Dialysis is a preference-sensitive decision where prognosis may play an important role. Although patients desire risk prediction, nephrologists are wary of sharing this information. They reviewed the performance of prognostic indices for patients starting dialysis to facilitate bedside translation.

However, several prognostic models performed well in predicting survival for patients starting dialysis; and could be used to inform decisions of whether to pursue the treatment; according to a systematic review and meta-analysis of published studies. “Dialysis is a lifesaving therapy for patients with end-stage kidney disease; but its utility is increasingly question for frail and multimorbid patients who experience rapid functional decline and have high symptom burden,” Ryan T. Anderson, MD,of Mayo Clinic School of Graduate Medical Education, and colleagues wrote.

Counseling patients with ESKD

“Ideally, clinicians counseling patients with ESKD would present patients with all treatment options; and likely outcomes before dialysis initiation and use shared decision-making focused on patients’ goals; and preferences. Prognostic information is an essential component of shared decision-making; and dialysis patients desire this information.”

But they excluded studies limited to prevalent dialysis patients; AKI and studies excluding mortality in the first 1–3 months. However, two reviewers independently screened abstracts; performed full text assessment of inclusion criteria and extracted: study design, setting, population demographics; index performance and risk of bias.

Pre-planned random effects meta-analysis performed stratified by index; and predictive window to reduce heterogeneity. To assess validated prognostic indices that predict mortality, researchers considered 36 studies that described the accuracy of 32 such tools in predicting outcomes of patient death (predictive windows ranged from 3 months to 10 years; cohort sizes ranged from 46 to 52,796).

Dialysis decision making

Therefore, researchers found discrimination area under the curve (AUC) was 0.71 with high heterogeneity. However, the Charlson comorbidity index was the most commonly use and show the highest AUC, along with the Obi and Ivory (AUC for all 0.74). The Kahn-Wright index (AUC = 0.68), Hemmelgarn modification of the Charlson comorbidity index (AUC = 0.66) and REINindex (AUC = 0.69) also frequently used.

Researchers noted the risk predictors varied between the indices with most including age and none including socioeconomic status. “Prognosis is a key element of shared decision making, yet poorly integrated into dialysis decision making,” the researchers wrote.

“Several validated indices with good prognostic ability are available to predict mortality at the start of dialysis. Little attention has been paid to bedside translation of these indices to determine the value of sharing prognostic information with patients. Further research is needed to understand how best to utilize these indices in patient care to support shared decision-making.”