Research in context
Evidence before this study
The Infectious Diseases Society of America and European Society for Clinical Microbiology and Infectious Diseases collaboratively updated the International Clinical Practice guidelines for the treatment of acute uncomplicated cystitis in women in 2010. They recommended nitrofurantoin 5-day, co-trimoxazole 3-day, fosfomycin trometamol single-dose, and pivmecillinam 3–7-day regimens as appropriate choices, and other antimicrobial regimens as alternatives. Evidence from various randomised clinical trials and previous meta-analyses suggest that shorter-term antimicrobial therapy are as effective as prolonged therapy, but not all shorter regimens were adopted into clinical guidelines. This study was done to reappraise the treatment duration of each of the antibiotics in current guidelines.
We searched PubMed, Embase, and Cochrane Library databases for randomised controlled trials published from the inception of each database until Dec 31, 2019. The following search terms were used: “uncomplicated”, “cystitis”, “urinary infection”, “antibiotic treatment”, and relevant variants (appendix p 1). Only randomised controlled trials that compared the clinical and microbial effectiveness of antibiotics in female patients with acute uncomplicated cystitis were included. A network meta-analysis was done to compare the efficacy of each antibiotic treatment regimen directly and indirectly.
Added value of this study
We reassessed the evidence supporting the current international guidelines with a network meta-analysis and rating the quality of evidence. Single-dose third-generation and fourth-generation fluoroquinolones had a similar effect on clinical and microbial responses as 3-day regimens, supported by a moderate quality of evidence. 3-day pivmecillinam had similar clinical and microbial responses to that of 5-day and 7-day regimens, supported by a moderate quality of evidence. Moderate quality of evidence suggests that the true effect is likely to be close to the estimate of the effect. However, there was no difference in both clinical and microbial responses between single-dose and 3-day regimens of third-generation cephalosporin, but the quality of evidence supporting this was either low or very low. This finding suggested a poor confidence in the effect estimate; therefore, there is insufficient power to affect the current clinical practices and guidelines. A 3-day regimen of nitrofurantoin appears to have a similar effect on clinical and microbial responses as the 5-day regimen, but the quality of evidence supporting this was very low. For co-trimoxazole and second-generation quinolones, single-dose therapy was less effective than the 3-day regimen, with moderate quality of evidence supporting this conclusion.
Implications of all the available evidence
In acute uncomplicated cystitis, treatment duration of pivmecillinam, third-generation and fourth-generation fluoroquinolones could be shorter than the recommended durations. The effect of shorter course regimens of nitrofurantoin (3-day), trimethoprim (3-day), and first-generation and third-generation cephalosporins (single-dose) might be similar to the conventional durations of each antibiotic; however, further research is needed because of the low quality of evidence associated with these conclusions. The currently recommended 3-day regimens of co-trimoxazole and second-generation quinolone appear to be appropriate. Given the significant rise in antibiotic resistance worldwide, our findings might add value to the literature because if regimens can be reduced in length, the adverse event and risk of resistance can be reduced.