Non-antibiotic options for recurrent urinary tract infections in womenBMJ 2017; 359 doi: https://doi.org/10.1136/bmj.j5193 (Published 23 November 2017) Cite this as: BMJ 2017;359:j5193
- Jonathan Barclay, specialty trainee in urology,
- Rajan Veeratterapillay, consultant urological surgeon,
- Chris Harding, consultant urological surgeon
- Correspondence to C Harding:
What you need to know
After treating the acute infection, low dose antibiotics given for 6-12 months are the most evidence based preventive measure for recurrent urinary tract infections (UTIs) in women, and are recommended by national and international guidelines as the standard of care
There is evidence of antimicrobial resistance development with prolonged low dose antibiotics and this has led to the investigation of non-antibiotic alternatives to prevent UTIs
Evidence for non-antibiotic treatments is variable: vaginal oestrogens, D-mannose, immunotherapy, and methenamine look most promising
Recurrent urinary tract infection in women is a common problem. There is no universally accepted definition, but most clinicians would classify recurrent UTI as being two episodes of infection in six months, or three episodes in one year.1 The annual incidence of a single UTI is 30 per 1000 women,2 with 44% experiencing recurrence within 12 months.3
When to investigate further
Most women presenting with recurrent UTI have uncomplicated infections (defined as no structural or functional urinary tract abnormalities).1 Current international guidelines do not recommend routine investigations of women with recurrent UTI because of their low diagnostic yield.1 Symptoms such as persistent loin pain, haematuria, or the presence of atypical infection would necessitate referral to secondary care, where further tests (renal tract ultrasound/computed tomography, flexible cystoscopy, and urine flow studies/post-void urine estimation) can be carried out.
Guidelines from the Scottish Intercollegiate Guidelines Network4 and the European Association of Urology1 recommend the use of low dose prophylactic antibiotics for preventive treatment of recurrent UTI. The largest systematic review and meta-analysis of the effect of prophylactic antibiotics on recurrence rates includes 19 randomised controlled trials with data from 1120 women5 and reports an 85% reduction in the incidence of symptomatic UTI compared with placebo (relative risk 0.15, 95% confidence interval 0.08 to 0.28). This review examined data from girls and women over 14 with a history of at least two episodes of uncomplicated UTI and calculated that the number needed to treat with prophylactic antibiotics to prevent recurrence over a 6-12 month period was 1.855 (however, the authors commented that the quality of the included studies was poor).
Prolonged antibiotic use in women has resulted in the emergence of resistant organisms in their urine. A randomised controlled trial of 252 women who had been through menopause showed that if resistance develops, it is often to multiple antimicrobial agents and not just the primary drug that the patient has been prescribed.6 This trial, comparing antibiotic prophylaxis against urinary tract infections with an oral Lactobacillus preparation, found that in the antibiotic arm (trimethoprim), resistance to other antibiotic agents such as amoxycillin and co-trimoxazole doubled over a period of one month.6 Resistant micro-organisms can develop after just a few weeks of treatment. A prospective cohort study involving 136 students (no ages given) who were given trimethoprim for two weeks found resistance to at least four antibiotics in 96% of Escherichia coli from faecal isolates.7 Consequently, other treatment options are of interest.
Emerging potentially efficacious non-antibiotic treatments (table 1⇓) for prevention of recurrent UTI could minimise the development of antimicrobial resistance in bowel reservoirs but there is a paucity of high level evidence to demonstrate their effectiveness.
What is the evidence for uncertainty?
Despite the existence of several up to date Cochrane meta-analyses on the subject of recurrent UTI prevention, uncertainty remains because of the lack of good quality comparative head to head trials. Consequently, precise quantification of the risks and benefits of prolonged non-antibiotic treatments is yet to be established.”
Sources and selection criteria
We searched only level 1a evidence in October 2016: systematic reviews (with homogeneity) of randomised controlled trials. We performed a digital search of the most recent issue of the Cochrane Database of Systematic Reviews via the Cochrane Library using the following MeSH terms: “urinary tract infection,” “recurrent urinary tract infection,” “bacterur*,” “urin* infection,” “preventing urin* infection” and “preventing urin*.” The results generated for each MeSH search were then evaluated and the systematic reviews relevant to the adult female population were selected. We also performed a literature search of PubMed, Embase, and Medline using the MeSH terms outlined above. We identified relevant randomised controlled trials or other systematic reviews that had been conducted since the Cochrane Systematic Reviews were published. We have not referenced them in this article, as they showed a vast heterogeneity of the sample groups, assessment criteria, and outcome measures assessed. Despite the existence of several up to date Cochrane meta-analyses on the subject of recurrent UTI prevention, uncertainty remains because of the distinct lack of good quality comparative head to head trials. Consequently, precise quantification of the risks and benefits of prolonged non-antibiotic treatments is yet to be established.
What are the non-antibiotic treatment options?
Urinary alkalisation is the administration of oral medications, such as potassium citrate, to reduce the acidity of urine. By raising the pH of urine, it is postulated that the severity of dysuria experienced by the patient is reduced.
Probiotic organisms (eg, Lactobacillus spp) are thought to establish a barrier against infectious pathogens ascending the urinary tract and subsequently causing infection. They modulate host defences by reducing pathogen adherence, growth, and colonisation.
Chinese herbal medicine
According to a 2015 Cochrane review, the Chinese herbal medicines most commonly used are Er Xian Tang, Bai Tou Weng Tang, and San Jin Wan.
Methenamine hippurate11 is hydrolysed to formaldehyde in the presence of acidic urine and exerts a bactericidal effect on E coli.
It is postulated that cranberries prevent bacteria (particularly E coli) from adhering to the uroepithelial cells.12
Topical application of vaginal oestrogen improves vaginal atrophia and increases vaginal lactobacilli. These promote a change in vaginal pH, suppressing Gram negative bacterial growth.
Intravesical hyaluronic acid aims to replenish the surface glycosaminoglycan layer of the urothelium and can prevent bacterial adherence.
Oral immunostimulants contain serotypes of heat killed/inert uropathogens designed to upregulate the patient’s immune response to urinary infections. Immunostimulants work by increasing the activity of components of the immune system, rather than providing an acquired immunity to a specific pathogen, like a vaccine does.
A meta-analysis examining the efficacy of urinary alkalinisation for prevention of recurrent UTI screened 172 randomised controlled trials, all of whichfailed to meet the inclusion criteria because of poorly defined study populations, interventions, and a lack of comparators (table 1⇑). No recommendations could be made, therefore, for the use of urinary alkalisation.8 No comment can be made regarding the risks or adverse events of urinary alkalinisation, but recognised side effects include gastrointestinal symptoms such as nausea and flatulence and mild diuresis.
Probiotics in recurrent UTI were the subject of a recent Cochrane systematic review and meta-analysis including nine randomised controlled trials of 735 patients.9 No statistically significant benefit (relative risk 0.82 v placebo [ 95% confidence interval 0.6 to 1.12]; relative risk 1.12 v antibiotics [95% confidence interval 0.95 to 1.33]) was shown for probiotics over placebo or no treatment, but the authors concluded that a benefit cannot be ruled out as the number of patients was small, and the trials had poor methodological reporting because of high reporting and attrition bias. Reported side effects of probiotics included vaginal discharge, genital irritation, and diarrhoea, but these only affected 23 out of 735 women.
Chinese herbal medicine
In a 2015 Cochrane review, meta-analysis of three randomised controlled trials involving 282 women that looked at Chinese herbal medicine compared with antibiotics suggested that the Chinese herbal medicine reduced recurrent UTI rates (relative risk 0.28, 95% confidence interval 0.09 to 0.82), but the authors advised caution because of the small study sizes, the high risk of bias, and the lack of evidence in the premenopausal population. Only two studies reported on adverse events, and neither found any change in liver or kidney function. It was not explained in the review how this was assessed.
In a 2012 Cochrane review, methenamine hippurate11 was associated with a reduction in risk of UTI in patients with normal renal tracts (relative risk 0.24, 95% confidence interval 0.07 to 0.89) and was found to be well tolerated. However, only one of the trials—of 30 patients—included in this review looked specifically at people with recurrent UTI, and found that methenamine was associated with a reduction in UTI (relative risk 0.46). Low rates of adverse events were reported across all included studies. Nausea was the most common symptom, and was noted in 12 of 2032 people. Single instances of rash, diarrhoea, sore throat, and bladder “stinging” were also described. Given the demonstrable benefit of methenamine hippurate and its excellent side effect profile, it is one of the most promising non-antibiotic treatments for recurrent UTI, and is currently subject to an ongoing multicentre UK randomised controlled trial.14
A Cochrane review in 200815 found a reduction in the incidence of recurrent UTI (relative risk 0.65, 95% confidence interval 0.46 to 0.9), but since the dropout rates for several of these studies was high, there was a high risk of bias. An updated 2012 Cochrane review including an additional 14 studies and 4473 participants (table 1⇑) suggested that cranberry juice was less effective than previously indicated; cranberry products (juice or dietary supplements based on fruit extracts) did not statistically significantly reduce the occurrence of symptomatic UTI overall (relative risk 0.86, 95% confidence interval 0.71 to 1.04) compared with placebo or no treatment. The subgroup analysis for women with recurrent UTIs also showed no difference (relative risk 0.74, 95% confidence interval 0.42 to 1.31). The authors concluded that cranberry juice could not be recommended for the prevention of UTI.
Although adverse events from taking cranberry products, such as gastrointestinal disturbance, were similar to those of placebo/no treatment, many studies reported low compliance and high dropout rates, which were attributed to palatability of the products.
A meta-analysis in 2009 (four randomised controlled trials, 891 patients) of oral immunostimulants looking at their effectiveness showed a benefit in decreasing recurrent UTI (relative risk 0.61, 95% confidence interval 0.48 to 0.78).16 However, a later randomised double blind trial (2015) involving 451 women with recurrent uncomplicated UTI found no benefit of oral immunostimulants compared with placebo.17 Guidelines from the European Association of Urology1 support their use.
Topical vaginal oestrogen, when used primarily as a treatment for recurrent UTI, was shown to be beneficial in reducing the incidence of infection in women who have been through menopause in a meta-analysis of five randomised controlled trials.13 The included populations were well defined as women with an existing diagnosis of recurrent UTI who were more than 12 months from their last menstrual period. The two randomised controlled trials analysed contained only small patient numbers with differing results (relative risk 0.25, 95% confidence interval 0.13 to 0.50, and relative risk 0.64, 95% confidence interval 0.47 to 0.86), and therefore no robust recommendations can be made on the basis of such small scale trials. Adverse events for vaginal oestrogens were breast tenderness, vaginal bleeding or spotting, non-physiological discharge, vaginal irritation, burning, and itching. These side effects were reported in a minority of participants, but exact numbers were not described.13
Is ongoing research likely to provide relevant evidence?
More than 190 studies on recurrent UTI are listed on clinicaltrials.gov.18 Emerging non-antibiotic prophylactic agents such as intravesical agents,19 vaccines,17 and D-mannose20 show promise in reducing the risk of recurrent UTI, but have yet be evaluated in robust randomised controlled trials. Further research is necessary to enable meaningful comparisons with prophylactic antibiotics.
A small meta-analysis of intravesical hyaluronic acid included four studies (two randomised, two non-randomised, 143 patients) looking at the use of this treatment in recurrent UTI, showing a statistically significant improvement in the rates of recurrent UTI (mean difference −3.4 episodes per patient year, 95% confidence interval −4.3 to −2.5).19
Vaginal vaccines (vaginal immunogens applied mucosally) have also been shown (three randomised controlled trials, 220 patients) to convey some benefit (relative risk 0.81, 95% confidence interval 0.68 to 0.96)21 in reducing the risk of recurrent UTI, especially if the initial inoculation is followed up with booster doses. Smaller studies examining sublingual vaccines20 and the use of D-mannose oral powder14 have also shown promising results and have been compared directly with the current standard of prophylactic antibiotics.
Sublingual vaccines and D-mannose
In a multicentre observational trial involving 319 women with recurrent UTI, the efficacy of a sublingual vaccine containing a suspension of inactivated E coli, Klebsiella, Proteus, and Enterococcus was examined. In the patients receiving vaccine there was a 75% reduction in mean number of infections compared with those receiving daily trimethoprim (0.36 v 1.6 episodes at 3 months).20 The efficacy of D-mannose has been studied in a recent three arm randomised controlled trial comparing D-mannose with nitrofurantoin with no prophylaxis.14 Overall, 98 patients (31.8%) had recurrent UTI: 15 (14.6) in the D-mannose group, 21 (20.4) in nitrofurantoin group, and 62 (60.8) in no prophylaxis group, with the rate statistically significantly higher in the no prophylaxis group compared with the active groups. There was no statistically significant difference in the efficacy of D-mannose when compared with prophylactic antibiotics.
What should we do in the light of the uncertainty?
Based on the currently available evidence, prophylactic antibiotics remain the first line preventive treatment for uncomplicated recurrent UTI in women, and this is reflected in practice guidelines (Cochrane reviews, detailed in this article, have concluded that there is insufficient evidence to make a conclusive recommendation for many of the non-antibiotic treatments described).
Nonetheless, increased awareness of antimicrobial resistance has led patients and clinicians to seek alternatives. It has been noted that non-antibiotic prophylactic agents seem generally well tolerated, with few adverse events or side effects reported in the preliminary work described. Some of the non-antibiotic treatments—for example, vaginal oestrogens and probiotics, have been used extensively for other indications and therefore there are no concerns regarding their safety. As a result, there might be a case to support the use of non-antibiotic prophylactic agents, on an individual patient basis, for those who do not wish to take, or have an allergy or intolerance to, antibiotics.
Recommendations for future research
Population: Female patients with recurrent UTI
Intervention: Prophylactic treatments to reduce/prevent infection
Comparison: Prophylactic antibiotics (current gold standard)
Outcome: Reduction in infection episodes, side effects of treatments, health economic analysis of cost effectiveness
Many of the studies in progress are small, and further research with larger numbers of participants is needed before any firm conclusions for prevention of recurrent UTI can be made. Specifically, new studies should be designed to examine patient reported outcomes and quality of life indices, as well as more conventional endpoints such as bacterial eradication. In addition, the use of current standard treatment of prophylactic antibiotics as a comparator should be encouraged, and studies should be designed to look at other endpoints such as antimicrobial resistance development as well as clinical benefit.
Education into practice
When a patient presents with recurrent UTIs, ask yourself:
Have I addressed all modifiable risk factors for this patient—eg, adequate fluid intake, diabetic control, oestrogen depletion, and constipation, before recourse to preventive measures?
Do I need to perform further investigations in this patient, or am I happy that this is an uncomplicated recurrent UTI?
Have I discussed the potential risks and benefits of the standard treatment of prophylactic antibiotics with the patient? Should we consider the use of a non-antibiotic alternative in the light of the evidence presented?
What to tell patients
Most women with recurrent UTI will have a structurally and functionally normal urinary tract, therefore treatment is based on prevention and reducing infection frequency
Low dose antibiotics for 6-12 months carry the highest chance of success in reducing the frequency of UTI
Avoiding the development of antibiotic resistance is best achieved with non-antibiotic measures, but it is unclear how effective these are
How patients were involved in this article
There was no patient involvement from the authors when constructing this article. However, the patient reviewer highlighted the importance of patient reported outcome measures in any future or planned trials in addition to traditional microbiological tests.
This is one of a series of occasional articles that highlight areas of practice where management lacks convincing supporting evidence. The series advisers are Sera Tort, clinical editor, and David Tovey, editor in chief, the Cochrane Library. To suggest a topic for this series, please email us at
Conflicts of Interest: CH reports speaker fees from Astellas, Pfizer, Ferring, Allergan, Medtronic, and is a member of the advisory boards for American Medical Systems and Pierre Fabre Pharmaceuticals. JB and RV have no conflicts to report.
Patient consent not applicable.
Provenance and peer review: commissioned, externally peer reviewed.