Urinary tract infection (lower): antimicrobial prescribing

Two randomised controlled trials (RCTs) (Bleidorn et al. 2010 and Gagyor et al. 2015) in non-pregnant women found conflicting evidence regarding the effectiveness of ibuprofen compared with antibiotics. One small RCT, Bleidorn et al. 2010, found no significant differences between ibuprofen and ciprofloxacin in reducing symptoms and symptom duration (very low to low quality evidence). However, a more recent larger RCT, Gagyor et al. 2015, found that women using ibuprofen were more likely to report a higher burden of symptoms over the first 7 days after the start of their treatment, compared with women using fosfomycin (moderate quality evidence).

Subgroup analyses, based on the results of urine culture prior to treatment, showed that women who received ibuprofen had significantly fewer antibiotic courses per patient compared with women who received fosfomycin (Gagyor et al. 2015; moderate quality evidence). However, this effect was driven mostly by the randomisation process, rather than the effect of the treatment itself.

Women who received ibuprofen (irrespective of urine culture) were significantly more likely to receive an additional antibiotic prescription to treat a urinary tract infection (UTI) during 12 months of follow-up (31.1% versus 12.3%; high quality evidence), but were less likely to experience a recurrent UTI between days 15 to 28 of follow-up (5.8% versus 11.1%; moderate quality evidence). The authors noted that this result could be because the baseline risk of having recurrent UTI was greater in the fosfomycin group, as more women had experienced a UTI in the past year.

Bleidorn et al. (2010) did not report any safety or tolerability outcomes. However, Gagyor et al. (2015) found no significant difference between ibuprofen and fosfomycin in the incidence of pyelonephritis, febrile UTIs, patient-reported adverse events, or serious drug-related adverse effects (low to moderate quality evidence).

Two RCTs (Wing et al. 2008 and Wing et al. 2015) assessed the clinical effectiveness and safety of cranberry products for preventing asymptomatic bacteriuria in healthy pregnant women.

Wing et al. (2008) found that cranberry juice was not effective in preventing episodes of asymptomatic bacteriuria or UTI. However, a significant reduction in adherence was seen in women receiving cranberry juice compared with placebo (very low quality evidence).

Wing et al. (2015) assessed the safety and tolerability of cranberry capsules and found no significant difference in the number of babies born with a 1-minute Apgar score <7 in women who received cranberry capsules compared with placebo (21.4% versus 0%; very low quality evidence).

No systematic reviews or RCTs of any non-pharmacological or non-antimicrobial interventions were identified in men, older people or children.

No systematic reviews or RCTs of paracetamol were identified.

No systematic reviews or RCTs of hydration were identified.

One systematic review of RCTs (Falagas et al. 2009) in non-pregnant women found that women who were treated with antibiotics were more likely to have complete symptom resolution (61.8% versus 25.7%; NNT 3 [range 3 to 4]; high quality evidence) and microbiological success (defined as negative urine culture) (90% versus 33.3%; NNT 2 [range 2 to 2]; moderate quality evidence), and less likely to experience relapse after the end of treatment (15.8% versus 41.6%; NNT 3 [range 3 to 5]; moderate quality evidence), compared with placebo. There was no significant difference between groups in the incidence of pyelonephritis (0.14% versus 0.75%; low quality evidence), although due to the very low incidence of pyelonephritis, it is likely the studies lacked statistical power to detect a clinically important difference.

One systematic review (Smaill et al. 2015) and 1 RCT (Kazemier et al. 2015) assessed antibiotics compared with placebo or no treatment for managing asymptomatic bacteriuria in pregnant women. Smaill et al. (2015) found that pregnant women who received antibiotics for asymptomatic bacteriuria had a reduced incidence of persistent bacteriuria (20.3% versus 66.3%; NNT 2 [range 2 to 3]; low quality evidence); were less likely to develop pyelonephritis (5.6% versus 20.8%; NNT 7 [range 6 to 9]; moderate quality evidence) or deliver a preterm baby (<37 weeks) (5.8% versus 22.1%; NNT 7 [range 4 to 13]; moderate quality evidence), compared with those who received no treatment.

Smaill et al. (2015) found no significant difference between antibiotics and placebo in serious adverse neonatal outcomes (very low quality evidence).

Kazemier et al. (2015) found no significant difference between nitrofurantoin and placebo in reducing the incidence of symptomatic UTI, pyelonephritis or preterm birth (<34 weeks; very low quality evidence). Significantly more women had non-spontaneous onset of labour with nitrofurantoin compared with placebo, but there is considerable uncertainty with these results (very low quality evidence).

Zalmanovici-Trestioreanu et al. (2015) found that there was a greater incidence of bacteriological cure in older people who received antibiotics for treating asymptomatic bacteriuria compared with those who received placebo or no treatment (61% versus 17%; NNT 3 [range 2 to 3]; high quality evidence). However, there was no significant benefit in reducing symptomatic UTI (very low quality evidence) and people who received antibiotics were more likely to report adverse events (high quality evidence).

Antibiotic-associated diarrhoea occurs in 2 to 25% of people taking antibiotics, depending on the antibiotic used (NICE clinical knowledge summary on diarrhoea – antibiotic associated).

About 10% of the general population claim to have a penicillin allergy; this is often because of a skin rash that occurred while taking a course of penicillin as a child. Fewer than 10% of people who think they are allergic to penicillin are truly allergic. See the NICE guideline on drug allergy for more information.

People with a history of immediate hypersensitivity to penicillins may also react to cephalosporins and other beta‑lactam antibiotics (BNF, August 2018).

Nitrofurantoin should be used with caution in those with renal impairment (MHRA Drug Safety Update, September 2014). It should be avoided at term in pregnancy because it may produce neonatal haemolysis. Adults (especially older adults) and children on long-term therapy should have monitoring for liver function and pulmonary symptoms (BNF, August 2018).

Trimethoprim has a teratogenic risk in the first trimester of pregnancy (folate antagonist; BNF, August 2018). The manufacturers advise that it is contraindicated in pregnancy (trimethoprim summary of product characteristics).

A systematic review (Falagas et al. 2009) in non-pregnant women found a significant increase in the total number of adverse events with antibiotics compared with placebo (19.2% versus 12.9%; NNH 15 [range 9 to 16]; moderate quality evidence). However there was no significant difference between antibiotics and placebo in the number of withdrawals due to adverse events.

Zalmanovici-Trestioreanu et al. (2015) assessed the safety of antibiotics in the management of asymptomatic bacteriuria in older people. There was a significant increase in the incidence of adverse events in those treated with antibiotics compared with placebo or no treatment (4.2% versus 1.0%; NNH 31 [range 19 to 82]; high quality evidence).

See the summaries of product characteristics for information on contraindications, cautions and adverse effects of individual medicines.

One RCT (Little et al. 2010) assessed various antibiotic prescribing strategies in non-pregnant women with acute uncomplicated lower UTI, where immediate antibiotic treatment was not necessary. The women included in the study had a mean age of 39 to 45 years and had moderate symptoms (mean score of 1.7 to 1.8, on a scale of 0=no problem, 1=mild problem, 2=moderately bad problem, 3=severe problem) at baseline. Although bacterial confirmation was not essential for inclusion into the study, the proportion of women who had urine culture ranged from 23 to 89% (p<0.001), across 5 treatment groups. Immediate empirical antibiotics were compared with back‑up (delayed by 48 hours) empirical antibiotics, and immediate antibiotics based on either a symptom severity score >2, a positive dipstick test, or a midstream urine culture result. Only two-thirds of women randomised to receive antibiotics based on a midstream urine result had a positive urine culture.

There was no difference between the different prescribing strategies (immediate empirical antibiotics; back-up empirical antibiotics; immediate antibiotics based on symptom severity score >2, positive dipstick test or midstream urine culture result) in the severity or duration of symptoms during follow-up, or in the time to reconsultation (low to very low quality evidence). However, significantly more women who were prescribed immediate antibiotics used them, compared with all other prescribing strategy groups, except immediate antibiotics based on symptom severity scoring (very low quality evidence). There were also significantly fewer women waiting 48 hours before taking their antibiotics in the immediate antibiotics group, compared with all other prescribing strategy groups, except immediate antibiotics based on symptom severity score (low to very low quality evidence).

Little et al. (2010) found that despite randomisation, all groups delayed starting their antibiotic course by at least 24 hours (very low quality evidence). However, a delay of more than 48 hours was associated with a longer duration of moderately bad symptoms (very low quality evidence).

No systematic reviews or RCTs of back‑up antibiotic prescribing strategies were identified in men, pregnant women, older people or children.