LAUSR

The article by Winnicki et al.1 and the related editorial by Larkins and Craig2 raises important questions about indiscriminately postponing dialysis in children in every clinical setting. Multiple studies have shown that the global variation in eGFR at dialysis initiation in children is large due to multiple reasons: lack of evidence base, complexity of ESKD in the pediatric population, and temporal trends following clinical guidelines recommendations. However, the current median eGFR at dialysis initiation is 7.8, 8.1, and 8.2 ml/min per 1.73 m in the United States, Canada, and Europe, respectively (all below the cutoff limit of 10 ml/min per 1.73 m).3 The only published randomized controlled trial (RCT) studying the effect of initiatingdialysis early (10–15ml/minper 1.73m) versus late (5–7 ml/min per 1.73 m)4 in adults revealed no differences in all-cause mortality, economics, or quality of life. Although the results of this RCT could not be entirely generalizable to children, we could conclude that, among patients with asymptomatic ESKD, initiation of dialysis can be safely postponed. Recently, the European Society of Paediatric Nephrology/ European Renal Association and European Dialysis and Transplant Association (ESPN/ERA-EDTA) Registry5 examined the same question in 2963 patients from 21 European countries using a eGFR cutoff value of 8 ml/min per 1.73 m (representing more accurately the current practice worldwide and following clinical practice guidelines as per Kidney Disease Outcomes Quality Initiative 2015, National Institute for Health and Care Excellence 2018, and Kidney Disease Improving GlobalOutcomes 2019).3 Therewas no difference in terms ofmorbidity, mortality, growth, and access to transplantation, except a higher risk of developing hypertension among late starters (64% versus 51%among early starters). However, it is important tomention that the International Pediatric Peritoneal Dialysis Network found in 1001 children and adolescents on peritoneal dialysis that late dialysis start was associated with underweight (11% and 5%inchildren starting peritoneal dialysis at an eGFRof,6ml/min per 1.73mandaneGFRof 9–12ml/minper 1.73m, respectively), which may favor early start in that setting. Last but not least, lack of clinical information (renal residual function, clinical indications for starting dialysis, uremic symptoms), and statistical inborn errors of registry and observational studies (lead time, indication and survivor biases) might lead to overor underestimation of the outcomes. Until a RCT is conducted in children with the current data, and knowing the complexity of ESKD in children, we should only advise that eGFR is a poor indicator of deciding whether to start dialysis if this is used as the sole/main criterion. Even if recent data have demonstrated an association between early dialysis start (eGFR.10 ml/min per 1.73 m) and mortality, clinicians might need to “think twice” about delaying dialysis if quality of life, growth, or uremic/hypertensive symptoms indicate the opposite.