The relative efficacy of antibiotic classes for pyelonephritis and complicated UTI (cUTI) remains unclear, as different antibiotic classes vary in antimicrobial activity, pharmacodynamics and pharmacokinetics.

Systematic review and meta-analysis. We searched PubMed, Cochrane CENTRAL, clinical-trial registries and conference abstracts through 30 November 2025. Randomised controlled trials comparing antibiotic classes in adults with pyelonephritis/complicated UTI (cUTI) were eligible (PROSPERO CRD42023448020). Five comparisons were analysed: quinolones vs β-lactams, cephalosporins vs penicillins, aminoglycosides vs quinolones or beta-lactams, trimethoprim-sulfamethoxazole vs quinolones or beta-lactams and ESBL-active agents vs non-ESBL-active antibiotics. The primary outcome was clinical cure by day 7. We compiled risk ratios (RRs) with 95% confidence intervals (CI) using fixed-effect meta-analyses. Risk of bias was assessed with ROB 2.0 and evidence certainty using GRADE.

Thirty-seven RCTs were included, comprising 7,904 mainly hospitalized participants (62.7% women). Most had some risk-of-bias concerns. No significant difference in clinical cure was observed with moderate evidence certainty, across all comparisons. Relapse increased significantly with TMP/SMX compared to quinolones or beta-lactams (RR 2.34 [1.29, 4.24]) and microbiological cure improved with ESBL-active antibiotics (RR 1.07 [1.02, 1.13], very low certainty evidence). TMP/SMX increased adverse events and aminoglycosides increased nephrotoxicity and resistance acquisition.

Clinical cure was comparable across all comparisons. This supports basing antibiotic selection on local resistance patterns, patient factors and toxicity profiles rather than presumed class superiority.