The most recent literature has shown extensively that a low-protein diet in patients with chronic kidney disease (CKD) delays the natural progression of end-stage renal disease and the necessary treatment of chronic dialysis. The aim of this study was to estimate the cost-effectiveness of a very low-protein diet compared with a moderate low-protein diet.


Markov model.


The study was conducted from the perspective of the Italian National Health Service.


Patients with CKD stages 4 and 5. Seven hundred and two patients were considered, and half were treated with a very low-protein diet and the other half with a moderate low-protein diet.


The Markov model estimates the costs and the quality-adjusted life years (QALYs) associated with a very low-protein diet versus a moderate low-protein diet for patients with CKD Stages 4 and 5. The transition probability was estimated on data from 7 studies, which determined the efficacy of very low-protein diets in delaying the need to start maintenance dialysis. Utilities and cost were estimated from literature review and projected for the lifespan considered in the model. The annual cost of dialysis per patient was approximately €34,072. The cost of a very low-protein diet was €1,440 per patient per year in the Lazio region (conservative assumptions). Probabilistic and deterministic sensitivity analyses were performed.

Main Outcome Measure

QALY, cost of treatment, incremental cost-effectiveness ratio.


The model estimates that a very low-protein diet should be more effective. Dietary treatment improved 0.09 QALYs after 2 years, 0.16 after 3 years, 0.36 after 5 years, and up to 0.93 incremental QALYs after the first 10 years. After 2 years, the model estimated increment cost in favor of the most restrictive dietary treatment of €826, €2,146, €5,203, and €10,375 for 2, 3, 5, and 10 years of follow-up respectively.


The results of these simulations indicate that the treatment of CKD patients with a very low-protein diet is cost-effective relative to a moderate low-protein diet in an Italian setting. Further studies should test this model in other countries with different dialysis costs and dietary support.