June, 18 years ago, I was standing on the front lawn of the First Baptist Church in Augusta, GA, having just recited the Hippocratic Oath as part of graduation ceremonies. My brother, having also just graduated from a different medical school asks me, “So what do you think you can actually treat?” My reply belies reality. “I guess I could treat a UTI.” For many medical professionals, a urinary tract infection (UTI) is simple. Throw some antibiotic at the problem and it will go away. Simple things, however, do not often remain so simple with improved understanding.

UTIs are a problem. About 20% of the US population (with >80% of that number being women) will have at least 1 UTI during their lifetime. Millions of office visits and dollars are spent on the problem with the therapies in many cases spawning new problems and more spending. The issue is particularly vexing in the context of gynecologic surgery. The most common complication of pelvic floor surgery is UTI. In the randomized Stress Incontinence Surgical Treatment Efficacy trial that compared Burch colposuspension to autologous fascial sling in women with stress urinary incontinence, 48% of women in the sling group and 32% women in the Burch group reported UTI within the first 24 months of follow-up. Among women pursuing obliterative surgery for pelvic organ prolapse, 45% had UTI within 3 months of surgery. Given the morbidity, and in some cases mortality–approximately 13,000 deaths attributable to UTI occur annually in the United States –finding an effective prophylaxis is a worthy goal.

Unfortunately, trials of UTI prophylaxis in the context of female pelvic floor surgery arrive at opposite conclusions. In a multicentered, randomized trial comparing nitrofurantoin to placebo among women with a suprapubic tube following pelvic floor surgery, antibiotics decreased UTI incidence. Similarly, a randomized placebo-controlled trial with nitrofurantoin among women undergoing a midurethral sling also concluded that antibiotics decreased UTI incidence in this high-risk context. In contrast, a recent randomized trial in women undergoing pelvic reconstructive surgery that compared daily nitrofurantoin to placebo during the time a transurethral catheter was used, for up to 7 days, found no benefit from antibiotics. All of these studies struggle to offer guidance for women >65 years who have the highest risk for UTIs. In addition, many clinicians want to avoid using nitrofurantoin in patients with a creatinine clearance <60 mL/min although there appears to be little evidence to support this contraindication. So now what?

Cranberry has a similar conflicted history with respect to UTI prevention. There are studies that find an effect and others that do not. Per a recent Cochrane review, support for cranberry prophylaxis has waned in light of more evidence. This review considered 24 studies with a total of 4473 participants and found “the benefit for preventing UTI is small” and “cranberry juice cannot currently be recommended for the prevention of UTIs.” The same authors as part of an earlier Cochrane review of only 10 studies, however, concluded that cranberries did possess some efficacy to reduce UTIs in women. As was found for antibiotic-based UTI prophylaxis…so now what?

The proposed mechanism behind cranberry’s effect on UTI prevention relates to the proanthocyanidin (PAC) content of the fruit. Apparently PAC can block bacterial adherence to the urothelium (note cranberries are not the only nor do they possess the highest concentration of PACs among foods). In a clever study in the pediatric literature a randomized trial compared cranberry juice with 37% PAC to cranberry juice without PAC. Over 1 year, the researcher found the high-PAC juice was associated with a 65% reduction in UTIs in the study population (mostly girls 5-18 years of age) relative to the no-PAC juice. The authors note that PACs are light- and heat-sensitive, offering some insight into why the grocery story variety cranberry juice might not work as well to prevent UTIs.

There is considerable heterogeneity among the studies investigating UTI prophylaxis. Among other things, there are often differences across studies in how UTI or recurrence is defined, how and for how long the study drug was administered, randomization techniques and masking efforts, subjects’ compliance with the use of study drugs, or which form of cranberry was used, making comparisons difficult. Yet despite the extensive, albeit diverse, scholarship the matter of UTI prophylaxis remains unsettled. To this situation, the study by Foxman et al in this month’s journal juices up the debate again.

Like prior studies, Foxman et al compares cranberry to placebo. In agreement with Afshar et al, the PAC content was standardized and presumably stable and in concentrations shown to be effective per other recent pro-cranberry studies. The study methods are good. Randomization and masking techniques are sound. In contrast to other studies there is a good effort to document drug compliance and avoid dietary biases. Although the UTI definition was clinical with respect to the primary outcome, good efforts were made to collect urine for culture. Given these efforts the results are remarkable particularly given the context of postgynecologic surgery. The incidence of UTI, by clinical or culture-based diagnosis was less (as much as 50% less) in the cranberry cohort. Despite the randomization efforts, women receiving cranberry were less likely to require self-catheterization and more likely to have reported a history of UTIs; yet, adjusting for these, the positive effects of cranberry remained. Adverse effects were similar between cranberry and placebo.

So now what? It has been quite some time since I stood on that lawn in Augusta, GA. Epistemologists define knowledge as justified “true” belief. That’s an awful definition and any freshman philosophy student should recognize why. The trouble is there is no better alternative. Despite reasonable efforts to gain knowledge regarding the matter of UTI prophylaxis it’s not clear what’s “true.” Based on the 2 most recent and arguably best-quality studies on antibiotic or cranberry UTI prophylaxis in the context of gynecologic surgery one could conclude nitrofurantoin does not work and cranberry does. Clearly a head-to-head trial comparing these 2 approaches would be helpful. If approaching the matter from the perspective of value, a 30-day supply of nitrofurantoin (100-mg pill/d) costs between $42 and $68. A 45-day supply of Theracran HP (Theralogix, LLC, Sharon Hill, PA) taken per the Foxman et al protocol costs $65 (note there are lots of cranberry products but few specify the delivered amount of PAC). The risk of adverse events would seem to favor the cranberry, albeit adverse events are low for both therapies. In balance, it could be argued the cranberry narrowly comes out ahead. So how would I respond to my brother’s query today? Enjoy the fruit of Foxman et al’s labor!