Study reveals need for better tools to guide kidney failure patients

A systematic review of mortality prediction models for kidney failure patients identified limitations in current tools used to inform treatment decisions. The study, published in JAMA Network Open, analyzed 50 prediction models covering more than 2.9 million patients.

Researchers found that existing models fall short of helping patients and healthcare providers make informed decisions about treatment options. The study identified that all current prediction models focused exclusively on patients already receiving dialysis treatment, leaving a critical gap for those still weighing their care options.

The research team screened 7,184 clinical study abstracts, ultimately including 50 studies that met their criteria. The patient population represented diverse age groups, with a median age of 64 years (ranging from 52 to 81 years), and included both men and women, with women making up 42% of participants.

The systematic review employed rigorous methodology to evaluate each model’s effectiveness. Researchers assessed studies for risk of bias in four key domains: study participants, predictors, outcomes and analysis methods. They found concerning trends across multiple areas, with more than half of the studies showing high risk of bias due to inadequate selection of study populations and significant shortcomings in measurement methods.

In addition, none of the existing models included patients who had chosen conservative care without dialysis, an option often considered by older adults with kidney failure. This gap in the research is especially significant given that people with kidney failure have higher morbidity and mortality than those with non-metastatic cancer and often face difficult treatment decisions.

“According to this systematic review of 50 studies, published mortality prediction models were at high risk of bias and had applicability concerns for clinical practice,” the authors noted. They found that current models showed limitations in several key areas, including inadequate selection of study populations, shortcomings in measurement methods, and flaws in analysis strategies.

The researchers conclude that new mortality prediction models could better support treatment decisions in people with kidney failure, especially those who have not yet started dialysis treatment.