Six Recurrences and No Clean Prophylaxis Left
A single patient, six recurrences into a year of urinary tract infections. The disagreement isn't about which antibiotic works best in general — it's about what's left to prescribe once the three usual first-line options have each been disqualified for a reason specific to her.
Renata O., a 52-year-old woman who works as a hospice chaplain, has spent the past year telling her own patients that some discomforts are simply the toll of getting older — advice she has begun resenting having to take herself. Six urinary tract infections in twelve months have interrupted her rounds at the inpatient hospice twice, each one meaning a same-day antibiotic and a rescheduled visit with a family she had promised to see that week. She has had type 2 diabetes for eleven years, reasonably controlled on metformin and semaglutide, and her renal function has drifted downward over that same stretch — her eGFR, in the high 50s three years ago, now sits at 42, a slow expected decline rather than a new insult, confirmed stable across two prior readings. A single dose of trimethoprim-sulfamethoxazole three years ago produced a diffuse hives eruption within hours, documented in her chart as a genuine allergy rather than an intolerance worth pushing through. Neither fact alone would be unusual. Together, they are what makes today's visit different from her first five: nitrofurantoin's urinary concentration depends on the kidney filtering it into the collecting system in meaningful amounts, and the FDA-approved label still lists a creatinine clearance below 60 mL/min as a contraindication outright. Her 42 is inside that labeled bar. It is also inside the range the 2015 Beers Criteria revised out of the bar, having moved the threshold to 30 mL/min after Oplinger and Prakash traced the original 60 figure back to a 1968 study that measured how much drug appeared in urine rather than whether patients got better. So her number is simultaneously a labeled contraindication and, on the geriatrics literature's own revision, not one — which is a genuinely different problem from a number that is merely borderline.
What is left, once trimethoprim-sulfamethoxazole is removed for allergy and nitrofurantoin is questioned for renal function, is fosfomycin — not FDA-approved for prophylactic dosing at any interval, only for single-dose treatment — and methenamine hippurate, a non-antibiotic option that had for years been treated as a weaker fallback rather than a real first choice. The fluoroquinolones she took without incident for her first two infections, years before the 2016 FDA boxed warning restricted their use in uncomplicated UTI, are no longer an option anyone at this clinic will reach for first. Published comparative work has found no meaningful difference in recurrence prevention among nitrofurantoin, trimethoprim-sulfamethoxazole, and several other agents head-to-head — reassuring in principle, and entirely silent on what to do once two of the three are off the table for reasons that have nothing to do with comparative efficacy.
At the follow-up visit, after the sixth recurrence
Start methenamine hippurate, twice daily. ALTAR randomized 240 women with recurrent UTI just like Renata to methenamine versus daily antibiotic prophylaxis and found it held up — non-inferior, absolute difference of about half a UTI per person-year, comfortably inside the trial's own predefined margin. It sidesteps her allergy and her renal number entirely instead of arguing around either one.
If she had normal renal function and no allergy, I might still reach for it first on stewardship grounds alone — ALTAR also found antimicrobial resistance ran proportionally higher in the antibiotic arm.
I manage her diabetes every quarter, and her diet log is not consistent about the acidic-juice supplementation her last visit recommended. Methenamine only converts to its active antibacterial form — formaldehyde — in acidic urine below about pH 5.5. If her urine runs alkaline more often than not, we're prescribing a drug that may simply not activate.
ALTAR's population was UK women with recurrent UTI generally, not selected for reliable urinary acidification — I'm not disputing the trial, I'm asking whether Renata specifically fits the population it actually validated.
Before we choose between two second-line options, I want to name something neither of you has, and I want to name both halves of it. The Macrobid label contraindicates nitrofurantoin below a creatinine clearance of 60, full stop — so on the label she is disqualified. But the 2015 Beers Criteria moved that threshold to 30 after the evidence behind the 60 figure was traced to a 1968 urinary-excretion study with no clinical endpoints, and retrospective series since have found cure rates in the 30-60 band comparable to normal renal function. We may be disqualifying our actual first-line drug on a label figure the geriatrics literature has already walked back.
I take the acidification point seriously, but it argues for confirming her urine pH before committing to methenamine, not for skipping straight to fosfomycin's off-label dosing — that has even less of a defined interval behind it than either alternative.
And I'll bound my own point before either of you does: what Beers actually licensed at a clearance above 30 is short-course treatment, ≤7 days. Renata needs twelve months of prophylaxis, and nitrofurantoin's pulmonary, hepatic, and neuropathic toxicity is specifically a function of long-term exposure. So the number may not disqualify the drug, and the duration may disqualify it anyway. Those are two separate questions and we have been running them together all visit.
Agreed: methenamine hippurate 1g twice daily started today, with a urine pH check at the two-week follow-up to confirm the drug is actually converting to its active form.
Not agreed, and left explicit rather than smoothed over:
The primary care physician's concern was unfounded for Renata specifically; move to reassessing whether methenamine's own dosing needs adjustment before abandoning it.
The mechanism genuinely can't activate; fosfomycin at a to-be-defined interval becomes the actual next step, and nitrofurantoin gets a second look with a repeat eGFR in hand.
The clinical pharmacologist's point about the eGFR range was accepted by both other voices as correct but not acted on this visit — nobody wanted to layer a second open question onto a plan already contingent on a pH result still pending.