Fig 1.
General structure of the modified model of transport of carbon dioxide and oxygen in the human body, summarizing the input and outputs of the model compartments.
The state variables are reported for each of the four main compartments, and their value is calculated by solving the blood chemistry system of equations using the inputs for each compartment. See Appendix D in S1 Text for the explanation of the symbols.
Table 1.
Patient-specific and treatment-specific parameters for the two clinical studies [26,31] to which the model was applied.
The values used here are the average of the data reported. For patient-specific parameters, initial values are reported unless otherwise specified. Measurements missing in the datasets were taken from the original description of the model or assumed based on standard values for the patient population.
Table 2.
Parameters estimated by the model.
The values estimated to fit the data from the studies by Sargent [26] and Park [31] are compared with the values assumed in the original model description by Andreassen and Rees [30] (only for the steady-state parameters and
). The dialysances DBic and DCO2 are shown first for the simulations with constant respiration rate and then when a linear increase in minute ventilation (
) was assumed.
Fig 2.
Simulations of several variables of the model, fitted to the data from the study by Sargent et al [26].
The dialysate concentration of bicarbonate was corrected for the Gibs-Donnan effect.
Fig 3.
Simulations of several variables of the model, fitted to the data from the study by Park et al [31].
The dialysate concentration of bicarbonate was corrected for the Gibs-Donnan effect.
Fig 4.
Simulations with a linearly increasing minute ventilation during the HD session.
The increase was up to 11% more than the initial value for Sargent data (top panels) and 22% for Park data (bottom panels). The dialysate concentration of bicarbonate was corrected for the Gibs-Donnan effect.
Fig 5.
Simulations of a week-long cycle of HD sessions.
Two postdialytic intervals were of 48 hours and one was of 72 hours, during which fluid volumes accumulates and bicarbonate buffer is depleted due to an endogenous acid production rate of 0.039 mmol/day per Kg of body mass. The dialysate concentration of bicarbonate was corrected for the Gibs-Donnan effect.
Table 3.
Dialysances calculated from measured inlet and outlet dialyzer concentrations in published studies.
Fig 6.
(Top) Bicarbonate and dissolved CO2 concentrations in Sargent data (assuming constant respiration rate), with dialysate concentration of bicarbonate corrected for the Gibs-Donnan effect. The simulated session was extended to 420 minutes to better show the leveling off of bicarbonate concentration profile, and the effects of hemoconcentration were removed by setting the value of UF to zero. Note that if acetate concentration is not set to zero (dashed line), plasma bicarbonate can exceed dialysate concentration because of the constant inflow from acetate, which in the model is independent of other variables. (Bottom) Solute flow rates from dialysis fluid to plasma.