Comparison of Prone and Supine Positions for Suction Mini-PCNL

Prone positioning matched supine for stone clearance (85% vs 81%) but had higher rates of blood transfusion, renal pelvic perforation, and pneumothorax

Journal: European Urology Focus | Published: 2025-10-29 | Type: Prospective Multicenter Observational Study, Comparative Study | PMID: 41168029 Authors: Castellani D et al., Endourology Section of the European Association of Urology and the Suction Mini-PCNL Collaborative Study Group (30 centers across Europe, India, Russia, Turkey, Philippines) Funding/COI: Not disclosed

Summary

This is the largest prospective dataset to date on patient positioning for suction mini-percutaneous nephrolithotomy, drawing on 1,534 patients across 30 centers over eight months. Both prone and supine positions achieved similar stone-free rates — 85% versus 81% — and the difference didn't reach statistical significance (p=0.052). However, prone position was associated with meaningfully higher rates of blood transfusion (2.8% vs 0%), renal pelvic perforation (2.8% vs 0.23%), and pneumothorax (1.5% vs 0%), though multivariable analysis did not show prone positioning as an independent predictor of overall complications.

Claims

• Stone-free rate (zero fragments at 30-day CT): 85% prone vs 81% supine (p=0.052)
• Blood transfusion: 2.8% prone vs 0% supine (p<0.001)
• Renal pelvic perforation: 2.8% prone vs 0.23% supine (p<0.001)
• Pneumothorax: 1.5% prone vs 0% supine (p<0.001)
• Spinal anesthesia use: 68% prone vs 29% supine (p<0.001)
• Supracostal access: 36% prone vs 22% supine (p<0.001)
• Surgical time, pain scores, hospital stay, and readmission rates: no significant difference between groups
• Multivariable OR for prone vs stone-free status: 0.92 (95% CI 0.66–1.29; p=0.6)
• Multivariable OR for prone vs overall complications: 0.87 (95% CI 0.59–1.28; p=0.5)

Study Quality

Prospective design and 1,534 patients across 30 international centers is genuinely large for a procedural study. The 30-day CT stone-free assessment is the right endpoint — better than plain X-ray, better than surgeon self-report. Multivariable adjustment was applied to control for confounders, which is appropriate given the non-randomized design.

The problem is that the two groups weren't remotely comparable at baseline. The prone group had a higher proportion of Guy's Stone Score 1 stones (simpler cases: 60% vs 47%), yet also had more supracostal access (riskier upper-pole punctures) and more frequent fluoroscopy-only guidance rather than ultrasound. The supine group had more Grade 4 stones (6.9% vs 3.2%). This is classic selection bias: different centers, different surgeons, different case mixes, different technique preferences — all collapsed into a "prone vs supine" comparison. When raw complication rates differ by that magnitude (blood transfusion 2.8% vs 0%, pneumothorax 1.5% vs 0%), and then multivariable analysis makes those differences disappear, the right reaction is skepticism about the model, not reassurance about the procedure.

Red Flags

• Non-randomized by design — the authors acknowledge selection bias directly, and it's severe; center-level variation is likely a major uncontrolled confounder
• Prone group had significantly simpler stones (more Grade 1) but also riskier access patterns (more supracostal), making the groups incoherent to compare directly
• Funding and conflict of interest not disclosed for a 30-center industry-adjacent study — this is a meaningful gap
• The stone-free rate p-value of 0.052 is one data point away from significance in a 1,534-patient study; calling it "comparable" is accurate but convenient
• Pneumothorax rate of 1.5% in the prone group (zero in supine) is a clinically significant finding that the multivariable model's null result does not explain away
• No data on surgeon experience, center volume, or learning curve — critical for a technique-sensitive procedure

Strengths

• Largest prospective dataset on this specific question (1,534 patients, 30 centers, 8 months)
• 30-day CT for stone-free assessment is rigorous
• Prospective data collection, not retrospective chart review
• Captures a real-world, multi-continent practice pattern — useful for understanding variation even if causality is unresolvable
• Reports specific complication types rather than just aggregate Clavien-Dindo scores

Verdict

This paper gives you good epidemiology of how suction mini-PCNL is practiced globally and confirms that both positions can achieve stone-free rates above 80% in experienced hands. What it cannot tell you is whether prone or supine positioning causes differences in outcomes — the case mix is too different, the centers too heterogeneous, and the analysis too reliant on a multivariable model that erases complication signals that look clinically real (zero transfusions in 881 supine patients vs 2.8% in 653 prone patients is not noise). Read it as a descriptive registry study, not a comparative effectiveness trial. The randomized controlled trial this question deserves still doesn't exist.