Errors in Computing the Normalized Protein Catabolic Rate due to Use of Single-pool Urea Kinetic Modeling or to Omission of the Residual Kidney Urea Clearance

Validation of formulas calculating normalized protein catabolic rate in patients undergoing home hemodialysis

Depner and Daugirdas developed a simplified formula to estimate the normalized protein catabolic rate in patients on twice- or thrice-weekly hemodialysis (JASN, 1996). The aim of our work was to establish formulas in more frequent schedules and validate them in home-based hemodialysis patients. We realized that the structure of Depner and Daugirdas' normalized protein catabolic rate formulas has a general meaning and can be expressed as PCRn = C0/[a + b*(Kt/V) + c/(Kt/V)] + d, where C0 is pre-dialysis blood urea nitrogen, Kt/V is dialysis dose, a, b, c, d are the specific coefficients for each combination of home-based hemodialysis schedules and the day of blood sampling. The same applies to the formula that adjusts C0 (C'0) for residual kidney clearance of blood water urea (Kru) and urea distribution volume (V): C'0 = C0*[1 + (a1 + b1/(Kt/V))*Kru/V]. On this basis, we computed the six coefficients (a, b, c, d, a1, b1) for each of the 50 possible combinations and simulated a total of 24,000 weekly dialysis cycles using the Daugirdas Solute Solver software recommended by the KDOQI 2015 guidelines. From the associated statistical analyses we obtained 50 sets of coefficient values, which were validated comparing the paired normalized protein catabolic rate values (i.e., those estimated with our formulas with those modeled with Solute Solver) in 210 datasets of 27 patients on home-based hemodialysis. The mean values ± SD were 1.06 ± 0.262 and 1.07 ± 0.283 g/kg/day, respectively, with a mean difference of 0.004 ± 0.034 g/kg/day (p = 0.11). The paired values were highly correlated (R2 = 0.99). In conclusion, even if the coefficient values were validated in a relatively small sample of patients, they allow an accurate estimation of normalized protein catabolic rate in home-based hemodialysis patients.

Longitudinal Causal Effects of Normalized Protein Catabolic Rate on All-Cause Mortality in Patients With End-Stage Renal Disease: Adjusting for Time-Varying Confounders Using the G-Estimation Method

In this study, we aimed to estimate the causal effect of normalized protein catabolic rate (nPCR) on mortality among end-stage renal disease (ESRD) patients in the presence of time-varying confounding affected by prior exposure using g-estimation. Information about 553 ESRD patients was retrospectively collected over an 8-year period (2011-2019) from hemodialysis facilities in Kerman, Iran. nPCR was dichotomized as <1.2 g/kg/day versus ≥1.2 g/kg/day. Then a standard time-varying accelerated failure time (AFT) Weibull model was built, and results were compared with those generated by g-estimation. After appropriate adjustment for time-varying confounders, weighted g-estimation yielded 78% shorter survival time (95% confidence interval (95% CI): -81, -73) among patients with a continuous nPCR <1.2 g/kg/day than among those who had nPCR ≥1.2 g/kg/day during follow-up, though it was 18% (95% CI: -57, 54) in the Weibull model. Moreover, hazard ratio estimates of 4.56 (95% CI: 3.69, 5.37) and 1.20 (95% CI: 0.66, 2.17) were obtained via weighted g-estimation and the Weibull model, respectively. G-estimation indicated that inadequate dietary protein intake characterized by nPCR increases all-cause mortality among ESRD patients, but the Weibull model provided an effect estimate that was substantially biased toward the null.

Routine assessment of kidney urea clearance, dialysis dose and protein catabolic rate in the once-weekly haemodialysis regimen

BACKGROUND

The dialysis dose (Kt/V) and normalized protein catabolic rate (PCRn) are the most useful indices derived from the urea kinetic model (UKM) in haemodialysis (HD) patients. The kidney urea clearance (Kru) is another important UKM parameter which plays a key role in the prescription of incremental HD. Ideally, the three kinetic parameters should be assessed using the complex software Solute Solver based on the double pool UKM. In the clinical setting, however, the three indices are estimated with simplified formulae. The recently introduced software SPEEDY assembles the aforementioned equations in a plain spreadsheet, to produce quite accurate results of Kru, Kt/V and PCRn. Unfortunately, specific equations to compute Kt/V and PCRn for patients on a once-weekly HD regimen (1HD/wk) were not available at the time SPEEDY was built-up. We devised a new version of SPEEDY (SPEEDY-1) and an even simpler variant (SPEEDY-1S), using two recently published equations for the 1HD/wk schedule . Moreover, we also added a published equation to estimate the equivalent renal clearance (EKR) normalized to urea distribution volume (V) of 35 L (EKR₃₅) from Kru and Kt/V . Aim of the present study was to compare the results obtained using the new methods (SPEEDY-1 and SPEEDY-1S) with those provided by the reference method Solute Solver.

SUBJECTS AND METHODS

One hundred historical patients being treated with the once-weekly HD regimen were enrolled. A total of 500 HD sessions associated to the availability of monthly UKM studies were analysed in order to obtain Kru, single pool Kt/V (spKt/V), equilibrated Kt/V (eKt/V), V, PCRn and EKR₃₅ values by using Solute Solver, SPEEDY-1 and SPEEDY-1S.

RESULTS

When comparing the paired values of the above UKM parameters, as computed by SPEEDY-1 and Solute Solver, respectively, all differences but one were statistically significant at the one-sample t-test; however, the agreement limits at Bland-Altman analysis showed that all differences were negligible. When comparing the paired values of the above UKM parameters, as computed by SPEEDY-1S and Solute Solver, respectively, all differences were statistically significant; however, the agreement limits showed that the differences were negligible as far as Kru, spKt/V and eKt/V are concerned, though much larger regarding V, PCRn and EKR₃₅.

CONCLUSIONS

We implemented SPEEDY with a new version specific for the once-weekly HD regimen, SPEEDY-1. It provides accurate results and is presently the best alternative to Solute Solver. Using SPEEDY-1S led to a larger difference in PCRn and EKR₃₅, which could be acceptable for clinical practice if SPEEDY-1 is not available.

The lacking equation that estimates the protein catabolic rate in patients on once-weekly haemodialysis

BACKGROUND

The normalized protein catabolic rate (PCRn) is one of the key indices derived from the urea kinetic model (UKM) in haemodialysis (HD) patients. Ideally, it should be assessed using the double pool UKM (KDOQI clinical practice guidelines, AIKD, 2015), as the web-based software Solute-Solver (SS) does (Daugirdas et al., AJKD, 2009). Simple formulae exist to compute PCRn for patients on thrice- or twice-weekly HD schedule, but not for patients on once-weekly HD schedule (1HD/wk). Aim of the present technical note was to introduce the lacking equation that estimates PCRn in the 1HD/wk regimen.

METHODS

Data of a single HD session associated to monthly UKM studies were retrieved from the electronic database of our dialysis unit for 80 historical patients on 1HD/wk regimen. The UKM parameters, as calculated with SS, were used in a subgroup of 40 randomly selected patients (group 1) to build-up a multiple regression model of PCRn. The latter was used to predict PCRn (PCRn_Pred) values in the cohort of the remaining 40 patients (group 2). The Bland-Altman plot was used to analyse the agreement between PCRn_Pred and the paired "observed" (PCRn_Obs) values, as measured with SS.

RESULTS

The following equation was established by means of the multiple regression analysis: PCRn = - 0.46 + 0.01 × C0 + 0.09 × eKt/V + 3.94 × Kru/V, where C0 is pre-dialysis blood urea nitrogen concentration, eKt/V is the equilibrated Kt/V, Kru is the residual renal urea clearance and V is the post-dialysis urea distribution volume. The PCRn_Pred values were 0.99 ± 0.24 g/kg/day; the PCRn_Obs values were 0.96 ± 0.23 g/kg/day (mean difference 0.03 ± 0.05 g/kg/day). Their difference at the Bland-Altman analysis ranged from - 0.08 to + 0.13 g/kg/day. Finally, a nomogram was drawn: it can be used to estimate not only PCRn from Kru/V and C0, but also C0 as a function of Kru/V and PCRn.

CONCLUSIONS

The equation here introduced allows a simple and accurate estimate of PCRn in patients on once-weekly HD regimen. The availability of the nomogram relating C0 to PCRn and Kru/V could be a further step to make safer and safer the once-weekly HD regimen. The following equation was established by means of the multiple regression analysis [Formula: see text] where PCRn is the normalized protein catabolic rate (PCRn), C0 is pre-dialysis blood urea nitrogen concentration (BUN), eKt/V is the equilibrated Kt/V, Kru is the residual renal urea clearance and V is the post-dialysis urea distribution volume. A nomogram relating pre-dialysis BUN to PCRn and Kru/V could be drawn: it can be used to estimate not only PCRn from Kru/V and pre-dialysis BUN, but also pre-dialysis BUN as a function of Kru/V and PCRn.

Equations to Estimate the Normalized Creatinine Generation Rate (CGRn) in 3/Week Dialysis Patients With or Without Residual Kidney Function

OBJECTIVE

Normalized creatinine generation rate (CGRn) can be computed for a variety of dialysis schedules using a recently described kinetic modeling program. However, the availability of estimating equations might facilitate broader study of this metric. We developed equations to estimate CGRn based on modeling and then tested them against modeled CGRn values in the Frequent Hemodialysis Network Nocturnal Trial baseline (3/week) dataset.

DESIGN AND METHODS

We used a "what-if" derivation of a previously published variable volume 2-pool creatinine kinetic model to generate predicted predialysis values of serum creatinine that would result from creatinine generation rates of 250-2000 mg/day in patients with creatinine distribution volumes of 20 to 50 L, dialyzed from 60 to 480 min per treatment three times a week. Then, in patients with residual kidney function, we calculated an "anuric expected predialysis serum creatinine value" before applying the same equations. We then compared estimated CGRn values as predicted by this approach with modeled values in patient data from the Frequent Hemodialysis Network Nocturnal Trial.

RESULTS

The estimating equations for CGRn yielded results similar to those obtained with formal modeling, in both anuric patients and those with residual kidney function, with mean percent error of 0.845 ± 6.15 (SD) in anuric patients, and ‒0.29 ± 4.90 in patients with a mean creatinine clearance of 5.44 ± 4.82 mL/min, with R-squared values of 0.96 in both anuric patients and those with residual renal clearance of creatinine.

CONCLUSIONS

In patients dialyzed 3/week, CGRn can be estimated using prediction equations. Use of these equations may facilitate broader investigation of CGRn as a measure of nutritional status and outcome.