Prognostic models for chronic kidney disease: a systematic review and external validation.

Area of Science:

• Nephrology
• Epidemiology
• Biostatistics

Background:

• Accurate risk prediction is crucial for personalized healthcare in chronic kidney disease (CKD) patients.
• Numerous prognosis studies are published, investigating biomarkers and risk prediction.
• A systematic appraisal of multivariable prognosis studies in CKD is needed.

Purpose of the Study:

• To systematically appraise published prognosis studies investigating multiple biomarkers in CKD.
• To assess the quality of reported prognostic models for CKD.
• To externally validate these prognostic models in a broad CKD population.

Main Methods:

• Systematic review and quality appraisal of studies reporting multivariable prognosis models for end-stage renal disease (ESRD), cardiovascular (CV) events, and mortality in CKD.
• External validation of identified models using data from the Multifactorial Approach and Superior Treatment Efficacy in Renal Patients with the Aid of Nurse Practitioners trial.
• Analysis of model performance, including calibration and variability.

Main Results:

• Identified 91 papers describing 36 models for ESRD, 50 for CV events, 46 for mortality, and 17 for composite outcomes.
• Most studies were of moderate quality, with inconsistent outcome definitions and incomplete model reporting (21% reported full equations).
• External validation showed variable performance, with adequate calibration for most models except for mortality prediction.

Conclusions:

• An abundance of multivariable prognosis models exist for the CKD population.
• Moderate study quality and reporting deficiencies limit the direct clinical application of these models.
• Further research is needed to develop high-quality, well-reported prognostic models for CKD patients.