It is estimated that 10 million people could die worldwide by 2050 from infections caused by multidrug-resistant pathogens if urgent measures are not taken to combat the problem of antimicrobial resistance, mainly among gram-negative bacteria and especially enterobacteria [1]. Antibiotic resistance is an individual health and public health problem; increasing community resistance can prolong the infectious process, increase the risk of infecting other individuals, and in the long term, lead to increased costs. Inappropriate use of antibiotics increases the frequency of related and sometimes serious side effects and causes increased mortality [2,3]. Unnecessary use of antimicrobials is the most important cause of resistance generation in the community, and it is estimated that 80% of antimicrobials are prescribed in primary health care (PHC) [4].

Urinary tract infections (UTIs) are a common infection in PHC, and most patients are treated with antibiotics [5-7]. In Spain, UTIs represent the second most frequent cause of antibiotic prescriptions [8]. Almost 60% of women experience at least one episode during their lifetime [9,10]. Antibiotic treatment of uncomplicated UTIs is, in most cases, empirical, and the selection of the antimicrobial is made according to the most frequently involved pathogen and local resistance. Several studies have suggested that up to 50% of antibiotic prescriptions are inappropriate [11-13]. There is a tendency to use broad-spectrum antibiotics, which do not usually provide substantial improvements in efficacy over narrower-spectrum antibiotics and instead favor increased resistance [4].

Approximately 80% of cases of uncomplicated UTI are caused by Escherichia coli, so empirical treatment should cover this pathogen [14-16]. The resistance of uropathogens to classical antibiotics has increased significantly in recent years in Spain [17]. Various studies in Spain have described resistant E coli infections and shown that a high percentage are resistant to penicillins. A study published in 2017 that was conducted in our setting among patients with community-acquired pyelonephritis showed that 23% were resistant to ciprofloxacin, while 1% were resistant to fosfomycin [18]. In the program of surveillance of nosocomial infections in the hospitals of Catalonia (Vigilància de les infeccions nosocomials als hospitals de Catalunya [VINCat; all non-English abbreviations are in Catalan unless otherwise noted]), a local database from Catalonia (an autonomous community in Spain) that is part of the Antimicrobial Optimization Program (Programa d’optimització de l’ús d’antimicrobians [PROA]), it was shown that in 2020 E coli had a high sensitivity to fosfomycin (96.6%) and nitrofurantoin (98.5%); by contrast, its sensitivity to the amoxicillin/clavulanic acid combination, cotrimoxazole, and quinolones was much lower (75%, 74%, and 72%, respectively) [19]. Resistance of enterobacteria to third-generation cephalosporins, mediated by the production of extended-spectrum β-lactamases, is a growing problem in E coli and Klebsiella pneumoniae strains. Data from the VINCat PROA 2020 report show that the proportions of β-lactamase-producing E coli and K pneumoniae urinary tract infections in the community in Catalonia were 9% and 11%, respectively [19]. UTIs caused by resistant microorganisms are associated with longer symptom duration than infections caused by sensitive strains, and treatment is more likely to fail [20].

According to the recommendations of the Infectious Diseases Society of America, empirical antibiotherapy should not be used when resistance rates exceed 20% for all strains [21]. This means that the use of combined amoxicillin/clavulanic acid, as well as quinolones and cotrimoxazole, should no longer be recommended for empirical treatment of UTIs in our country. Current guidelines in PHC recommend prescribing a single dose of 3 grams of fosfomycin trometamol or nitrofurantoin for 5 to 7 days [22,23]. The rationale for this strategy is to specifically target the etiologic agents causing acute cystitis based on knowledge of the local resistance patterns to antimicrobials of the most frequent uropathogens [24]. In recent years, the use of fosfomycin as the therapy of choice for these infections has increased significantly in Spain. However, more than half of all physicians prefer the use of short-course therapies to single doses [25]. One of the reasons mentioned by professionals for not giving first-line treatments is to prevent complications, the most frequent being acute pyelonephritis; however, there is currently no predictive scale available to determine the risk that a UTI will worsen to pyelonephritis. Only 26% of antibiotic prescriptions for uncomplicated UTI were for fosfomycin trometamol (3-gram single dose) or nitrofurantoin in primary care management in Barcelona in the entire year 2020 (unpublished data).

Another dilemma faced by practitioners is the recommendation to give suppressive antibiotic treatment lasting more than 6 months to women who have repeated UTIs (defined as more than 2 UTIs in the last 6 months or more than 3 in the last year) [26]. However, this practice is not exempt from potential risks. In a recent article, a clear correlation was reported between a history of antibiotic use in the previous 2 years and the severity of COVID-19 (with severity defined as a variable combining death, hospitalization, and the presence of pneumonia), based on data from the Information System for the Development of Research in Primary Care (Sistema d’informació per al desenvolupament de la investigació en atenció primària [SIDIAP]) [27]. This correlation was very strong when antibiotics were taken in the previous 2 months, when there were more than 5 antibiotic regimens in the last 2 years, and when antibiotics were used that are critical and should be reserved for special cases, such as cephalosporins and quinolones. In addition, prudent use of antibiotics is necessary due to the increase in antibiotic resistance [3,28].

Urine culture requests are not always made according to guidelines. As early as 2000, it was documented that 44% of requests were made for cases of asymptomatic bacteriuria, a condition for which they should not be ordered and there should be no treatment [29]. In 2014, 92% of urine cultures that were ordered for a second time for patients with a UTI were not necessary [30]. In 2019, a range of 21% to 42% of requests for urine cultures were for uncomplicated UTIs in Barcelona city primary care management (in Spanish: gerencia de atención primaria de Barcelona ciudad), depending on the primary care center (unpublished data). The ordering of urine cultures for women with cystitis is only recommended in cases of (1) uncomplicated cystitis that, despite adequate antibiotic treatment, remains symptomatic (posttreatment urine cultures are used), (2) recurrent cystitis (pretreatment urine cultures are used), and (3) cystitis in pregnancy (both pretreatment and posttreatment urine cultures are used) [14]. The use of these tests when they are not necessary leads to antibiotic overprescription; moreover, since in many cases they are requested for cases of bacteriuria in elderly people who have been treated with antibiotics many times, the results of antibiograms recommend the prescription of second-line antibiotics for uropathogens that are often multidrug-resistant.

For all the above reasons, we propose the current protocol for the Urinary Tract Infections in Catalonia (Infeccions del tracte urinari a Catalunya [ITUCAT]) study, consisting of 4 work packages (WPs) that will be implemented successively in 3 phases, to evaluate different objectives related to UTIs in the adult population in Catalonia and investigate specific aspects of this condition in this population. The first phase of the project, which is presented in this paper, includes the first 2 WPs. WP 1 will evaluate the management of UTIs in Catalonia, and WP 2 aims to evaluate the correlation of antibiotic use for repeated UTIs with the presence of infectious complications in adult women. Subsequently, we will carry out the second phase with the data obtained from the first phase. At this stage, the aim is to create a scale to predict acute pyelonephritis in patients with UTI by evaluating clinical variables; this will be WP 3. Finally, in the third phase, we will perform WP 4, which aims to determine the degree of inadequacy in requests for urine cultures based on clinical practice guidelines and to establish the basis for a training intervention.

The study described in this paper corresponds to the first phase, which includes WP 1 and WP 2.

We expect to describe the epidemiology of UTIs in Catalonia during the period from 2012 to 2021, including the characteristics of the population, diagnoses of UTI, the use of urine cultures or laboratory tests, and different antibiotics prescribed according to treatment guidelines for dose and duration. Furthermore, we will report the degree of adequacy of the prescribed treatments and the clinical evolution of the patients (ie, the rates of clinical resolution, recurrence, hospitalization, and complications, such as acute pyelonephritis, urosepsis, pneumonia, COVID-19, and death).

The first phase of the ITUCAT project is expected to be completed during the second half of 2023. Phase 2 will start after the data from phase 1 are obtained and used to create the predictive scale for acute pyelonephritis in patients with UTI; phase 2 is expected to be completed in 2024. Phase 3 will be carried out during the second half of 2023 and during 2024 to evaluate the degree of inadequacy in requests for urine cultures according to the clinical practice guidelines and establish the basis for a training intervention.

The results obtained will be presented according to the Reporting of Studies Conducted Using Observational Routinely Collected Health Data Statement for Pharmacoepidemiology (RECORD-PE) recommendations [42].

We expect to find inadequate management in a high percentage of UTI cases, considering that on many occasions patients are treated with second- or third-line antibiotic therapies with a preference for the longest regimens. Likewise, the use of suppressive antibiotic therapy as a prophylaxis in cases of recurrent UTI is expected to vary in our study.

In addition, we aim to determine whether women with recurrent UTIs treated with antibiotic suppressive therapy have a higher incidence and severity of potentially serious future infections, with special attention to acute pyelonephritis, urosepsis, COVID-19, and pneumonia, in comparison to women who are treated with antibiotics only when they present with a UTI.

All the information obtained in this first phase will allow us to continue with the following WPs of the project. Furthermore, if we find the inappropriate use of antibiotics, we will work on the creation of proposals to promote the correct treatment of UTIs in our environment according to clinical guidelines [40,41]. In this way, we will promote responsible management of UTIs in PHC among health care personnel and patients. In addition, our study will provide valuable information for health care policy makers and national programs to optimize the use of antimicrobials.

Because this is an observational study based on data from administrative databases, it will not allow causality analysis. However, the results will be derived from health care carried out under conditions of routine clinical practice with data from electronic primary-care registries, including the SIDIAP, CMBD-HA, and CMBD-UR databases, which have proven to be valid and representative of the population in numerous previous studies [31,34,35]. Some limitations of these studies arose from underreporting of some variables and diagnoses that we will try to avoid with the use of the different databases. Another limitation of the databases is the lack of links between diagnoses and treatments. In order to deal with bias from this specific limitation, we plan to use proxy values available in the databases, such as dates of prescriptions and dates of diagnoses. An inherent limitation of observational studies is the presence of unmeasured confounding variables. A complete adjustment for all possible confounders would require detailed information on clinical parameters, lifestyles, socioeconomic conditions, use of over-the-counter medications, and personal circumstances, which is not routinely present in electronic health records, causing residual confounding bias. Nevertheless, this bias would affect all included populations equally. To minimize confounding, we will conduct sensitivity analyses and use external adjustments or specific statistical methods when appropriate [43]. We will also take this aspect into account in the scientific discussion of the results, as it is an inherent limitation of studies conducted with electronic health records.