bottled drinking water in small- and medium-sized facilities. The findings of the sensitivity study revealed that the pH of the water samples played a significant role in the procedure. This result implies that preserving the pH of the water throughout the procedure, as shown by this study, is a helpful technique for lowering the risk of contamination. However, harmful bacteria have the potential to taint finished goods if the water quality is inadequate.