Fig 1.
Conceptual framework of the quantitative microbial risk assessment model.
Framework includes the computational process, parameters, and variables included in estimating risk of infection for Cryptosporidium or for Giardia from contaminated drinking water. The driving input parameter (raw concentration of parasites) is identified at the left most section of the figure. Boxes represent modeled variables (e.g., modified concentration and load of parasites), ovals represent exogenous parameters and assumptions affecting modeled variables (e.g. HUM, MST, and ENV scenarios represent the fraction of parasites belonging to a species infectious-to-humans), and the dashed box on the right represents the model output. Units are displayed within brackets under each description. Variable and parameter symbols shown directly above their respective variable or parameter are used in Eq 1, 2 and 4.
Fig 2.
Conceptual framework of the susceptible-infected-recovered model.
Framework includes the steps, parameters, and variables included in estimating diarrheic infections of Cryptosporidium and Giardia from contaminated drinking water for each individual (i) in the study population. Boxes represent modeled variables (e.g. state of child) and ovals represent exogenous parameters and assumptions affecting modeled variables, i.e. P(i), P(s), and D. Units are displayed within brackets under each description.
Table 1.
Main parameters used in the quantitative microbial risk assessment and susceptible-infected-recovered models developed to simulate waterborne diarrheal infections in children < 5 in Puri District due to Cryptosporidium and Giardia in drinking water from deep (DTW) and shallow (STW) tubewells over two monsoon seasons.
Table 2.
Concentrations of Cryptosporidium oocysts and Giardia cysts detected in tubewells during the 2012 and 2013 monsoon season.
Table 3.
Animal host mean prevalence rates (Zi) of shedding zoonotic species infectious-to-humans (i.e., C. parvum or G. lamblia assemblage A or B) estimated from published literature and used in Eq 3 to calculate the overall zoonotic fraction ( in Table 1) of Cryptosporidium and Giardia parasites shed by animals in Puri District able to infect humans.
Fig 3.
Boxplots of daily risk of Cryptosporidium and Giardia infection.
Daily risk is for an individual child drinking deep and shallow tubewell water in 2012 (A & B) and 2013 (C & D), accounting for boiling rates (9% and 16% by deep and shallow tubewell users respectively), under different scenarios representing the fraction of parasites infectious-to-humans (HUM = H, MST = M, ENV = E). US EPA tolerable daily risk of infection from drinking water is denoted with the dashed line at log10 (0.0001%).
Fig 4.
Boxplots of simulated diarrhea daily point prevalence (% of children).
Prevalence is shown for the study population during the 90-day monsoon season estimated from 1,000 simulations of the quantitative microbial risk assessment and susceptible-infected-recovered models. Results shown for deep (DTW) and shallow (STW) tubewell users (either infection), and for Cryptosporidium (Crypto.) and Giardia (Gia.) infections (either tubewell type) for the monsoon season in 2012 and 2013, under different scenarios used to assign the fraction of parasites infectious-to-humans (HUM = H: orange, MST = M: green, ENV = E: blue). Combined DTW & STW results are for an average child in a tubewell using household, accounting for rates of deep (48%) and shallow (52%) tubewell usage in the study population. Boiling rates (9% and 16% for deep and shallow tubewell users respectively) are accounted for in all results. Different vertical scales for 2012 and 2013 plots highlight the upper limits of burden between years.
Table 4.
Fraction of the observed 7-day recall child diarrhea period prevalence in the Odisha Sanitation Trial (12% in 2012 and 9% in 2013) estimated to be attributable to drinking tubewell water contaminated with Cryptosporidium and Giardia under different human pathogenicity scenarios examined (median % and interquartile range (IQR)).
Fig 5.
Sensitivity analysis results for Cryptosporidium and Giardia.
Sensitivity analysis for Cryptosporidium (purple) and Giardia (green) averaged over year and tubewell type showing the sensitivity index of 7-day recall diarrhea period prevalence to each parameter, where a sensitivity index closer to one indicates greater model sensitivity to that parameter.