Related Experiment Videos
A population-specific formula predicting creatinine excretion in continuous peritoneal dialysis
Antonios H Tzamaloukas1, Glen H Murata
1Renal Section, New Mexico Veterans Affairs Health Care System, and University of New Mexico School of Medicine, Albuquerque 87108, USA. Tzamaloukas.Antonios@albuquerque.va.gov
Insights
A new formula accurately predicts creatinine excretion in continuous peritoneal dialysis (CPD) patients, outperforming the Cockcroft-Gault equation. This population-specific model improves estimations for creatinine excretion studies in CPD.
Area of Science:
- Nephrology
- Clinical Chemistry
Background:
- The Cockcroft-Gault formula systematically overestimates creatinine excretion decline with age in continuous peritoneal dialysis (CPD) patients.
- This inaccuracy makes the formula unsuitable for research involving creatinine excretion in CPD.
Purpose of the Study:
- To develop and validate a population-specific formula for predicting average creatinine excretion in CPD patients.
- To establish a more accurate method for assessing creatinine excretion compared to existing formulas.
Main Methods:
- Creatinine excretion was measured in 885 CPD patients, randomly divided into derivation (n=432) and validation (n=453) groups.
- Stepwise multiple linear regression was used to develop a predictive formula based on weight, age, gender, and diabetes in the derivation group.
- The derived formula (CrExcr1) and the Cockcroft-Gault formula (CrExcr2) were compared against measured creatinine excretion (CrExcr) in the validation group.
Main Results:
- The best-fit regression model included weight, age, gender, and diabetes: CrExcr1 = 302.150 - 4.380A + 171.234G - 39.041D + 11.730W (r²=0.477).
- In the validation set, CrExcr1 showed higher accuracy (r²=0.447) than CrExcr2 (r²=0.340) when predicting creatinine excretion.
- The population-specific formula (CrExcr1) demonstrated significantly smaller prediction errors compared to the Cockcroft-Gault formula.
Conclusions:
- A novel, population-specific formula for predicting creatinine excretion in CPD patients was successfully derived.
- This new formula offers superior accuracy to the Cockcroft-Gault formula for use in CPD studies.
- The developed formula can be reliably utilized in research focusing on creatinine excretion within the CPD population.
Objective:
The Cockroft-Gault formula was shown to systematically overestimate the decline in creatinine excretion with age in continuous peritoneal dialysis (CPD) patients and is, therefore, not suitable for studying creatinine excretion. The purpose of the present study was to develop and test a population-specific formula predicting average creatinine excretion in CPD.
Methods:
Creatinine excretion in urine plus dialysate was measured in 925 CPD patients. Forty patients were excluded because of evidence of noncompliance. The remaining 885 subjects were randomly grouped into a derivation group (n = 432) and a validation group (n = 453). Stepwise multiple linear regression models were used to predict creatinine excretion in the derivation group. The candidate variables, chosen because they were previously shown to be predictors of creatinine excretion in CPD, included weight (W), age (A), gender (G), diabetes (D), and interaction terms between these four variables. Estimates of creatinine excretion from the best-fit regression formula (CrExcr1) and from the Cockroft-Gault formula (CrExcr2) were compared to creatinine excretion (CrExcr) in the validation group.
Results:
The best-fit regression model in the derivation group included all four candidate variables (W, A, G, D), but no interaction terms. This model was as follows: CrExcr1 = 302.150 - 4.380A + 171.234G - 39.041D + 11.730W (r2 = 0.477, p < 0.001). In the validation set, CrExcr = -15.795 + 0.988CrExcr1 (r2 = 0.447, p < 0.001), and CrExcr = -303.823 + 0.732CrExcr2 (r2 = 0.340, p < 0.001). When the differences between measured and predicted creatinine excretion did not take into account the sign of each individual difference, CrExcr - CrExcr1 = 201 +/- 156 mg/24 hours, and CrExcr - CrExcr2 = 235 +/- 174 mg/24 hr (p < 0.001) in the validation group. When the sign of the difference was taken into account, CrExcr - CrExcr1 = -28 +/- 149 mg/24 hr, and CrExcr - CrExcr2 = 63 +/- 295 mg/24 hr (p < 0.001).
Conclusions:
A population-specific formula predicting creatinine excretion in CPD was derived. This formula has greater accuracy than the Cockroft-Gault formula and can be used in studies of creatinine excretion in CPD.