Diagnosis and management of male urinary tract infections: a need for new guidelines. Study from a French general practice electronic database

Abstract

Background

The definition and the treatment of male urinary tract infections (UTIs) are imprecise. This study aims to determine the frequency of male UTIs in consultations of general practice, the diagnostic approach and the prescribed treatments.

Methods

We extracted the consultations of male patients, aged 18 years or more, during the period 2012–17 with the International Classification of Primary Care, version 2 codes for UTIs or associated symptoms from PRIMEGE/MEDISEPT databases of primary care. For eligible consultations in which all symptoms or codes were consistent with male UTIs, we identified patient history, prescribed treatments, antibiotic duration, clinical conditions, additional examinations and bacteriological results of urine culture.

Results

Our study included 610 consultations with 396 male patients (mean age 62.5 years). Male UTIs accounted for 0.097% of visits and 1.44 visits per physician per year. The UTIs most commonly identified were: undifferentiated (52%), prostatitis (36%), cystitis (8.5%) and pyelonephritis (3.5%). Fever was recorded in 14% of consultations. Urine dipstick test was done in 1.8% of consultations. Urine culture was positive for Escherichia coli in 50.4% of bacteriological tests. Fluoroquinolones were the most prescribed antibiotics (64.9%), followed by beta-lactams (17.4%), trimethoprim-sulfamethoxazole (11.9%) and nitrofurantoin (2.6%).

Conclusions

Male UTIs are rare in general practice and have different presentations. The definition of male UTIs needs to be specified by prospective studies. Diagnostic evidence of male cystitis may reduce the duration of antibiotic therapy and spare critical antibiotics.

Background

Frequency, diagnosis, classification and treatment of male urinary tract infections (UTIs) remain unclear across countries. Definitions and classifications change according to guidelines (1–9). In 2014, the Society of Infectious Pathology of French Language removed the diagnosis of prostatitis from its guidelines and male UTIs are now classified according to the intensity and tolerance of patients’ symptoms (10). In the Netherlands and the UK, cystitis, prostatitis and pyelonephritis are clearly defined (1–4).

Quality data on the management of male UTIs in primary care are still lacking. The incidence of male UTIs is poorly known and varies by country and definition (7,11–13). The incidence of cystitis is 19.2 per 1000 person-years at risk (11) and increases with age, 78 per 1000 person-years at risk in men aged 85 years or more (14).

Several countries recommend 7 days of treatment with nitrofurantoin for male cystitis (1,4,7,8). Pivmecillinam is commonly used in Nordic countries (15–17). In France, the gold standard is 14 days of treatment with fluoroquinolone (FQ) for male UTIs (10). However, antibiotic resistance to FQ is increasing with a significant risk of treatment failure (18). Improved knowledge of male UTIs in general practice could lead to a more rationalized prescription of antibiotics.

The main objective of this study was to determine the frequency of male UTIs in primary care. Secondary objectives were to describe their clinical presentation and management.

Material and methods

We conducted a retrospective study based on a general practice database created from electronic health records.

Creation of the database

This database was the result of the merger of two existing databases. The first database (PRIMEGE) was created in the region of Provence-Alpes-Côte d’Azur, France. Data came from the consultations of 11 general practitioners (GPs), including supervisors of GP residents at the Faculty of Medicine of the University of Nice Sophia Antipolis. The PRIMEGE database contains the data of 323 928 consultations with 45 944 patients conducted between January 2012 and December 2015.

The second database (MEDISEPT) was created in a single practice in the region of Normandy, France. Data came from the consultations of 14 GPs, including 12 supervisors of GP residents at the Faculty of Medicine of the University of Rouen. The MEDISEPT database contains the data of 305 222 consultations with 20 055 patients conducted between January 2012 and November 2017. In total, the merged database contains the data of 629 204 consultations with 65 999 patients.

Content of the database

For each consultation, we recorded socio-demographic data, reasons for consultation, observations, conclusions, prescribed treatments, possible prescriptions of further examinations and sometimes biology results. The prescriptions were recorded by presentation identifier code with the anatomical, therapeutic and chemical class according to a table provided by Theriaque®.

The reasons for consultation and diagnoses were coded using the 2nd version of the International Classification of Primary Care (ICPC-2), developed by the World Organization of Family Doctors and recognized by the World Health Organization in 2003 as the international gold-standard classification for primary care.

The coding of reasons for consultation and diagnoses could be carried out either by the GP during the consultation or posteriorly using an automatic procedure. In each medical practice, at least one GP was trained in medical coding. The automatic coding procedure was only used when the variables of the consultation were not directly coded by the GPs. It compared the supplied labels with a list of character strings, regardless of case or accents. This procedure was validated by comparing the ICPC-2 codes obtained with those found in the ECOGEN study, as well as those provided by the Health Terminology/Ontology Portal (HeTOP) for the same labels (19). Its main objective was to describe the distribution of reasons for consultation associated with major health conditions managed by GPs. HeTOP is a multilingual multiterminal server to search MeSH keywords to query CISMeF, LiSSa and PubMed (20).

Inclusion and exclusion criteria

Our study included consultations with male patients aged 18 years or more and who had symptoms, diagnoses and/or management compatible with a UTI. A male UTI consultation was defined as a consultation involving at least one anamnestic and/or clinical and/or paraclinical element compatible with a UTI or a diagnostic coding of a UTI. The GP’s diagnosis of UTI was not questioned.

A UTI episode was defined as one or more consultations for a UTI within 8 days or less. When two consultations were included with an interval greater than 8 days, two episodes of UTI were counted.

We differentiated three types of male UTIs:

• -

  Diagnosed UTI: presence of clinical elements compatible with a UTI or a diagnosis coded by the GP without associated antibiotic therapy

• -

  Treated UTI: a diagnosed UTI where urinary antibiotic therapy had been prescribed.

• -

  Confirmed UTI: a treated UTI and presence of a positive urinalysis (positive culture or significant leukocyturia/haematuria).

Data extraction and selection

An extraction algorithm specified the inclusion and exclusion criteria. Data screening was done in two steps: data extraction (extraction of consultations compatible with a male UTI) and data selection (analysis of the extracted data to identify and include consultations for a male UTI).

Data extraction was performed by a public health physician who is also a computer engineer. This was achieved in two parallel ways: from ICPC-2 (e.g., U01 Dysuria) and by lexical analysis (e.g., ‘%dysur%’; Supplementary material). Each term related to a UTI was searched in each section of the consultation (reason, result, examination, conclusion, etc.). All the terms compatible with a UTI were retained. The purpose of data extraction was to obtain the most exhaustive information.

Data selection was carried out independently by two GP researchers with a 96% match. The disagreements were resolved by consensus between the two researchers (Fig. 1—flow chart]).

General comorbidities were screened for in the patient’s history, as well as pathologies of the urinary tract. Drug prescription, bacteriology and imaging were sought at each diagnosis of a UTI.

Statistical analysis

We used χ2, Fisher’s exact test to analyse qualitative variables. We used Student’s test and, if necessary, analysis of variance, for comparisons of quantitative variables. The step of statistical significance was set at P < 0.05. Data were analysed using RStudio® software (version 1.0.153).

Ethics statement

A statement was made to the National Commission for Informatics and Liberties (No. 1585962) regarding the purpose of the database: to evaluate and analyse the practices of GPs. A patient information poster, explaining how to access and correct the data, was available from investigating physicians. Patients were able to refuse authorization for the use of their personal health information.

Results

Characteristics of consultations

Our study included 610 consultations (24%) with 396 male patients (Fig. 1): 450 (73.8%) from the PRIMEGE database and 160 (25.2%) from the MEDISEPT database. Male UTIs accounted for 0.097% of visits and 1.44 visits per GP per year. Considering consultations within an 8-day period as one episode, these 610 consultations corresponded to 494 episodes in 396 patients.

The mean age of patients was 62.5 years (interquartile range 51–76): minimum 18 years, maximum 95 years (Table 1). The most common medical history was a UTI (21.7%), all-cause cancers (19.2%, of which 11.9% were urologic cancers) and benign prostatic hyperplasia (17.8%).

Characteristics of UTI episodes

Of the 494 UTI episodes, 81 were diagnosed but not treated (16.4%), 316 diagnosed and treated (64%) and 97 treated and confirmed (19.6%). The GP specified the site of the infection in 237 cases (48.1%): 176 prostatitis (35.6%), 43 cystitis (8.7%) and 18 pyelonephritis (3.6%; Table 2).

Diagnostic approach

Fever.

Fever was recorded in 68 cases (14%). There was a two-fold chance that a febrile infection would be labelled as prostatitis than any other diagnosis [hazard ratio (HR): 2.15; 95% confidence interval (CI): 1.38; 3.33]. When fever was identified, a urine culture was done in 47 cases (69%).

Functional signs

Functional signs were urinary burning in 77 (16%), dysuria in 63 (13%), urinary frequency in 46 (10%), pain in 40 (8%), haematuria in 26 (5%), malodorous urine in 3 (1%) and urinary urgency in 3 cases (1%). There was no association between the site of pain and the accuracy of diagnosis.

Digital rectal examination

A digital rectal examination (RE) was done in 33/494 episodes (6.6%). It was painful in more than one-third of cases. A diagnosis of prostatitis was twice as common in the case of RE (HR: 1.89; 95% CI: 1.41; 2.52). When RE was painful, 11/13 (84%) were labelled as prostatitis. Half of the episodes with painless RE (10/20) were diagnosed as prostatitis.

Additional examinations

We identified 9/494 urine dipstick tests (1.8%), 324/494 urine cultures (65.6%) and 140/494 blood tests (28%). Prescription of urine culture was associated with a diagnosis of undifferentiated UTIs in 185 (72%) and prostatitis in 109 cases (62%; P = 0.005). Prescription of blood test was associated with a diagnosis of prostatitis in 59 (34%) and undifferentiated UTIs in 75 cases (29%; P = 0.002). Urinary or prostatic ultrasound was found in 49/494 (10%).

Antibiotic treatments

Among the 413/494 treated infections (84%), the treatment duration was specified in 46.3%. Cystitis was associated with short treatment durations (<7 days, P = 0.002) and prostatitis was associated with long treatment durations (>21 days, P = <10^–4). The most commonly prescribed antibiotics were FQs in 262 (64.9%), beta-lactams in 72 (17.4%), trimethoprim-sulfamethoxazole in 51 (11.9%) and nitrofurantoin in 11 cases (2.6%) (Table 3). Only four infections were treated with bi-antibiotic therapy (0.97%). Prostatitis and cystitis were treated predominantly with FQs (72%, P = 0.02 and 47%, not significant, respectively) and pyelonephritis with beta-lactams (54%, P = 0.001).

Bacteriology

Of the 117 positive urine cultures, Escherichia coli was the most common germ (50.4%), then Enterococcus (13.7%), Klebsielle (7.7%), Staphylococcus (6.8%), Pseudomonas (4.3%) and Proteus (3.4%; Table 2).

Discussion

Frequency of UTIs in general practice

The frequency of male UTIs in our study was 0.097% of consultations, making it a rare disease in general practice. This result seems to be consistent with the rare available data (11–14). The incidence of male UTIs in general practice could not be established in our study due to a lack of data on the number of patients living in the two French areas investigated and a lack of information on the actual duration of follow-up of all patients. The French ECOGEN study found a similar frequency of 0.1% of prostatitis in general practice consultations (13).

Elements of semiology

Fever was observed in few UTI episodes. On a consensual basis, fever indicates tissue damage. The majority of international guidelines define a febrile UTI in men as prostatitis (2,5,9,10). Our data show that fever appears to be the only component of the clinical examination associated with a diagnosis of prostatitis. More than half of the diagnoses in the database had no organic localization (bladder, prostate or renal). This is related to the lack of clinical criteria for distinguishing cystitis from pyelonephritis or prostatitis. In a context of suspicion of UTI, spontaneous or costovertebral angle tenderness is strongly suggestive of pyelonephritis. In our study, only one diagnosis of pyelonephritis was associated with low back pain (6%, 1/18) but this result is difficult to analyse because the majority of pyelonephritis cases were from the MEDISEPT database (69%, 9/13) where observation data were lacking. Besides, pain does not appear to be a discriminating factor in identifying renal or prostate disease (21).

We found 43 infections identified as cystitis. In France, the existence of cystitis has long been ignored in comparison with the risk of prostatitis infection (22). Although the term cystitis is absent from the French guidelines, several countries mention the existence of uncomplicated cystitis or cystitis-like infection in men (1,4,6,9).

Diagnostic approach

Rectal examination was little used (7%) in the diagnostic approach. Its usefulness is low and does not change diagnosis or management (23,24). The urine dipstick test was rarely used in consultation. This practice appears to be consistent with the literature (13). The positive predictive value for urine dipstick positive nitrite in men is considered acceptable (78–96%), while its negative predictive value is poor (46–60%) (23–26). Unlike urine dipstick, urine culture was prescribed in at least 65% of cases. It seems essential to prove infection (10). Less than 20% of UTI diagnoses were confirmed on bacteriological examination but there were many missing values and multiple biases and techniques (information bias, structuration bias and mis-structured data) to conclude to the accuracy of urinalyses.

Antibiotic treatments

FQs: a first line to preserve

In our study, FQs were the most prescribed antibiotics (64.9%) and this seems to make sense according to French guidelines. FQs are the first-line treatment of male UTIs in general practice (27,28). Ciprofloxacin and levofloxacin are the two FQs recommended for probabilistic treatment in France (10) and studies in indication (29,30). Ofloxacin was the most prescribed FQ (26.7%) because it was still recommended until 2018 (10). The prescription of ofloxacin is predominant in France but is much lower in Germany and in the UK (27). The results recovered 25 prescriptions of norfloxacin (5% of antibiotics), which is not recommended in male UTIs in France. Norfloxacin does not have marketing authorization for male UTIs and its medical benefit is insufficient (31). Resistance to E. coli with FQ has been increasing for two decades in France (18). FQs must be preserved and their prescription rationalized.

Nitrofurantoin: a French paradox?

In our data, 2.7% of UTIs were treated with nitrofurantoin. Unlike in women, nitrofurantoin is not recommended in male UTIs in France. However, nearly 15% of the prescriptions for nitrofurantoin from 2012 to 2015 were for men (half by GPs) but the reason for their prescription is unknown (32). Two hypotheses can be put forward: a resistance to antibiotics recommended for urinary germs or the analogy of its use in women and in other countries for cystitis. Nitrofurantoin is frequently recommended for mild UTIs in other countries (1,4,7,8) but not in France due to the lack of intervention studies testing it.

Other treatments, not found in our study, are used elsewhere, such as trimethoprim in the UK and Norway (5,7) and pivmecillinam in Norway, Sweden and Denmark (15–17). Pivmecillinam, trimethoprim and nitrofurantoin are associated with a higher risk of antibiotic switching than FQs. However, the rates of switch with these male UTI antibiotics are relatively low (15).

Treatment duration

Antibiotic treatment duration was highly variable. Nearly a quarter of the prescriptions were for less than 7 days. The French guidelines recommend a treatment duration of 14 days minimum regardless of the intensity of UTI symptoms due to the lack of evidence-based studies to propose alternative durations. Several analyses of retrospective databases suggest that short-term treatments lasting less than 7 days would not be at risk of recurrence (28,33,34). Two interventional studies under recruitment, in the USA and in France, will evaluate the efficacy of a shorter treatment duration in male UTIs (35,36).

Strengths and limitations

Limitations

The limitations of these results are mainly inherent to the quality of the databases through two complementary aspects: standardization and data structuring. We developed a data extraction algorithm in order to obtain a good recall to identify all of the UTIs in the database, explaining the many secondary exclusions (76% of the extracted data were excluded). Furthermore, a lack of data structuring may have allowed for information bias. Handwritten prescriptions and home visits reports were probably not included by GPs in the database. We were unable to estimate the rate of underreported or misreported data but it was probably marginal. We decided to include UTIs diagnosed but not treated with antibiotics. This decision was based on the 2014 French guidelines updated in 2018 to defer treatment in the case of minor clinical signs (10,37). We were not able to analyse all the data of consultations within the MEDISEPT database. In addition, the duration of treatments was misinformed within the prescription software for this same base. The absence of urine culture results in this database did not allow us to secondarily invalidate some infections, which may have been included as treated UTIs, resulting in an overestimation of the frequency of male UTIs. Data related to the results of additional examinations (urine culture, ultrasound, etc.) were few in number as filling depended on the synchronization of results provided by a laboratory or radiology office via secured email in patients’ records. Patients’ follow-up could not be analysed in the care pathway (urologist consultations, emergency department admissions or even hospitalizations). These elements would have allowed us to assess the severity and recurrence of UTI episodes in general practice management.

Strengths

The strengths are related to the completeness of the data collected over several years and the 25 medical investigators in two different French regions. The automatic coding tools used improved the quality of the data extracted (38). This sample of UTIs could be used later to test other extraction algorithms or even simulate and predict the performance of an algorithm compatible with a search in a French medico-administrative database. The results are representative of a pragmatic approach to UTI management in general practice.

Conclusion

This study is one of the first to describe male UTIs in general practice, with a large number of patients. We were able to highlight the heterogeneity of the UTIs encountered, with many UTIs with few symptoms and little fever. The urine dipstick test is not used in general practice, unlike urine culture. Urine culture is essential to confirm the diagnosis of infection and to manage antibiotic therapy. In an Irish qualitative study, GPs expressed concern about the relevance of treatment recommendations and prescribing guidelines for the diagnosis of UTIs particularly in elderly fragile patients (39). The definition of male UTIs needs to be specified by prospective studies that could support a diagnosis of cystitis in men. The duration and type of antibiotic prescribed could then be modulated according to the intensity of UTI symptoms. Prospective studies in general practice are essential to confirm these hypotheses.

Acknowledgements

The authors are grateful to Nikki Sabourin-Gibbs, Rouen University Hospital, for her help in editing the manuscript, the PRIMEGE team for their collaboration and to Prof. Manuel Etienne for his expertise and insights on male urinary tract infections.

Declaration

Funding: none.

Ethical approval: any necessary ethical approval(s).

Conflict of interest: the authors declare that they have no conflict of interest in relation to the content of this article.

Data availability

The agreement was obtained from the National Commission for Informatics and Liberties (No. 1585962).

References