Evaluation of the nephroprotective strategy effectiveness in the late stages of chronic kidney disease

Background. Classical nephroprotection reduces its effectiveness at the late CKD stages; the search for effective algorithms is hampered by accelerating decline in GFR, therefore there are no generally accepted ways to evaluate the effectiveness.

Aim: to build a model for predicting the GFR decline rate in order to assess the effectiveness of the intensive follow-up.

Methods. A representative group of regular follow-up (N=540) was allocated from the city database (N=7696) to built-up the polynomial model that predicts GFR annual decline. We used the model to evaluate the intensive monitoring effectiveness (N=100) by the difference between predicted and actual rates of GFR decline. We also selected well matched subgroup of 200 patients for direct comparison of hard and surrogate outcomes.

Results. During last year before need in dialysis, the rate of GFR decline in intensive group was 5.98±1.69 vs. the predicted 9.06±0.59ml/min/1.73 m²/year. We used that assessment of the intervention effectiveness as dependent variable in regression and categorical analysis. The significant components of the nephroprotection: phosphatemia decrease (0.25 mmol/l), hemoglobin increase (1 g/dl), effective administration of RAAS blockers (to reduce proteinuria by 0.1 g/l), systolic blood pressure decrease (5 mmHg), calcemia deviations decrease from the target (0.1 mmol/l), acidosis correction (2 mmol/l), inflammation reduction and albumin increase (1.5 g/l) -were associated with the smaller GFR decrease rate by 15%. In intensive group, the dialysis risk was 2.2 times lower, the death risk was 4 times. The only planned dialysis start was ensured in intensive group, 67% chose peritoneal dialysis.

Conclusions. The prediction of GFR decline rate calculated by nonlinear model in comparison with the actual one can evaluate the nephroprotection effectiveness; it differs significantly from the classical ones at the CKD late stages.