Twenty-Four Hours or Seven Days, for a Kidney That Keeps Growing Stones Around Bacteria
Two real bodies of guidance exist for antibiotics before PCNL, and they've never been reconciled. The question is whether this particular patient is the one the longer course was actually built for.
Carol S., a 66-year-old woman, has spent the last four years as her husband's primary caregiver since his Parkinson's diagnosis made driving, cooking, and most mornings a two-person job by necessity. She retired early from the postal service to do it, and has mentioned twice in clinic that her own appointments get rescheduled more often than she'd like because his are less predictable than hers. She has a 2.1cm right renal pelvis stone with a struvite composition, her third infection stone in six years, and her urine has grown Proteus mirabilis on two separate cultures — a urease-producing organism that alkalinizes urine and drives exactly the kind of crystal formation that keeps rebuilding this stone around the bacteria itself, not just alongside them. She's scheduled for PCNL.
The AUA's Best Practice Statement on antimicrobial prophylaxis recommends 24 hours or less of perioperative antibiotics for PCNL, a recommendation built on the general PCNL population and reinforced by real antibiotic-stewardship concerns about driving resistance with longer courses. The EDGE endourology consortium has run this question as two separate multi-institutional randomized trials, and the answer depends entirely on which patient is asked. Chew and colleagues tested the exact regimen now being proposed for Carol — nitrofurantoin 100mg twice daily for seven days — against no oral antibiotic in patients defined as low risk by negative preoperative cultures and no urinary drain, and found nothing: sepsis in 12% against 14%, p of 1.0, with the conclusion that perioperative dosing per the AUA statement was sufficient. Sur and colleagues then ran the same drug in moderate-to-high-risk patients and found the opposite, seven days beating two days for sepsis prevention, and closed by recommending that guidelines start stratifying by infectious risk at all. What matters for Carol is which side of that line two Proteus cultures put her on: the negative trial would have excluded her at the door for exactly the finding that brought her here.
Pre-admission planning
The EDGE consortium answered this in two trials, not one. Chew's trial found nothing in low-risk patients, and I don't dispute it. Sur's trial, same consortium, same drug, in moderate-to-high infectious-risk patients, found seven days of preoperative nitrofurantoin beat two days on sepsis — and its authors ended by saying guidelines ought to stratify by infectious risk rather than issue one number for everybody.
Carol has grown Proteus mirabilis on two separate cultures. Chew's trial required a negative preoperative culture to enroll; she would have been screened out of it. I'd give her a week of preoperative oral nitrofurantoin on top of the standard single perioperative dose.
The AUA Best Practice Statement is the actual named guideline for PCNL prophylaxis, and it recommends 24 hours or less — for real reasons, not just caution, and Chew's trial is the one that actually tested a week against it and found nothing. I'd also note nitrofurantoin concentrates in urine and essentially nowhere else, which is a limitation for organisms living inside a struvite matrix rather than swimming in front of it. Carol has already grown the same organism twice. Every additional week-long course we give her is another opportunity to select for a resistant strain in a patient who is, almost by the nature of recurrent struvite stones, going to need antibiotics again.
I'm not saying the extended-course data is wrong. I'm saying we're weighing a real, immediate stewardship cost against a benefit measured in trials that, however well-targeted, are still a distinct and separate body of evidence from the guideline actually governing this procedure.
I don't think these two positions are actually in conflict once you read each trial's own enrolled population. Chew's negative result is a true finding about culture-negative patients, and the AUA statement it endorses was written for the same group. Sur's positive result is a true finding about the group Carol is in. You're both citing the same consortium and reaching opposite conclusions because you're citing different patients.
Both can be right at once, for different jobs: a single perioperative dose covers the immediate surgical-site risk the AUA statement addresses, and the preoperative oral course treats a documented, twice-cultured bacteriuria rather than prophylaxing against a hypothetical one. And I'd hold onto the stewardship objection rather than dissolve it — Sur's comparator was two days, not none, so what the trial actually licenses is a longer course over a shorter one, not a free hand.
Agreed: seven days of preoperative oral nitrofurantoin, plus the standard single-dose IV cefazolin within an hour of incision — both, not one instead of the other. The Primary Care Physician's reframing (two guidelines answering two different questions) is what actually resolved the apparent disagreement, rather than either side conceding ground.
The Antimicrobial Stewardship Pharmacist's underlying concern about her long-term antibiotic exposure wasn't dismissed — it's now flagged in her chart as a reason to pursue definitive stone clearance aggressively at this procedure, since recurrent partial treatment is what drives the repeat-course pattern stewardship is actually worried about.