Insufficient participating sites, and insufficient participants from participating sites mainly due to COVID-19 workload. The number of necessary participants could not be reached within the anticipated timelines and the allocated budget.

Cystitis is the most frequent reason for women to visit their general practitioner. More than 600.000 women suffer from urinary tract infections in The Netherlands each year. Currently, the 1st choice treatment for uncomplicated cystitis is nitrofurantoin (NIT) for 5 days. The second choice is 3 gram fosfomycin-trometamol (FT) in a single dose. FT is increasingly prescribed because it has few side-effects and it has a patient-friendly dosing scheme. Previous research did not show significant difference in efficacy between fosfomycin and nitrofurantoin, but a clinical trial from 2018 claims a single dose of FT might be inferior to 5 days of nitrofurantoin. Pharmacodynamic and pharmacokinetic research suggests that a single dose of FT may be insufficient to cure cystitis. Overall, it remains unknown whether a single gift of FT is as efficacious as 5 days of nitrofurantoin for uncomplicated cystitis with regard to clinical cure and if an additional gift of FT would overcome this. A clinical trial is therefore warranted.

Objective: To investigate the comparative effectiveness and side-effects of 5 days of nitrofurantoin, single dose FT, and extended use of FT in uncomplicated cystitis in primary care.

Study design: An open-label randomized non-inferiority / superiority study with 3 arms.

Study population: 777 non-pregnant women with symptoms of uncomplicated cystitis, with 259 subjects in each study arm.

Intervention: (A) FT in a single dose of 3000mg on day 1; (B) extended dosing of 3000mg FT on day 1 and 3 (C) nitrofurantoin 100mg bid (slow release) for 5 days.

Main study parameters/endpoints: primary: days of absence of cystitis symptoms within 28 days. Secondary: clinical failure on day 28, microbiological failure on day 28, incidence of side-effects, cost-effectiveness Burden and risks associated with participation, benefit and group relatedness: A potential risk of participation is that the treatment arm to which the patient is allocated is either less efficacious, has more adverse events or higher recurrence rate than the other treatment arms. However, NIT and FT are both frequently used for urinary tract infections and considered safe and effective compounds for uncomplicated cystitis. According to previous studies, a second dose of FT is well tolerated. The potential risks of participation on severe adverse events is expected to be negligible as the risk of severe clinical failure after cystitis treatment is only 1% according to previous studies and differences between NIT and FT have not been observed previously. A potential benefit of participating to this study is that a more patient friendly treatment scheme is equally effective. For future patients the guidelines could be improved and become more patient-friendly. The burden of participation is considered low. Study participants need to complete a short daily questionnaire on a mobile application up to 28 days.

1. INTRODUCTION AND RATIONALE

   Cystitis is the most common problem in primary care in the Netherlands for women (3.7% of all consultations in women).0 Women are disproportionally affected with about 70/1000 new lower urinary tract infections (UTIs) per year, compared to 10/1000 new UTIs per year for men.1 In the Netherlands this implies more than 600,000 women suffer from a UTI every year.

   Empiric treatment of lower urinary tract infection (UTI) targets Enterobacterales and in particular E.coli, the most prevalent causative pathogen. In the Netherlands, the current first choice treatment for uncomplicated cystitis is nitrofurantoin for 5 days (four times daily 50mg or twice daily 100mg in slow release form, mainly depending on availability at the pharmacy), the second choice is fosfomycin-trometamol (FT) in a single dose and the third choice is trimethoprim for 3 days.1 FT is increasingly prescribed, probably because it is easy to administer, well tolerated and patient friendly due to its shorter treatment duration. In a large database of prescribed drugs in primary care (Utrecht region) there is a clear increase in FT use since (in 2013) FT became 2nd choice for the treatment of uncomplicated cystitis, at the expense of nitrofurantoin (NIT) (now 1st choice) and trimethoprim (now 3rd choice). FT as a second choice is much more frequently used than trimethoprim as a second choice in 2013 (see figure 1, below)

   In two randomized controlled trials, performed more than 20 years ago, efficacy (based on participant reported symptoms) between a single gift of FT compared to nitrofurantoin four times 50mg or twice daily 100mg for 7 days did not differ statistically significant.4,5 Yet, one of these studies did not meet the necessary number of inclusions and was thus underpowered 4 and the other was performed in a population in which causative pathogens had higher levels of resistance to nitrofurantoin than is usual in the Netherlands. 5 A recent open-label randomized controlled trial suggests that a single gift of FT is less effective compared to thrice daily 100mg nitrofurantoin for 5 days for uncomplicated cystitis in non-pregnant women.3 Compared to the Dutch recommendation they applied a higher dose of nitrofurantoin, reported nitrofurantoin adherence was very high, and the study also included hospitalized patients (7,4% of study population) whereas in the Netherlands nitrofurantoin for uncomplicated cystitis is almost exclusively prescribed to outpatients.3 This implies that the findings of this study may not be fully generalizable to the Dutch situation, and call for a reevaluation and possible optimization of fosfomycin treatment for UTI.

   Although usually prescribed in a single dose, the optimal dose of FT for cystitis is still unknown.3 The pharmacodynamic parameter most strongly linked to effectiveness is the area under the curve / minimal inhibitory concentration (AUC/MIC) ratio. High inter-individual variability in urinary FT concentrations in healthy female volunteers were observed after a single-gift FT with fosfomycin concentrations in urine below the EUCAST breakpoint in two-thirds of the volunteers after 72 hours.8 Moreover, in vitro bladder models have shown that a single dose of oral FT 3000mg is insufficient to kill E.coli strains with a minimal inhibitory concentration (MIC)>4mg/L in urine. Consequently, a single dose FT may be insufficient and it is suspected that extended dosing of FT improves outcomes.8,7 Besides, in vitro activity of a single dose of FT for cystitis correlates badly to in vivo bacteriological efficacy, even when using agar dilution, a reference standard for susceptibility testing.3,9 For that reason, dose-finding studies in clinical patients such as in our trial are indispensable for cystitis.

   In summary, the current recommendation for Dutch GPs is to treat uncomplicated cystitis with a 5-day course of nitrofurantoin (2x100mg or 4x50mg daily). It is unknown whether a single dosage might be equally effective and if the efficacy of FT could be improved by an additional dosage on day 3. Efficacy reflects the rapidity in which complaints disappear and the likelihood of not developing a relapse or recurrence cystitis or pyelonephritis within 28 days.

   We, therefore, designed a non-inferiority/superiority trial for the treatment of uncomplicated cystitis in the Dutch community, in which the investigators compare a 1-day and 3-day regimen of FT to a 5-day regimen of nitrofurantoin to investigate the effect on time to resolution of symptoms and the occurrence of recurrence/relapse within 28 days.

2. OBJECTIVES

2.1. Primary Objective:

To investigate the effect of the different treatment arms on:

• the number of days with full resolution of cystitis symptoms within the period of 28 days 2.2. Secondary Objective(s):

• Rate of clinical failure within 28 days, irrespective the causing organism

• Rate of microbiological failure at day 28

• Rate of relapses at day 28,

• Rate of reinfections at day 28

• Rate of aggravation to pyelonephritis or urosepsis at day 28

• Rate of hospital admission at day 28

• Rate of mortality at day 28

• Incidence of (severe) adverse events (including fever, diarrhea, nausea, vaginitis, headache, dizziness) on day 7

• Self-reported therapy adherence on day 7

• Satisfaction with the treatment received after 28 days

• The number of days of absenteeism (from payed work or volunteer work)

• Correlation between in vivo and in vitro activity for the investigational treatments

• If funding permits:

  • A pharmacodynamic evaluation of in vitro activity measurement of fosfomycin for E. coli and Klebsiella spp.
  • Micro-organisms associated with clinical failure as identified with molecular microbiological methods
  • Develop novel approaches to identify human antibodies recognizing E. coli

2.3. Hypotheses

The investigators will test the following hypotheses related to the primary objective:

1. Fosfomycin in a single dose is non-inferior to nitrofurantoin,
2. Fosfomycin in two dosages on day 1 and 3 is non-inferior to nitrofurantoin
3. Fosfomycin in two dosages on day 1 and 3 is superior to fosfomycin in a single dose.

3. STUDY DESIGN

An open-label non-inferiority/superiority randomized clinical trial. Patients will be randomized in a 1:1:1 ratio to either of:

A) Fosfomycin in a single dose on day one (FT1) B) Extended use of fosfomycin during three days with a dose given on day one and three (FT1+3); C) Nitrofurantoin for 5 consecutive days (2 times daily 100mg in slow release form (Furabid®) (NIT1-5)*

In case of treatment failure the treating physician will act according to their medical standards. In general this implicates that study medication will be ceased and other antibiotics will be started according to the professional judgment of the physician.

4. STUDY POPULATION 4.1 Population (base) The study population consists of non-pregnant adult (>=18 years of age) women with cystitis. 5-10 general practices will participate. An average general practice performs about 8966 consulting moments a year (NIVEL) which implies at least 165 UTIs in women per general practice yearly. Of all eligible patients an inclusion rate of 25-30% is assumed.

4.4 Sample size calculation The investigators performed sample size calculation for both duration of symptoms and clinical failure using sample size calculations for non-inferiority. The sample size was driven by clinical failure. Assumptions for duration of symptoms were a standard deviation of 2.2 days10, clinically relevant non-inferiority margin of 0.8 days, two-sided alpha of 0.017, power of 90%, and 20% loss of efficiency due to skewness of the outcome parameter and potentially missing outcome data, which yielded a total sample size of 741 (247 per arm).

Assumptions for clinical failure were a failure rate at day 28 of 7.5% based on data provided by the Julius Huisartsen Netwerk, clinically relevant non-inferiority margin of 7.5%, two-sided alpha of 0.017, and power of 80%, which yielded a total sample size of 777 (259 per arm).

Hence the investigators aim to enroll 777 patients in total. The investigators did not take into account missing outcome data for this endpoint since these data will be derived from the GP medical records with a short follow-up period. The sample size has been calculated for the non-inferiority analyses, and ensures sufficient power for the superiority analyses.

6. INVESTIGATIONAL PRODUCT

6.1 Name and description of investigational product(s) 6.1.1 Fosfomycin Fosfomycin is a phosphoenolpyruvate (PEP) analogue that is produced by Streptomyces spp.. It has a bactericidal action, primarily by inhibiting bacterial cell wall (peptidoglycan) synthesis. In the Netherlands, Fosfomycin is orally available as fosfomycin-trometamol (Monuril). Fosfomycin-trometamol is identical to fosfomycin-tromethamine. Fosfomycin-trometamol is a phosphoric acid derivative of fosfomycin, available in a single dose sachet containing white granules. One sachet contains 5.63 g of fosfomycin-trometamol, corresponding with 3000mg fosfomycin.

6.1.2 Nitrofurantoin Nitrofurantoin belongs to the group of nitrofuranes. Both nitrofurantoin and its metabolites have antibacterial activity, which is enhanced under acidic conditions. Nitrofurantoin is reduced to (active) metabolites by bacterial enzymes. These metabolites inhibit bacterial enzymes that are essential for energy metabolism and inhibit bacterial protein synthesis by binding to ribosomes. Active therapeutic concentrations are only reached in urine and not in other tissues. Several manufacturers produce generic nitrofurantoin in the solid oral form. Three different formulations exist: a macrocrystalized form, nitrofurantoin monohydrate (microcrystals) and a mixture of macrocrystals (75%) and monohydrate 25%) contained in a delayed-release gel matrix (Furabid®) In this trail only Furabid® will be used. Macrocrystals are slower dissolved and absorbed in comparison to the monohydrate.

6.4 Summary of known and potential risks and benefits 6.4.1 Fosfomycin Fosfomycin is contraindicated in patients with known hypersensitivity, but is generally considered safe. The allergic risk is very low.19 Mild and self-limiting gastro-intestinal disturbances, such as diarrhoea, nausea, abdominal pain and dyspepsia are the most common. Headaches, dizziness, upper respiratory tract infections, vaginitis, bacterial and fungal superinfections have been reported. Transient laboratory alterations concerns all blood series (neutropenia, eosinophilia, anaemia, low platelet count, increased liver enzymes, bilirubin), but no renal insufficiency.20 A full list of adverse reactions can be found in the SPC and investigator brochure.

6.4.2 Nitrofurantoin Nitrofurantoin is contra-indicated in patients with renal impairment (creatinine clearance <30 mL/min), a history of lung or liver reaction or peripheral neuropathy, hypersensitivity of nitrofuranen, acute porphyria, G6PD-deficiency, newborns (<3mnd). Common side-effects include superinfection by fungus or resistant bacteria in urogenital tract, benign intracranial hypertension and (temporary) alopecia. A full list of possible side effects can be found in the SPC.

6.5 Description and justification of route of administration and dosage 6.5.1 Fosfomycin In the Netherlands FT has been registered as an oral single dose for symptomatic UTI, on 29 march 1990.21 Before registration, 3 clinical trials on fosfomycin-tromethamine (=fosfomycin trometamol) were performed in the US, demonstrating safety and efficacy.22 Fosfomycin has been marketed as fosfomycin-calcium previously, but as a result of low bio-availability, it is not registered in the Netherlands nor internationally.23 Fosfomycin is also registered as an intravenous administration, fosfomycin-disodium (in the Netherlands registered as Fomicyt) for the treatment of complicated UTI, osteomyelitis, nosocomial infections, lower airway infections, bacterial meningitis, or bacteremia in conjunction with one of the above infections. It is only indicated when it is deemed inappropriate to use the commonly used antibiotics or when these antibiotics have failed.24 6.5.2 Nitrofurantoin In the Netherlands, nitrofurantoin is available in 1) a slow-release formulation in capsules of 100mg, 2) in macrocrystalline form in capsules of 50mg and 100mg and 3) in a suspension of 10mg/mL (100mL per dose).

7. NON-INVESTIGATIONAL PRODUCT

Not applicable.

8. METHODS

8.2 Randomisation, blinding and treatment allocation

Participants will be randomized to fosfomycin FT1:FT1+3:NF1-5 in a 1:1:1 ratio. Randomization will be performed by the UMCU study team after obtainment of informed consent. Randomization will be done centrally and computer generated via the eCRF. The randomization sequence is created by computer software using block randomization with unequal block size.

8.3 Study procedures

The patient's GP will verify the eligibility criteria. Eligible patients interested in participation in the study will receive from the GP, the study's patient information form as well as a link to an information video. Eligible patients will receive a telephone call from the UMCU study team for the informed consent procedure. If the u to the pharmacy to obtain the study medication. If co-medication is used and the study pharmacist is not the patient's regular pharmacist, the study pharmacist will request a medication overview from her regular pharmacist. The patient is asked for consent to do so. If no consent to participate in the study is obtained, the patient will receive antibiotics in accordance to the GPs prescription.

If the patient consents to study participation, the UMCU study team will perform the following study procedures:

• Randomize the patient
• Notify the GP, the pharmacist and Saltro laboratory about participation and allocation, and request GP to send urine samples to Saltro laboratory.
• Fill in the eCRF
• Register the participant for the mobile application

Follow-up data will be collected by a research application for mobile phones. A daily notification will be sent to participants in the subsequent 28 days with three additional questionnaires on day 0, day 7 (adverse events and treatment adherence) and day 28 (absenteeism, treatment satisfaction, hospital admission). For participants not able to use the mobile application, the alternative of a paper diary will be provided including the same questions and with weekly phone contacts to obtain the follow-up data. The content of the mobile application and tasks per day is provided in attachment 1. Secondary endpoints and SAE's that cannot be reported (timely) with the mobile application, will be collected by the GP (e.g. mortality, hospital admission, pyelonephritis).

8.4 Withdrawal of individual subjects Participants can leave the study at any time for any reason if they wish to do so without any consequences. Participants will be asked to provide the reason for withdrawal but are not obliged to disclose it. Participants that want to withdraw from the study will be asked permission to obtain follow-up data from the medical records for secondary endpoints. The investigator can decide to withdraw a participant from treatment for urgent medical reasons. Urgent medical reasons include: an AE that requires immediate change of study medication (e.g. allergic reaction) or a medical complication (e.g. pyelonephritis, urosepsis) that requires a change of study medication. In these cases the participant is however still requested to participate in the daily completion of the questionnaires.

8.4.1 Specific criteria for withdrawal (if applicable) No specific criteria applicable 8.4.2 Replacement of individual subjects after withdrawal Withdrawn participants will not be replaced as the sample size calculation already incorporates this.

9. SAFETY REPORTING 9.1. Temporary halt for reasons of subject safety It is highly unlikely that for reasons of subject safety the trial needs to be halted. If this is however the case, then in accordance to section 10, subsection 4, of the WMO, the sponsor will suspend the study if there is sufficient ground that continuation of the study will jeopardise subject health or safety. The sponsor will notify the accredited METC without undue delay of a temporary halt including the reason for such an action. The study will be suspended pending a further positive decision by the accredited METC. The investigator will take care that all participants are kept informed.

9.5. Safety Committee This study is performed in a (generally) healthy population in primary care. There is extensive experience in the population with both nitrofurantoin and fosfomycin. Both compounds are considered safe. For these reasons a DSMB is not necessary. Instead an annual safety report will be sent to the accredited METC and competent authority (see 9.3)

10. STATISTICAL ANALYSIS

10.1. Primary study parameter(s) Duration of absence of cystitis symptoms will be analyzed using linear regression analysis, adjusted for stratification variable GP practice. Putative violations of the homoscedasticity assumption will be corrected using robust standard errors. Differences in duration of cystitis symptoms with corresponding 98.3% confidence intervals (to account for multiple testing) will be calculated for each comparison. Non-inferiority will be tested for single-dose FT vs. nitrofurantoin and double-dose FT vs. nitrofurantoin, using a non-inferiority margin of 0.8 day, and superiority will be tested for double-dose FT vs. single-dose FT. Hence a two-sided alpha will be used. Two different populations will be analysed: the intention-to-treat (ITT) population and the microbiologically evaluable, per-protocol population. The ITT population includes all patients who were enrolled in the study. The per-protocol analysis excludes patients with a negative urine culture immediately prior to starting antibiotic treatment and patients that have received a new antibiotic treatment since the investigational treatment.

10.2. Secondary study parameter(s) The rates of clinical failure, microbiological failure, relapses, reinfections, complete self-reported therapy adherence, aggravation to pyelonephritis or urosepsis, hospital admission or mortality will be analyzed using generalized linear models, adjusted for stratification variable GP practice, using a binomial identify link function and will be reported as risk difference with 98.3% confidence interval. Differences in duration of adverse events, the days of absenteeism, within the population that is employed (including voluntarily work) or studies, and the degree of satisfaction with the received treatment on a scale from 1 to 5 will be calculated for each comparison and will be reported with corresponding 98.3% confidence intervals. The incidence and nature of severe adverse events will be described.

Correlation between in vivo and in vitro activity will be reported as categorical and essential agreements, which is done separately for the investigational treatments.

10.3. Other study parameters Incremental differences in costs and health outcomes (days with complaints) will be compared among the 3 different treatments and plotted in a cost-effectiveness plane / cost-effectiveness frontier. Uncertainty analysis will be performed with a bootstrap procedure.

10.4. Interim analysis (if applicable) Not applicable. No interim analysis will be performed.

10.5. Missing data Missing observations due to loss to follow-up will be examined to determine both its extent and whether it is missing at random or has a structural underlying reason. If data are missing, the use of appropriate multiple imputation techniques will be considered. Data management processes will include checking for data outliers and unusual data patterns.

11. ETHICAL CONSIDERATIONS

11.1 Regulation statement The trial will be conducted according to the principles of the Declaration of Helsinki 64th WMA General Assembly, Fortaleza, Brazil, October 2013, ICH GCP E6(R2) Guideline, and in accordance with the Medical Research Involving Human Subjects Act (WMO).

11.2 Recruitment and consent

Assistants in participating general practices are instructed to inform patients about the EXFOCY study when a woman seeks help for cystitis.

During the appointment the GP checks if patient meets inclusion criteria and no exclusion criteria are present. The GP asks if the patient would like to participate in principle and prescribes a study medication prescription which is sent digitally to the central pharmacy or printed and given to the patient.

While waiting for the dipstick result at home, information about the study can be read (patient letter and informed consent form), and an information video about the study can be viewed. Waiting time is usually a couple of hours.

Eligible patients will be called by the UMCU study team for the informed consent procedure. After signing the informed consent the patient will be assigned to one of the three arms randomly and given a study number. Under supervision of the UMCU study team the mobile research application is downloaded and together they check if it works properly. The patient is then referred to the pharmacy to obtain the study medication.

If the patient initially agreed to participate, but has second thoughts and does not want to participate when arriving at the pharmacy the study medication prescription will be interpreted as a nitrofurantoin prescription (current 1st choice for cystitis).

11.3 Objection by minors or incapacitated subjects (if applicable) Not applicable.

11.4 Benefits and risks assessment, group relatedness The investigators are investigating 3 arms. Both fosfomycin and nitrofurantoin are already frequently used for cystitis so risks are small, but benefits for FT1+3 arm could be great.

Benefits for NIT1-5 are that it is currently first choice. Benefits for FT1 are that this is the most patient friendly dosing scheme with few side-effects and possibly non-inferior to NIT1-5.

Benefits for FT1+3 are that it is still a patient-friendly dosing scheme with few side-effects and is possibly superior to FT1.

Risks for NIT1-5 are side-effects and a higher risk of forgetting a pill due to the high number to be taken (N=10).

Risks for FT1 are side-effects and potentially a lower efficacy than NIT1-5. Risks for FT1+3 are a higher risk of side effects compared to FT1.

12.6 Public disclosure and publication policy Prospective Trial registration will be performed before inclusion of the first patient in order to comply with requirements from the The International Committee of Medical Journal Editor.

When the study is finished the results will be published in a peer-reviewed journal.

All findings are of interest on a national and international level considering the debate about a single dose of fosfomycin and considering cystitis is such a frequently seen problem.

The study resembles current clinical practice in the Netherlands as much as possible and therefore findings can be easily interpreted and placed into national context.

Before initiation of this study the NHG will be informed about the trial. The findings of this study will actively be communicated to GPs and will help shape an updated Dutch NHG-guidelines on UTIs.

13. STRUCTURED RISK ANALYSIS

13.1 Potential issues of concern See section 13.2 Synthesis

13.2. Synthesis Nitrofurantoin is the current 1st choice antibiotic and fosfomycin is the current 2nd choice for uncomplicated cystitis. Both compounds are frequently prescribed in primary care in the Netherlands and both are in general considered safe for the population studied.

From an effectiveness perspective a single dose of fosfomycin has been considered non-inferior to nitrofurantoin with equal effectiveness in treatment of uncomplicated UTI based on trials with some quality issues. In 2018 it was claimed by Huttner et al. 2018 that fosfomycin is inferior to nitrofurantoin, but it is unknown whether this finding would also apply to the Dutch primary care setting. Therefore a reevaluation and possible optimization of fosfomycin treatment for UTI is warranted.

All study arms have particular benefits and possible disadvantages. FT1: Based on PK/PD studies a single dose FT may be insufficient. The pharmacodynamic parameter most strongly linked to effectiveness is the area under the curve / minimal inhibitory concentration (AUC/MIC) ratio. For this reason it is possible that extended or higher dosing of FT improves outcomes. Additionally, high inter-individual variability in urinary FT concentrations in healthy female volunteers were observed after a single-gift FT with fosfomycin concentrations in urine below the EUCAST breakpoint in two-thirds of the volunteers after 72 hours. It is however a very patient friendly due to its simple dosing and limited side-effects which might explain why FT is increasingly used for treatment of uncomplicated UTI.

FT1+3: An extended dose of fosfomycin could have a higher efficacy than a single dose of FT for above mentioned reason, but also have a higher chance of side-effects.

NIT1-5: Is the 1st choice and possibly has the highest efficacy, but it is also the longest treatment with most pills.

Resistance rates for fosfomycin and nitrofurantoin are similar and low (1% and 2% resistance rate in primary care in 2018) and does not favour fosfomycin nor nitrofurantoin.

In summary, the current recommendation for Dutch GPs is to treat uncomplicated cystitis with a 5-day course of nitrofurantoin (at least 10 pills), and it is unknown whether a single dosage or 2 dosages of FT in 3 days might be equally effective.

Inclusion Criteria:

Adult women (>18 years of age) with a diagnosis of uncomplicated cystitis in primary care

Exclusion Criteria:

• Presence of signs of tissue invasion according to NHG guideline: fever, malaise, chills, flank or perineal pain, signs of sepsis or delirium

• Pregnancy or nursing

• Diabetes Mellitus

• Immunocompromised state

  • Untreated infection with human immunodeficiency virus (hiv)
  • Use of high-dose systemic corticosteroids
  • Use of other immunosuppressive medication (see table below)

• Presence of an indwelling urinary catheter

• History of abnormalities in urinary tract or kidneys

• Neurogenic bladder dysfunction

• UTI in past 28 days before inclusion

• Antibiotic prophylaxis (current or in past 28 days)with nitrofurantoin or fosfomycin or trimethoprim.

• Known GFR <30mL/min

• Contra-indication for nitrofurantoin or fosfomycin use (e.g. allergic reactions, lung or liver reaction or peripheral neuropathy after previous use in clinical history, acute porphyria, G6PD deficiency)

• Current use of an antibiotic for any reason

• Previous inclusion in EXFOCY

• Presence of urine cultures showing resistance for nitrofurantoin or fosfomycin in the last 12 months