urinary tract infections review article

JAMA just published a good up-to-date review article on the diagnosis and treatment of uncomplicated urinary tract infections, defined as acute onset of dysuria, frequency or urgency in a healthy, nonpregnant woman without known functional or anatomical abnormalities of the urinary tract (see urinary tract infection review jama 2014​ in dropbox, or doi:10.1001/jama.2014.12842), incorporating all articles published until july 2014 (and not including treatment articles before 2000 since the bacterial resistance patterns have changed so much). they assessed 27 RCTs and clinical trials with 6463 patients, 6 systematic reviews and 11 observational studies with >250K patients.  their conclusions:

--diagnosis: women with at least 2 symptoms of UTI (dysuria, urgency or frequency) and the absence of vaginal discharge have >90% probability of acute cystitis. adding a urine dipstick for leukocyte esterase does not really change the probability, and studies show no benefit of urine culture. so, unless this is relapse/recurrent infection (>2 UTIs within 6 months), complicated infections, or women with a history of multi-drug resistant organisms, telephone-based therapy is as good as coming in to see a clinician (documented by studies). also patient-initiated therapy is good (small studies, but women were correct in diagnosis >90% of the time) -- ie, giving the woman antibiotics to have at home and take as soon as they get symptoms (not-so-shockingly, the sooner they start therapy, the sooner they feel better).
--treatment: best are trimethoprim/sulfamethoxazole (TMP/SMX) DS bid x3 days, with clinical cure rates 85-100%; nitrofurantoin 100mg bid for 5-7 days (one study showing 5 days was as good as 7). no difference in clinical cure rates compared to TMP/SMX​; fosfomycin 3-g as a single dose. may be a little less effective (i have prescribed it only once in my life, under slight duress from an insistent patient). of note, the relative cost for a course of therapy for self-paying patient at our pharmacy: TMP/SMX $20.45, Nitrofurantoin $44.10, Fosfomycin (can only get as 3-pak) $198.20
--NOT use: fluoroquinolones (unlike the above, this gets good serum levels and should be reserved for complicated UTIs. does get good cure rates, but there is increasing bacterial resistance in the community); b-lactams (amoxicillin-clavulanate and cefpodoxime-proxetil) are less effective. non-antibiotics, including ibuprofen, placebo, cranberry products seem mostly to delay antibiotics.
--in terms of resistance: good to know what the sensitivities are in your area to the different antibiotics and contour therapy to that. it should be noted, however, that many UTIs respond to resistant antibiotics: eg, for TMP/SMX in one study 84% or susceptible organisms responded vs 41% of resistant ones, presumably because there is such concentration of antibiotics in the urine that they still work in many cases, which seems better than other studies with placebos. one rule-of-thumb is to use another antibiotic when the resistance rate is above approx 20%.
--other groups -- not much data. the sense of the authors (under the guise of "expert opinion") is that men do well with a 14-day course (1 RCT, 1 observational study) in those with febrile UTI involving the prostate, but the experts suggested a 7-day course was likely as good; for diabetic women (1 observational trial), the experts side with treating them as above for nondiabetics.

so, this article basically reinforces immediate treatment of women with uncomplicated UTI symptoms with mostly TMP/SMX or nitrofurantoin, keeping in mind the local resistance patterns. my personal approach for women with relatively frequent infections is to give them a big bottle of TMP/SMX and have them dose themselves.  seems to work well, assuming one chooses appropriate patients (which is really a pretty large % in my practice).