Roxadustat raised haemoglobin more than erythropoietin-type drugs in peritoneal dialysis, but evidence is low certainty

Design

Systematic review and meta-analysis of randomised studies

Population

19 studies, 1,636 adults on peritoneal dialysis with renal anaemia

Primary outcome

Post-treatment haemoglobin, roxadustat vs erythropoiesis-stimulating agents

Effect

Mean difference 7.30 g/L (95% CI 5.43 to 9.16); adverse event RR 0.45 (95% CI 0.27 to 0.74), definitions inconsistent

This systematic review and meta-analysis searched to December 2025 and pooled 19 randomised studies with 1,636 adults on peritoneal dialysis, comparing roxadustat-based regimens with erythropoiesis-stimulating agent regimens.

Average post-treatment haemoglobin was 7.30 g/L higher with roxadustat (95% CI 5.43 to 9.16; I2 = 62.6%). Target attainment was not defined consistently enough to pool. The pooled adverse event rate favoured roxadustat (RR 0.45, 95% CI 0.27 to 0.74), but definitions and ascertainment varied. Certainty of evidence was low for haemoglobin and low to very low for most secondary outcomes. The authors conclude the evidence is insufficient to establish clinical superiority, because final values were pooled and comparator regimens varied, and that long-term safety is not established.

A gain of 7 g/L is modest and could vary with how the comparator was dosed. For prescribers it supports roxadustat as a reasonable oral option in peritoneal dialysis, with attention to cardiovascular and thrombotic risk.

• Haemoglobin was 7.30 g/L higher with roxadustat (95% CI 5.43 to 9.16) across 19 trials of 1,636 patients.
• Heterogeneity was moderate to high (I2 62.6%) and certainty of evidence low.
• The adverse event difference (RR 0.45) rested on inconsistent definitions.
• Long-term safety, including cardiovascular risk, remains unproven.
• Consider roxadustat as an oral option rather than a proven superior drug.

Why it matters

Oral therapy is attractive in home dialysis, but a modest haemoglobin gain does not on its own show better outcomes.

Don't overread it

The pooled data are low certainty, and the authors themselves say they cannot establish clinical superiority or long-term safety.

The statistics, in plain English

A haemoglobin difference of 7 g/L equals 0.7 g/dL. The pooled studies compared final values against different comparator regimens, so the difference partly reflects how aggressively each comparator was dosed. I2 of 62.6% means the trials differ appreciably from one another.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.