Tratamiento antimicrobiano para la prostatitis bacteriana crónica
Resumen
Antecedentes
La prostatitis bacteriana crónica (PBC) se diagnostica con frecuencia en los hombres en edad fértil y se caracteriza por una variedad de síntomas incapacitantes, incluido dolor en la zona pelviana (por ejemplo, perineo, testículos), síntomas de vaciamiento (aumento de la polaquiuria y tenesmo, también de noche; dolor o malestar a la micción) y disfunción sexual. La cura de la PBC se puede intentar mediante el tratamiento a largo plazo con agentes antibacterianos, pero las recurrencias son frecuentes. Pocos agentes antibacterianos pueden difundir al tejido prostático y lograr concentraciones suficientes en el sitio de la infección. Estos agentes incluyen fluoroquinolonas, macrólidos, tetraciclinas y trimetoprima. Después de la introducción de las fluoroquinolonas en la práctica clínica se han realizado varios estudios para optimizar el tratamiento antimicrobiano de la PBC y mejorar las tasas de erradicación y el alivio de los síntomas.
Objetivos
Evaluar y comparar la eficacia y los efectos perjudiciales de los tratamientos antimicrobianos para la prostatitis bacteriana crónica.
Métodos de búsqueda
Se realizaron búsquedas en el Registro Cochrane Central de Ensayos Controlados (Cochrane Central Register of Controlled Trials, CENTRAL), MEDLINE (PubMed), EMBASE, otras bases de datos nacionales e internacionales y resúmenes de actas de congresos el 8 de agosto de 2012.
Criterios de selección
Se incluyeron todas las comparaciones controladas aleatorias de un agente antimicrobiano versus placebo o uno o más agentes antimicrobianos comparadores, combinados o no con fármacos no antimicrobianos. También se incluyeron los ensayos que compararon diferentes dosis, duraciones del tratamiento, frecuencias de las dosis o vías de administración de los agentes antimicrobianos. Se excluyeron los estudios en los que los pacientes no se diagnosticaron según criterios internacionalmente recomendados o no se les realizaron pruebas segmentadas de las vías urinarias bajas.
Obtención y análisis de los datos
Dos revisores extrajeron los datos del estudio de forma independiente. Los resultados del estudio fueron eficacia microbiológica (erradicación de los agentes patógenos), eficacia clínica (curación o mejoría del síntoma, o puntuaciones de los síntomas) a las visitas para pruebas de curación o al seguimiento, o ambas, y efectos adversos del tratamiento. Los resultados secundarios incluyeron las tasas de recurrencia microbiológica.
El análisis estadístico se realizó mediante un modelo de efectos fijos para los resultados microbiológicos y un modelo de efectos aleatorios para los resultados clínicos y los efectos adversos. Los resultados se expresaron como cocientes de riesgos para los resultados dicotómicos (con intervalos de confianza del 95%) o como diferencias de medias estandarizadas para las variables continuas o no dicotómicas.
Resultados principales
Se identificaron 18 estudios con un total de 2196 pacientes asignados al azar. Se compararon las fluoroquinolonas orales ciprofloxacino, levofloxacino, lomefloxacino, ofloxacino y prulifloxacino. No hubo diferencias significativas en la eficacia clínica o microbiológica ni en la tasa de efectos adversos entre estas fluoroquinolonas. En la prostatitis clamidiácea, (i) la azitromicina mostró mejores tasas de erradicación y tasas de curación clínica comparada con ciprofloxacino, sin diferencias significativas con respecto a los efectos adversos; (ii) la azitromicina fue equivalente a la claritromicina, microbiológica y clínicamente; (iii) el prulifloxacino pareció mejorar los síntomas clínicos, pero no las tasas de erradicación, en comparación con la doxiciclina. En la prostatitis por ureaplasma las comparaciones de ofloxacino versus minociclina y azitromicina versus doxiciclina mostraron perfiles microbiológicos, clínicos y de toxicidad similares.
Conclusiones de los autores
La eficacia microbiológica y clínica, así como el perfil de efectos adversos de diferentes fluoroquinolonas orales son equivalentes. No se pueden establecer conclusiones con respecto a la duración óptima del tratamiento con fluoroquinolonas en el tratamiento de la PBC causada por agentes patógenos tradicionales.
Los agentes antimicrobianos alternativos probados para el tratamiento de la PBC causada por agentes patógenos tradicionales son cotrimoxazol, betalactámicos y tetraciclinas, pero no fue posible extraer pruebas definitivas con respecto a la función de los antibióticos no fluoroquinolonas en el tratamiento de la PBC causada por los agentes patógenos tradicionales.
En los pacientes con PBC causada por agentes patógenos intracelulares obligatorios, los macrólidos mostraron mayores tasas de curación microbiológicas y clínicas en comparación con las fluoroquinolonas.
PICO
Resumen en términos sencillos
Intervenciones para tratar la infección crónica de la glándula de la próstata (prostatitis bacteriana crónica)
La prostatitis bacteriana crónica (PBC) incluye la infección y la inflamación de la glándula de la próstata en los hombres de todas las edades. Puede causar problemas al orinar que incluyen malestar y dolor, aumento de la frecuencia y la urgencia, o problemas con la evacuación de la vejiga. Las bacterias que infectan la próstata causan la PBC. Estas bacterias pueden ser de transmisión sexual. Para curar la PBC, los antibióticos se deben administrar durante períodos prolongados (cuatro semanas o más), pero no siempre se garantiza una curación permanente. Otros fármacos se pueden combinar con los antibióticos para mejorar los síntomas de la PBC. Esta revisión encontró que las fluoroquinolonas como ciprofloxacino, levofloxacino, lomefloxacino, ofloxacino o prulifloxacino tienen efectos y tasas de éxito equivalentes en los pacientes con PBC. Si se sospecha que bacterias atípicas como la clamidia son la causa de la PBC, los antibióticos macrólidos como la azitromicina pueden lograr mejores resultados comparados con la fluoroquinolona ciprofloxacino. Hay que considerar que algunos de los estudios realizados son de calidad deficiente o se han realizado en un escaso número de participantes. Se necesitan más estudios que se centren en los agentes nuevos o en dosis optimizadas de los antibióticos actualmente prescritos.
Authors' conclusions
Summary of findings
| Levofloxacin versus ciprofloxacin for chronic bacterial prostatitis | |||||
| Patient or population: patients with chronic bacterial prostatitis | |||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No. of participants | Quality of the evidence | |
| Assumed risk | Corresponding risk | ||||
| Ciprofloxacin | Levofloxacin | ||||
| Microbiological efficacy ‐ pathogen eradication | 667 per 1000 | 787 per 1000 | RR 1.18 | 669 | ⊕⊝⊝⊝ |
| Clinical efficacy ‐ cure or improvement at end of treatment | 722 per 1000 | 838 per 1000 | RR 1.16 | 669 | ⊕⊕⊝⊝ |
| Clinical efficacy ‐ cure or improvement at follow‐up (6 months) | 710 per 1000 | 823 per 1000 | RR 1.16 | 669 | ⊕⊝⊝⊝ |
| Adverse effects of treatment ‐ any adverse effects | 266 per 1000 | 229 per 1000 | RR 0.86 | 785 | ⊕⊕⊕⊝ |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence | |||||
| 1Bundrick 2003 ‐ high risk of reporting bias. | |||||
| Lomefloxacin versus comparator fluoroquinolone for chronic bacterial prostatitis | ||||||
| Patient or population: patients with chronic bacterial prostatitis | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No. of participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Comparator fluoroquinolone | Lomefloxacin | |||||
| Microbiological efficacy ‐ pathogen eradication at follow‐up (6 months) | 804 per 1000 | 771 per 1000 | RR 0.96 | 116 | ⊕⊕⊝⊝ | |
| Clinical efficacy ‐ cure or improvement at end of treatment | See comment | See comment | Not estimable | 0 (0) | See comment | No study reported or provided useable data for this outcome |
| Clinical efficacy ‐ cure or improvement at follow‐up (6 months) | See comment | See comment | Not estimable | 0 (0) | See comment | No study reported or provided useable data for this outcome |
| Adverse effects of treatment ‐ any adverse effects | 212 per 1000 | 135 per 1000 | RR 0.64 | 215 | ⊕⊕⊕⊝ | |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 The comparator fluoroquinolone was ofloxacin (Koff 1996) or ciprofloxacin (Naber 2002). | ||||||
| Ciprofloxacin versus comparator fluoroquinolone for chronic bacterial prostatitis | |||||
| Patient or population: patients with chronic bacterial prostatitis | |||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No. of participants | Quality of the evidence | |
| Assumed risk | Corresponding risk | ||||
| Comparator fluoroquinolone | Ciprofloxacin | ||||
| Microbiological efficacy ‐ pathogen eradication at end of treatment | 806 per 1000 | 733 per 1000 | RR 0.91 | 851 | ⊕⊝⊝⊝ |
| Clinical efficacy ‐ cure or improvement at end of treatment | 879 per 1000 | 791 per 1000 | RR 0.9 | 851 | ⊕⊝⊝⊝ |
| Clinical efficacy ‐ cure or improvement at follow‐up (6 months) | 808 per 1000 | 752 per 1000 | RR 0.93 | 851 | ⊕⊝⊝⊝ |
| Adverse effects of treatment ‐ any adverse effects | 212 per 1000 | 246 per 1000 | RR 1.16 | 967 | ⊕⊕⊕⊝ |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence | |||||
| 1 The comparator fluoroquinolone was levofloxacin (Bundrick 2003; Zhang 2012) or lomefloxacin (Naber 2002). | |||||
| Levofloxacin versus comparator fluoroquinolone for chronic bacterial prostatitis | ||||||
| Patient or population: patients with chronic bacterial prostatitis | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No. of participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Comparator fluoroquinolone | Levofloxacin | |||||
| Microbiological efficacy ‐ pathogen eradication | 674 per 1000 | 755 per 1000 | RR 1.12 | 758 | ⊕⊝⊝⊝ | |
| Clinical efficacy ‐ cure or improvement at end of treatment | See comment | See comment | Not estimable | 0 (0) | See comment | No study reported or provided useable data for this outcome |
| Clinical efficacy ‐ cure or improvement at follow‐up (6 months) | See comment | See comment | Not estimable | 0 (0) | See comment | No study reported or provided useable data for this outcome |
| Adverse effects of treatment ‐ any adverse effects | 258 per 1000 | 227 per 1000 | RR 0.88 | 874 | ⊕⊕⊕⊝ | |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 The comparator fluoroquinolone was ciprofloxacin (Bundrick 2003; Zhang 2012) or prulifloxacin (Giannarini 2007). | ||||||
Background
Description of the condition
Prostatitis syndromes represent the most frequent urological diagnosis in men below 50 years of age, and they are the third most common diagnosis among individuals beyond that age (Collins 1998).
The prevalence of chronic prostatitis‐like symptoms ranges between 2% and 13% worldwide, depending on the type of study and on the population examined (Bartoletti 2007;Ejike 2008;Ferris 2010;Krieger 2008;Liang 2009;Mehik 2000;Nickel 2001; Rizzo 2003;Wallner 2009). Analysis of the United States (US) Kaiser Permanente Northwest database (Portland, Oregon) showed that between 2002 and 2004 the incidence of physician‐diagnosed prostatitis was 4.9 per 1000 person‐years (Clemens 2005).
The age‐adjusted annualized visit rate for prostatitis is 17,980 per million population in the US, and prostatitis accounted for a total of 8,021,396 physician office visits (with any diagnosis) between the years 1992 and 2000. The total US spending for the diagnosis and management of prostatitis in year 2000, not including pharmaceutical expenses, was 84 million USD (McNaughton‐Collins 2007; Pontari 2007). The economic impact of visits and tests for prostatitis in the US, assessed in year 2009, ranged between 3017 USD (Medicare rates) and 6534 USD (non‐Medicare rates) per patient per year (Clemens 2009).
Chronic prostatitis syndromes are traditionally classified as 'bacterial' or 'abacterial'.
According to the most recent National Institutes of Health ‐ National Institute of Diabetes and Digestive and Kidney Diseases (NIH‐NIDDK) consensus definition, category II chronic bacterial prostatitis (CBP) occurs when patients experience recurrent symptomatic episodes of urinary tract infection caused by the same organism (usually E. coli, or another Gram‐negative organism (for example, Klebsiella spp.,Proteus spp.,Pseudomonas spp.) or Enterococcus faecalis). Between symptomatic episodes of bacteriuria, lower urinary tract cultures can document an infected prostate gland as the focus of these recurrent infections (Krieger 1999). Besides these commonly recognized pathogens, Staphylococcus aureus is frequently included among the causative agents of CBP (for example, Naber 2008; British National Guidelines: www.bashh.org/guidelines). Other bacteria have been investigated in recent years, but a general consensus on their pathogenic role in CBP is still awaited.
The vast majority of men with chronic prostatitis (about 90% of all prostatitis cases) (Lipsky 2010; McNaughton‐Collins 2007) present with pelvic pain and voiding symptoms without evidence of bacterial infection, and are diagnosed with chronic (abacterial) prostatitis/chronic pelvic pain syndrome (CP/CPPS, NIH‐NIDDK category III) (Krieger 1999;McNaughton‐Collins 2007). CP/CPPS is mainly characterized by pain in the perineum, prostate, rectum, penis, testicles and abdomen. It is often associated with dysuria (painful voiding), with symptoms of obstruction on voiding (for example, hesitancy, weak or intermittent stream), with irritative symptoms (for example, increased frequency, urgency, nocturia (night‐time urination)), and sometimes with sexual dysfunction (Mehik 2001).
Symptoms of CBP and CP/CPPS frequently overlap. Since the clinical presentation of patients with CBP or CP/CPPS is similar, and given that there is no gold standard diagnostic test for the latter, CP/CPPS is mainly diagnosed by excluding the presence of category II CBP (McNaughton‐Collins 2007).
Although patients suffering from prostatitis with a recognized bacterial etiology are only 5% to 10% of all men showing symptoms of chronic prostatitis (McNaughton‐Collins 2007), bacterial infection is reputed to be a possible pathogenic factor in the early 'etiological pathway' of CP/CPPS (Daniels 2007; Nickel 2010;Shoskes 2009). However, CP/CPPS is defined and diagnosed as an abacterial form of chronic prostatitis and antimicrobial treatment is not effective for this specific syndrome (Cohen 2012).
Microbiological diagnosis of CBP is based on finding substantially lower (one tenth or less) bacterial counts in urine specimens from the urethra (first‐voided urine, or VB1) and bladder (midstream urine, or VB2) compared with counts in prostatic secretions expressed during prostatic massage (EPS) or in post–massage voided urine (VB3). Such segmented microbiological analysis of men's lower urinary tract is commonly referred to as the 'four‐glass test' according to Meares and Stamey (Stamey 1981). Although never validated in a randomized setting, the four‐glass test is considered to be a standard analytical procedure for diagnosing CBP as well as for discriminating between CBP and CP/CPPS. A test based on bacteriological culture of the pre‐massage and post‐prostatic massage voided urine (PPMT, or 'two‐glass' assay) has been proposed as a simplified alternative to the four‐glass test (Nickel 2006). Although a study comparing the two tests in patients diagnosed with CP/CPPS showed that the PPMT could detect uropathogens in fewer cases (44%) compared to the traditional four‐glass assay (Nickel 2006), the former is considered a preferable alternative to simple urine or semen cultures.
Description of the intervention
The therapy of CBP is based on the administration of antibacterial agents for several weeks.
Fluoroquinolones are currently indicated as first‐choice antibacterial agents for treatment of category II CBP. International recommendations and guidelines indicate a four to 12‐week course of ciprofloxacin, lomefloxacin, ofloxacin, levofloxacin or norfloxacin for the eradication of susceptible pathogens (European Association of Urology (EAU) Urological Infections Guidelines: www.uroweb.org/guidelines/online‐guidelines/; United Kingdom (UK) National Guidelines: www.bashh.org/guidelines; Canadian guidelines: www.phac‐aspc.gc.ca/std‐mts/sti‐its/guide‐lignesdir‐eng.php; Lipsky 2010;Wagenlehner 2007).
Trimethoprim, combined or not with sulfamethoxazole, was formerly the most prescribed drug for the treatment of CBP (Meares 1975). Due to the low eradication rates achieved with trimethoprim, this drug is now indicated as a second‐choice agent in case of bacterial resistance to fluoroquinolones or in case of poor tolerability of the first‐choice agents (EAU Guidelines: www.uroweb.org/gls/pdf/18_Urological%20infections_LR.pdf).
Macrolides and tetracyclines are also recommended for treatment of CBP, but their use is presently restricted to special indications (for example, chlamydial infection) (EAU Guidelines: www.uroweb.org/gls/pdf/18_Urological%20infections_LR.pdf; Lipsky 2010; Nickel 2008b).
Patients with frequent recurrences may be placed on antibiotic prophylaxis for several months (for example, low‐dose co‐trimoxazole). However, evidence‐based proof of efficacy of such a strategy is lacking.
How the intervention might work
The number of antibacterial agents suitable for treatment of category II CBP is very limited. Fluoroquinolones, trimethoprim and macrolides are among the few antibacterial agents that can penetrate the prostate sufficiently to reach levels exceeding the minimal concentrations inhibiting the growth (MIC) of most infecting pathogens (Foulds 1991).
Lipophilicity and a high pKa (acid dissociation constant) are considered important features of antibacterial agents for the treatment of CBP (Shoskes 2001). To achieve suitable prostatic concentrations, a drug must be sufficiently lipophilic to cross the many barriers separating the prostatic vasculature from the target site of action, and to reach the pathogens infecting the prostatic glands and ducts and, in some cases, the intracellular compartments (Naber 2003;Perletti 2009).
The pH at the site of action can also affect the pharmacodynamic properties of antibacterial agents. The milieu of the infected human prostate is alkaline (pH = 8.34) (Naber 2003). It has been demonstrated that alkalinization of the pH can significantly decrease (10‐ to 30‐fold) the MICs of fluoroquinolones against E. coli and other uropathogens (Aagaard 1991;Gesu 1987; Kamberi 1999). The activity of macrolides is also influenced by the pH at the site of action; for example, the MIC of azithromycin (pKa = 9.5) against Staphylococcus aureus is 64, 1 and 0.03 at pH 6, 7 and 8, respectively (Dalhoff 2005).
Why it is important to do this review
Firstly, most current therapeutic recommendations for CBP are based on data from randomized trials, non‐randomized clinical evidence, or on the clinical experience and opinion of leading experts. The fact that contemporary guidelines are not based on systematic reviews of the literature and the meta‐analysis of available data represents a major limitation in this regard. One example is 'suppressive' long‐term therapy with low‐dose trimethoprim. This recommendation is not substantiated by clinical data.
Secondly, the current antibacterial dosing regimens for CBP are mainly based on 'trial‐and‐error' empirical strategies, regimens adopted for other infectious diseases, or on safety data. Pharmacokinetic and pharmacodynamic parameters for the treatment of CBP, focusing on dosage issues, are almost non‐existent. The results of studies comparing different doses of antibacterial agents or different durations of therapy should be analyzed and reviewed.
Thirdly, the adjuvant effect of compounds administered in combination with antibacterials (for example, alpha‐adrenoceptor blockers) is controversial. Few of these combinations have been tested in the framework of randomized controlled trials (RCTs). The results of these studies must be thoroughly analyzed to improve clinical decision‐making and patient management.
Fourthly, the efficacy of interventions different from established long‐term oral antibacterial regimens (for example, intraprostatic injection of antibacterial agents) is debated. The results of RCTs involving alternative administration routes for antibiotics must be reviewed and thoroughly analyzed to improve clinical decision‐making and patient management.
Finally, the world‐wide increasing fluoroquinolone resistance in Gram‐negative pathogens poses new therapeutical problems also in the antibacterial treatment of CBP. For example, the activity of second‐generation fluoroquinolones like ciprofloxacin is hampered by the novel resistance determinant aac (6')‐Ib‐cr, whereas unique structural features make molecules like levofloxacin unaffected by aac (6')‐Ib‐cr. Thus, fluoroquinolone clinical trials published in the past must be reviewed in a contemporary perspective. Moreover, studies published in the past may have lost relevance due to diffuse drug resistance.
In conclusion, a systematic review may help improve current therapeutic guidelines on the basis of the evidence available from quality RCTs, and may improve the nature and grade of clinical recommendations and the management of patients affected by CBP.
Objectives
To assess and compare the efficacy and harms of antimicrobial treatments for CBP.
Methods
Criteria for considering studies for this review
Types of studies
All RCTs in which antimicrobial therapy was used to treat CBP.
Types of participants
Patients with category II (NIH‐NIDDK) CBP (Krieger 1999), or with CBP according to the earlier classification by Drach et al (Drach 1978).
According to the Drach definition, CBP is diagnosed when pathogenic bacteria are recovered in significant numbers from a purulent prostatic fluid in the absence of concomitant urinary tract infection or significant systemic signs (Drach 1978).
A clinical diagnosis of CBP is mainly based on three criteria: a history of CBP, current clinical signs and symptoms of prostatitis, and laboratory evidence of prostatic infection in expressed prostatic secretions or post‐massage voided urine.
Studies focusing on patients affected by category I acute bacterial prostatitis, category III CP/CPPS, or category IV asymptomatic inflammatory prostatitis (NIH‐NIDDK criteria), or by acute bacterial prostatitis, chronic non‐bacterial prostatitis or prostatodynia (Drach 1978 classification) were excluded.
Studies not providing microbiological findings from adequate lower urinary tract segmented tests (Meares and Stamey '4‐glass' test, '2‐glass' pre‐ and post‐massage test) were excluded. Studies including patients with poorly defined infections or conditions (for example, unclassified 'prostatitis' or 'chronic prostatitis'; 'prostato‐epididimo‐vesiculitis'; 'genital tract infection including prostatitis'; etc.) were excluded.
Types of interventions
-
We considered all randomized controlled comparisons of one antimicrobial agent versus placebo, versus a different antimicrobial agent, or versus two or more combined antimicrobial agents.
-
Trials comparing different doses, different treatment durations, different dosing frequencies, or different routes of administration of antimicrobial agents were also considered to be acceptable for inclusion, as these regimens are likely to differ in their pharmacodynamic and pharmacokinetic properties and thus may differ in their efficacy.
-
We also considered randomized controlled comparisons of antimicrobial agents alone with antimicrobial agents combined with non‐antibacterial drugs or physical interventions aimed at improving the microbiological or clinical efficacy of therapy as well as drug pharmacokinetics.
Types of outcome measures
Primary outcomes
-
Microbiological efficacy, defined as yielding at test‐of‐cure (TOC) visit sterile cultures of expressed prostatic secretions or post‐massage urine, or positive cultures with a bacterial load inferior to a defined threshold (e.g., 103 colony‐forming units (CFU)/mL).
-
Clinical efficacy, defined as cure, resolution or improvement of signs and symptoms of CBP at the TOC visit or at follow‐up, or assessed with strategies based on subjective or objective findings:
-
subjective clinical outcomes included symptom scores, bother scores, quality of life (QoL) scores, global urinary or systemic symptom reports, or patient self‐declared status (e.g., improved, unchanged or worsened);
-
objective clinical outcomes included the results of urodynamic or sonographic evaluations, prostate examination (tenderness, size, consistency, symmetry), microscopy of specimens of lower urinary tract segmented tests (white blood cell counts), biochemical markers (e.g., prostate‐specific antigen (PSA)).
-
-
Adverse effects of treatment subgrouped or not for type, severity, or drug class.
Secondary outcomes
Microbiological recurrence, defined as reappearance of a pathogen or increase of its load over a defined threshold (for example, > 103 CFU/mL) after (apparent) eradication, assessed at the TOC visit.
Search methods for identification of studies
Electronic searches
Clinical trials for CBP were identified through MEDLINE (1966 to 8 August 2012) by crossing the sensitivity‐maximizing version of the Cochrane highly sensitive search strategy for identifying randomized trials in MEDLINE (2008 revision) (Higgins 2011) with the Boolean logic structure (item #6 of the following list):
-
'prostatitis[MeSH]', (including all subheadings)
-
'(prostatitis) NOT (prostatitis[MeSH Terms])'
-
'(prostato‐vesic*[Title/Abstract]) OR (prostatovesic*[Title/Abstract]) OR (prostato ADJ vesic*[Title/Abstract])'
-
'(vesiculo‐prostat*[Title/Abstract]) OR (vesiculoprostat*[Title/Abstract]) OR (vesiculo ADJ prostat*[Title/Abstract])'
-
'(prostate OR prostate[MeSH Terms]) AND (bacterial infections and mycoses[MeSH Terms])'
-
#1 OR #2 OR #3 OR #4 OR #5
The specialized PROSTATE register of the Cochrane Prostatic Diseases and Urologic Cancers Group, Cochrane Central Register of Controlled Trials (CENTRAL) and EMBASE databases were searched in an analogous fashion.
Searching other resources
The meta‐register of Current Controlled Trials (controlled‐trials.com) and the US registry of clinical trials (clinicaltrials.gov) were searched for protocols and results of RCTs on CBP.
International and national databases (for example, LILACS, Panteleimon, IMSEAR, WPRIM, IndMed, KoreaMed, PASCAL, Australasian Medical Index, Eastern‐Mediterranean Index Medicus) were also searched.
Handsearching was performed on the web pages containing the abstracts of all scientific contributions presented at international meetings of the European Association of Urology (http://www.uroweb.org/), American Urological Association (http://www.auanet.org/), International Society of Chemotherapy (http://www.ischemo.org/), and the International Continence Society (http://www.icsoffice.org/Events/EventsIndex.aspx). The general term 'prostatitis' was used for the abstract search.
One systematic review of the literature was retrieved (Erickson 2008). This review was also searched for studies.
Data collection and analysis
Selection of studies
-
The titles and abstracts obtained with the search strategy described above were screened independently by two review authors (GP, EM). Studies deemed to be not eligible for the systematic review were excluded. Reviews or manuscripts that might include relevant data or information on studies were retained initially.
-
Two review authors (GP, FMEW) independently assessed the retrieved abstracts and, if necessary, the full text of these studies to determine which studies satisfied the inclusion criteria.
-
Discrepancies in the eligibility of retrieved studies were resolved by discussion. If necessary, the Cochrane Prostatic Diseases and Urologic Cancers Group was involved for arbitration.
Data extraction and management
Data extraction was performed independently by two review authors (GP, FMEW), using a modified version of a standard data extraction form provided by the Cochrane Renal Group.
Studies were eligible if they were randomized, involved a placebo control group or an active drug comparison group, involved patients with CBP diagnosed according to NIH or Drach 1978 criteria, and if diagnosis at enrolment was performed using an adequate lower urinary tract segmented bacteriological test (4‐glass or 2‐glass).
Discrepancies or disagreements were resolved by discussion and, if necessary, by arbitration involving the Cochrane Prostatic Diseases and Urologic Cancers Group. Studies reported in non‐English language journals were tentatively translated before assessment, asking for the collaboration of the original authors of the reports. Any further information required from the original authors was obtained by correspondence and, if relevant, was included in the review. Where more than one publication of one trial was found, reports were grouped together and the most complete data set was used.
Assessment of risk of bias in included studies
The risk of bias of included studies was assessed by two independent review authors (FMEW, GP), without blinding to authorship or journal. The following items were assessed using the Cochrane Collaboration tool for assessing risk of bias (Higgins 2011).
-
Random sequence generation.
-
Allocation concealment.
-
Blinding of participants and personnel.
-
Blinding of outcome assessment.
-
Incomplete outcome data.
-
Selective reporting.
-
Other sources of bias.
'Risk of bias' tables were generated for each included study and were summarized in a 'Risk of bias' summary figure.
In the present review, risk of bias was a fundamental component of the analysis of the quality of evidence according to the GRADE approach (Higgins 2011). Quality of the evidence was graded as high, moderate, low or very low. In the case of low risk of bias, no downgrading of a study was deemed necessary. In the case of unclear risk of bias, or in the presence of biases raising doubts about the estimate of the effect and the results, a study was downgraded one level (for example, 'moderate' to 'low'). In the case of high risk of bias, the quality of the evidence was downgraded one or two levels (for example, 'moderate' to 'very low') depending on the severity of biases seriously weakening confidence in the results.
Measures of treatment effect
For dichotomous outcomes (for example, microbiological efficacy (number of patients showing eradication versus persistence), clinical efficacy (number of patients undergoing cure or improvement versus failure), adverse effects (number of patients showing the adverse effect)) results were expressed as risk ratios (RRs).
In the presence of ordinal outcomes (for example, mild, moderate or severe symptoms), these were dichotomized (for example, mild versus moderate or severe symptoms).
Where non‐dichotomous scales were used to assess the effects of treatment (for example, symptom, bother or QoL scores), the mean difference (MD) was calculated. If different scales were adopted for the same outcome, the standardized mean difference (SMD) was used for analysis.
Both dichotomous and categorical outcomes were expressed with 95% confidence intervals (CIs).
Unit of analysis issues
Cluster‐randomized trials were excluded from the meta‐analysis as they are in general more prone to bias and, in the context of meta‐analysis, they may cause overestimation of the effect of interventions due to the tendency to show narrow CIs and smaller P values (Chapter 16.3.1, Higgins 2011).
Cross‐over trials were planned to be incorporated in meta‐analyses by including only data from the very first period of randomized treatment (for continuous outcomes). In addition, cross‐over trials were planned to be assessed for risk of bias by analyzing the following items in the report and the protocol of the study.
-
Was use of a cross‐over design appropriate?
-
Is it clear that the order of receiving treatments was randomized?
-
Can it be assumed that the trial was not biased from carry‐over effects?
-
Are unbiased data available?
-
Are results of the second treatment period concealed?
If multiple treatments were compared within a single study, indirect comparisons were planned to be performed to provide an indirect estimate of the relative effect of the single interventions. The limits of this approach were taken into account during evaluation of the quality of evidence, according to GRADE criteria (Guyatt 2008).
Dealing with missing data
Missing studies
The comprehensive search strategy described above has been designed to minimize missing studies.
Missing outcomes
Studies not reporting information on a primary outcome were not excluded from the present systematic review. The lack of relevant outcomes from a study of interest was addressed in the discussion section and during 'Risk of bias' assessment.
Missing data or missing individuals
We attempted to request relevant missing data from the original authors or trialists. If data were apparently missing at random, we analyzed only the available information.
Because imputation strategies may significantly increase heterogeneity, we limited our analysis to participants for whom outcomes were obtained (available case analysis).
A high risk of selective reporting bias was assigned to trials when study outcomes were described in the methods paragraph but were not reported in the results section of the same article.
Assessment of heterogeneity
Evidence of heterogeneity was initially assessed by visual inspection of the forest plots. Heterogeneity was analyzed by calculating the I2 statistic. A 50% threshold was set for further investigation of heterogeneity by subgroup analysis.
Combined endpoints (for example, 'any adverse effects', including different lists of adverse effects for each trial) were assessed by the random‐effects model.
Assessment of reporting biases
To identify reporting biases, we performed a comprehensive search of clinical trials registers (http://clinicaltrials.gov/; www.controlled‐trials.com) in order to compare the original protocols with published reports of the same trials. Reporting bias was assessed by generating funnel plots in RevMan 5.1 and by testing for funnel plot asymmetry (for example, Egger test). These tests were planned to be performed if at least 10 studies were included in the meta‐analysis.
Data synthesis
We compared dichotomous as well as non‐dichotomous outcomes at the endpoint. We assessed effect size inconsistency as well as clinical study design and statistical heterogeneity.
We used a fixed‐effect model to compare microbiological efficacy as standardized pathogen cultures performed on patients' biological samples evaluate exactly the same effect, and variations in this case are likely to be due to sampling issues. Conversely, a random‐effects model was adopted to evaluate clinical efficacy, to take into account the diverse strategies used in the included studies in order to assess general clinical endpoints (for example, cure or improvement definitions). Adverse effects were also analyzed using a random‐effects model. Finally, data analyzed with a fixed‐effect model were analyzed using a random‐effects model to investigate heterogeneity among studies.
We reported non‐dichotomous clinical outcomes (for example, questionnaire scores) by comparing the SMDs. We were aware that a limitation to this kind of analysis is the fact that the scores of clinical questionnaires are often based on ordinal scales.
For both continuous and dichotomous outcomes we calculated 95% CIs.
Relevant data from pooled analyses were reported in 'Summary of findings' tables.
Subgroup analysis and investigation of heterogeneity
Where substantial heterogeneity was found among pooled studies evaluated with a fixed‐effect model, we repeated the analysis using a random‐effects model.
To explore possible sources of heterogeneity, we planned to perform subgroup analysis if an adequate number of pooled studies were available. Heterogeneity among study participants might be related to the following criteria:
-
age of participants (< 55 years versus ≥ 55 years) (Berges 2011);
-
prostate volume (< 25 mL versus ≥ 25 mL) (Berges 2011);
-
severity of symptoms at baseline or on study enrolment (assessed with National Institutes of Health Chronic Prostatitis Symptom Index (NIH‐CPSI) or other symptom scores, bother scores, QoL scores, or patient self‐declared status) (e.g., NIH‐CPSI total score < 15 versus ≥ 15) (Nickel 2001);
-
type, sensitivity or specificity of microbiological diagnostic tests (4‐glass versus 2‐glass);
-
previous antibacterial treatment (naïve versus heavily or chronically pretreated participants);
-
duration of antibacterial treatment (< 4 weeks versus ≥ 4 weeks);
-
duration of follow up (< 1 month versus ≥ 1 month);
-
different criteria for microbiological outcome (e.g., different bacterial load cutoff to define pathogen eradication, for example, 105 versus 103 CFU/mL);
-
different tests used to measure clinical outcomes (e.g., NIH‐CPSI versus International Prostate Symptom Score (IPSS)).
Subgroup analysis was performed only in the presence of an adequate number of studies and if subgroup data were available.
Sensitivity analysis
Sensitivity analysis was used to explore the robustness of the meta‐analysis in the presence of I2 values beyond the 50% threshold.
We performed sensitivity analysis in the presence of a sufficient number of included studies by repeating the analysis taking into account one or more of the following items:
-
specific parameters of study quality (e.g., low versus moderate or high risk of bias);
-
different measures of effect size (secondary analysis performed with odds ratios (ORs) in the case of the primary analysis performed using RRs);
-
different statistical models (secondary analysis performed with a random‐effects model in the case of the primary analysis performed using a fixed‐effect model).
Results
Description of studies
See: Characteristics of included studies; Characteristics of excluded studies.
Results of the search
We identified 3394 potential studies from the article databases and 12 potential studies from the online congress abstract databases. The databases searched and the number of retrieved articles for each database are listed in Figure 1. From 104 potentially relevant studies selected after title and abstract review, 12 articles were not evaluable and 41 articles were excluded. Among the excluded papers, four did not involve antibacterial treatment, 26 were non‐RCT studies, and 11 did not include patients with CBP or included patients showing CBP together with other concomitant conditions. Among the remaining 51 articles, 18 were finally included in this systematic review, 32 were excluded, and one awaits classification (Drasa 2009).
Flow chart of included and excluded studies.
Included studies
Eighteen studies, including a total of 2196 randomized participants, met all inclusion criteria. Among these studies, 14 compared two different antibacterial agents (AAs) in treatment arms containing participants with CBP caused by different pathogens (Bundrick 2003; Bustillo 1997; Cox 1989; Giannarini 2007; Koff 1996; Naber 2002; Paulson 1986; Zhang 2012), or CBP caused by a single pathogen (Cai 2010; Ohkawa 1993; Skerk 2002; Skerk 2003; Skerk 2004a; Skerk 2006). One RCT compared an AA with an AA combined with a phosphodiesterase‐5 inhibitor (PDE5‐I) (Aliaev 2008). Two RCTs compared two courses of different lengths with the same AA (Skerk 2004b; Smith 1979). One article compared the combination of an AA plus herbal supplement with an AA administered as single‐agent (Cai 2009).
Excluded studies
Thirty‐two studies did not meet inclusion criteria. Two articles (Nickel 2008a; Schaeffer 2005) presented subset analyses of an included study (Bundrick 2003). Four articles were not focusing on CBP (Gleckman 1979; Martino 1993; Sabbaj 1986; Shen 2004), and one included participants with chronic prostatitis involving protozoans as the etiological agents (Vickovic 2010). One study included CBP patients (n = 2) within a treatment arm containing men and women with various urinary tract infections (Childs 1983). Two studies were non‐comparative (Baert 1983; Wedren 1989), four were non‐randomized (Brannan 1975; Colleen 1975; Kozdoba 2007; Smelov 2004), and five were non‐RCTs (Cox 1991; Kunishima 2008; Lee 2006; Panagopoulos 2009; Shafik 1992). In particular, in the study by Lee et al, participants affected by category II and IIIa prostatitis were pooled together (Lee 2006). In three studies, participants were affected by CBP associated with other conditions, namely, vesiculitis (Kim 2006), urethritis (Zhang 2004), and genital infection with oligoasthenoteratozoospermia (Cai 2011). In three articles, a lower urinary tract segmented test was not mentioned or described in the methods section (Deng 2004; Hu 2002; Vicari 2000). In one study, a microbiological diagnostic test was not required at enrolment, and a past history of CBP was deemed sufficient to qualify a patient as having CBP (Paglia 2010). In four studies, participants belonging to a single treatment arm were treated with various antibiotics, and the names of the drugs or the number of participants treated with a given drug were not specified, or subgroup analysis was not performed (Ateya 2006; Barbalias 1998; Liao 2004; Trapeznikova 2007). For one (Chinese) study, translation was not available at the review authors’ institutions, and was not provided by the authors when requested (Xu 2010). Moreover, this study, together with the Zhang 2004 and Liao 2004 trials, involved traditional Chinese medications. One RCT compared two different techniques of intraprostatic administration of AA. Thus, this was a trial evaluating neither alternative antibiotics, different doses or dosages, nor different routes of administration (Yavaşçaoğlu 1998).
Risk of bias in included studies
The risk of bias analysis is summarized in Figure 2 and Figure 3.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Allocation
Random sequence generation
Nine trials described adequately the procedure used for generation of the randomization sequences (Bundrick 2003; Giannarini 2007; Naber 2002; Skerk 2002; Skerk 2003; Skerk 2004a; Skerk 2004b; Skerk 2006; Zhang 2012). In seven reports (Aliaev 2008; Cai 2009; Cai 2010; Cox 1989; Ohkawa 1993; Paulson 1986; Smith 1979), randomization procedures were not described in detail, though it was clearly stated that participants were randomized. In two studies, the sequence generation procedure was not adequate (Koff 1996) or not disclosed (Bustillo 1997).
Allocation concealment
Allocation concealment procedures were not disclosed in all 18 included studies. Though concealment was probably adequate in one study (Giannarini 2007), in the remaining 17 studies it was unclear as to whether allocation was concealed or not (Aliaev 2008; Bundrick 2003; Bustillo 1997; Cai 2009; Cai 2010; Cox 1989; Koff 1996; Naber 2002; Ohkawa 1993; Paulson 1986; Skerk 2002; Skerk 2003; Skerk 2004a; Skerk 2004b; Skerk 2006; Smith 1979; Zhang 2012).
Blinding
Three studies were double‐blinded (Bundrick 2003; Giannarini 2007; Smith 1979). One study was single‐blinded (Paulson 1986). The remaining 14 studies were open‐label (Aliaev 2008; Bustillo 1997; Cai 2009; Cai 2010; Cox 1989; Koff 1996; Naber 2002; Ohkawa 1993; Skerk 2002; Skerk 2003; Skerk 2004a; Skerk 2004b; Skerk 2006; Zhang 2012).
Lack of blinding was not deemed to be a major determinant of performance bias in one study having as the sole primary outcome a non‐subjective endpoint, namely microbiological eradication (Koff 1996). Similarly, in the Smith paper (Smith 1979) the sole outcome of the trial was not subjective (microbiological eradication following antibiotic treatment), and the risk of both performance and detection biases was deemed to be low. On the contrary, primary outcomes based on clinical signs and symptoms or QoL scores (Aliaev 2008; Bustillo 1997; Cai 2009; Cai 2010; Cox 1989; Naber 2002; Ohkawa 1993; Skerk 2002; Skerk 2003; Skerk 2004a; Skerk 2004b; Skerk 2006; Zhang 2012) were considered at risk of bias in the absence of blinding.
The Paulson study (Paulson 1986) was deemed to be at high risk of bias. Although the study was single‐blinded, participants in group 1 (oral minocycline twice daily) did not receive two additional placebo tablets to equal participants in group 2 (cephalexin four times/day).
The risk of detection bias was considered unclear if the blinding of outcome assessors was not described or disclosed in the study reports (Bundrick 2003; Bustillo 1997; Cai 2009; Cai 2010; Cox 1989; Koff 1996; Naber 2002; Ohkawa 1993; Paulson 1986; Skerk 2002; Skerk 2003; Skerk 2004a; Skerk 2004b; Skerk 2006; Zhang 2012). One open study was at high risk of bias due to the specific nature of the experimental drug combination (Aliaev 2008).
Incomplete outcome data
Two studies included an intention‐to‐treat (ITT) analysis (Bundrick 2003; Naber 2002). Three studies (Bustillo 1997; Cai 2010; Giannarini 2007) were considered as having low risk of attrition bias due to the low impact of missing data on microbiological outcome estimates and the high expected frequency of the outcome (pathogen eradication after fluoroquinolone therapy). In seven studies, withdrawals and dropouts were not described or were indefinite; the risk of attrition bias was unclear in these studies (Aliaev 2008; Koff 1996; Skerk 2002; Skerk 2003; Skerk 2004a; Skerk 2004b; Skerk 2006). In one study, the risk of bias was high due to the high rate of withdrawals (22.2% per treatment arm) (Ohkawa 1993). Two studies were considered at high risk of attrition bias as the reasons for study withdrawals were not presented separately according to treatment group (Cai 2009) or to disease group (Paulson 1986). One study showed high dropout frequencies in both treatment arms (59% and 41.6%) but lacked an ITT analysis (Cox 1989). This study was considered as having a high risk of bias. Similarly, in the Smith study (Smith 1979) a high number of withdrawals in the treatment groups and the low expected therapeutic success of the experimental drug (co‐trimoxazole) suggested high attrition bias. In the Zhang 2012 study, almost 40% of the isolated pathogen strains were resistant to ciprofloxacin. Nevertheless, patients harbouring resistant strains were apparently treated with ciprofloxacin. Subgroup analysis on eradication rates only in patients harboring sensitive strains was not disclosed.
Selective reporting
Three trials were considered to be free of selective reporting (Bustillo 1997; Cai 2010; Naber 2002). In the Smith paper, a section addressing clinical results was not presented although clinical assessments were described in the methods section (Smith 1979). Fourteen trials were considered at high risk of reporting bias (Aliaev 2008; Bundrick 2003; Cai 2009; Cox 1989; Giannarini 2007; Koff 1996; Ohkawa 1993; Paulson 1986; Skerk 2002; Skerk 2003; Skerk 2004a; Skerk 2004b; Skerk 2006; Zhang 2012).
Other potential sources of bias
In one study, the per protocol‐like design did not allow evaluating the presence or absence of baseline imbalances (Aliaev 2008). Risk of bias was unclear. In one high‐risk trial, patients with different urological conditions were pooled at enrolment (Paulson 1986). This made evaluating baseline values impossible. In one study, the trial design and methods were not described in sufficient detail (Koff 1996), and risk of bias was rated 'unclear'. One high risk of bias study was poorly designed as participants with acute prostatitis were included in a cohort of CBP participants (Cox 1989). In one high‐risk study, 'additional agents' were administered to a fraction of the participants in both treatment arms; names and dosages of these agents were not disclosed (Zhang 2012). Moreover, the design of the 4‐glass lower urinary tract diagnostic segmented test was modified, and assessment of concomitant bacterial urethritis was impossible (Zhang 2012). In the same study, resistance to study drugs was not an exclusion criterion. In the remaining studies, risk of bias was considered low (Bundrick 2003; Bustillo 1997; Cai 2009; Cai 2010; Giannarini 2007; Naber 2002; Ohkawa 1993; Skerk 2002; Skerk 2003; Skerk 2004a; Skerk 2004b; Skerk 2006; Smith 1979).
Effects of interventions
See: Summary of findings for the main comparison Levofloxacin versus ciprofloxacin for chronic bacterial prostatitis; Summary of findings 2 Lomefloxacin versus comparator fluoroquinolone for chronic bacterial prostatitis; Summary of findings 3 Ciprofloxacin versus comparator fluoroquinolone for chronic bacterial prostatitis; Summary of findings 4 Levofloxacin versus comparator fluoroquinolone for chronic bacterial prostatitis
Eighteen RCTs were included in this review.
Different antibacterial agents
Fourteen parallel‐group studies compared different antibacterial agents.
Comparisons between different fluoroquinolones
Levofloxacin versus ciprofloxacin
Two studies, involving a total of 791 participants (Bundrick 2003, 383 participants; Zhang 2012, 408 participants), compared levofloxacin (500 mg once daily for four weeks in both trials) to ciprofloxacin (500 mg twice daily for four weeks in both trials) in patients affected by CBP (summary of findings Table for the main comparison). The Bundrick study was double‐blind, whereas the Zhang study was open‐label. Both studies included an ITT analysis. The studies had similar microbiological and clinical outcomes (microbiological eradication at the end of therapy; clinical success (cured or improved) at the end of therapy and after a six‐month follow‐up; adverse effects of treatment).
-
Microbiological efficacy (pathogen eradication) did not differ significantly between groups in the Bundrick study (RR 0.98, 95% CI 0.85 to 1.12), whereas in the Zhang study levofloxacin was found to significantly increase the RR for microbiological eradication (RR 1.42, 95% CI 1.25 to 1.61) (Analysis 1.1). When data were pooled (levofloxacin versus ciprofloxacin comparison), a significant increase in RR for eradication was observed (RR 1.22, 95% CI 1.11 to 1.34, fixed‐effect model). Substantial heterogeneity was found between the studies (Chi2 = 15.82 (P < 0.0001); I2 = 94%). When a random‐effects model was adopted to analyze the pooled eradication data, the difference lost statistical significance (RR 1.18, 95% CI 0.81 to 1.71) (Analysis 1.2). Sensitivity analysis was performed by calculating ORs in place of RRs for the microbiological efficacy primary outcome. Results from the Bundrick study were not substantially affected (OR 0.91, 95% CI 0.51 to 1.60 (fixed‐effect and random‐effects models; forest plot not shown)). Conversely, results from the Zhang study were substantially influenced by this strategy (OR 3.99, 95% CI 2.43 to 6.35 (fixed‐effect model); OR 3.93, 95% CI 2.43 to 6.35 (random‐effects model; forest plot not shown)). Consequently, pooled results were also affected (OR 2.16, 95% CI 1.52 to 3.07 (fixed‐effect model); OR 1.90, 95% CI 0.45 to 8.01 (random‐effects model; forest plot not shown)).
-
In the Bundrick study, clinical efficacy (cure or improvement) did not differ significantly between groups when assessed at the end of therapy (RR 1.03, 95% CI 0.89 to 1.19) or after a six‐month follow‐up (RR 0.99, 95% CI 0.85 to 1.16) (Analysis 1.3). In the Zhang study, levofloxacin was found to significantly increase the RR for clinical efficacy (cure or improvement) both at the end of therapy (RR 1.30, 95% CI 1.18 to 1.43) and at follow‐up (RR 1.33, 95% CI 1.21 to 1.46). When data were pooled, clinical efficacy did not differ significantly between treatment arms, both at the end of therapy (RR 1.16, 95% CI 0.93 to 1.46) and at follow‐up (RR 1.16, 95% CI 0.86 to 1.55) (Analysis 1.3). Also in this case significant heterogeneity was observed between the studies (end of therapy: Chi2 = 7.06 (P = 0.008); I2 = 86%; follow‐up: Chi2 = 10.35 (P = 0.001); I2 = 90%). Sensitivity analysis was performed by calculating ORs in place of RRs for clinical efficacy (forest plots not shown). Results from the Bundrick study were not substantially affected (OR 1.12, 95% CI 0.64 to 1.95 (end of therapy); OR 0.97, 95% CI 0.57 to 1.66 (follow‐up)). Results from the Zhang study were substantially influenced by this strategy (OR 5.45, 95% CI 2.92 to 10.18 (end of therapy); OR 6.75, 95% CI 3.50 to 13.04 (follow‐up)). Consequently, pooled results were also affected (OR 2.46, 95% CI 0.52 to 11.63 (end of therapy); OR 2.54, 95% CI 0.38 to 17.14 (follow‐up)).
-
The rate of adverse effects did not differ significantly between treatment groups in both studies (Analysis 1.4). With the exception of the 'dizziness' effect (Chi2 = 3.68 (P = 0.06); I2 = 73%), heterogeneity was not detected for the adverse effects of treatment outcome.
Prulifloxacin versus levofloxacin
In one study involving 96 participants (Giannarini 2007), prulifloxacin (600 mg once daily for four weeks) was compared to levofloxacin (500 mg once daily for four weeks).
-
Microbiological efficacy (pathogen eradication) did not differ significantly between groups (RR 1.02, 95% CI 0.79 to 1.33) (Analysis 2.1).
-
Clinical efficacy (total NIH‐CPSI scores) did not differ significantly between groups when assessed at the end of therapy (SMD ‐0.03, 95% CI ‐0.45 to 0.39) (Analysis 2.2).
-
The rate of adverse effects did not differ significantly between treatment groups (RR 0.82, 95% CI 0.36 to 1.88 (any adverse effects)) (Analysis 2.3).
Lomefloxacin versus ofloxacin
In one study involving 33 participants (Koff 1996), lomefloxacin (400 mg once daily for six weeks) was compared to ofloxacin (200 mg twice daily for six weeks).
-
Microbiological efficacy (pathogen eradication) after a six‐month follow‐up did not differ significantly between groups (RR 1.11, 95% CI 0.66 to 1.88) (Analysis 3.1).
-
The rate of adverse effects did not differ significantly between the treatment groups (RR 0.42, 95% CI 0.16 to 1.12 (any adverse effects)) (Analysis 3.2).
Lomefloxacin versus ciprofloxacin
In one study involving 182 participants (Naber 2002) lomefloxacin (400 mg once daily for four weeks) was compared to ciprofloxacin (500 mg twice daily for four weeks). In this study, equivalence between lomefloxacin and ciprofloxacin was defined as a 95% CI within 15% of the observed differences.
Intention‐to‐treat (ITT) analysis
-
Microbiological efficacy (pathogen eradication) did not differ significantly between groups at the end of therapy (RR 0.96, 95% CI 0.82 to 1.11) or for follow‐up at four weeks (RR 0.87, 95% CI 0.72 to 1.06), three months (RR 0.90, 95% CI 0.74 to 1.09) or six months (RR 0.87, 95% CI 0.67 to 1.12) (Analysis 4.1).
-
Clinical efficacy (cure or improvement) did not differ significantly between groups when assessed at the end of therapy (RR 1.01, 95% CI 0.94 to 1.09) or for follow‐up at four weeks (RR 0.91, 95% CI 0.78 to 1.05), three months (RR 0.97, 95% CI 0.82 to 1.15) or six months (RR 0.91, 95% CI 0.75 to 1.11) (Analysis 4.3).
-
The rate of adverse effects did not differ significantly between treatment groups (RR 0.82, 95% CI 0.40 to 1.68 (any adverse effects)) (Analysis 4.5).
Per protocol analysis
-
Microbiological efficacy (pathogen eradication) did not differ significantly between groups at the end of therapy (RR 0.98, 95% CI 0.89 to 1.09) or for follow‐up at four weeks (RR 1.00, 95% CI 0.94 to 1.07), three months (RR 1.03, 95% CI 0.94 to 1.12) or six months (RR 0.94, 95% CI 0.80 to 1.09) (Analysis 4.2).
-
Clinical efficacy (cure or improvement) did not differ significantly between groups when assessed at the end of therapy (RR 0.96, 95% CI 0.89 to 1.03) or for follow‐up at four weeks (RR 1.00, 95% CI 0.86 to 1.18), three months (RR 1.07, 95% CI 0.94 to 1.21) or six months (RR 0.88, 95% CI 0.77 to 1.01) (Analysis 4.4).
Lomefloxacin versus comparator fluoroquinolone
Two trials (Koff 1996; Naber 2002) compared a cycle of treatment with lomefloxacin (400 mg once daily) with a comparator second‐generation fluoroquinolone (Koff 1996: ofloxacin; Naber 2002: ciprofloxacin) (summary of findings Table 2).
-
The trials were pooled for microbiological efficacy (pathogen eradication) at follow‐up (six months). The RR analysis showed no significant difference between the treatment arms (RR 0.96, 95% CI 0.80 to 1.16) (Analysis 5.1).
-
The trials were also pooled for adverse effects. Men in the lomefloxacin arm were not at a significantly different risk than men in the comparator fluoroquinolone arm for total adverse effects (RR 0.64, 95% CI 0.34 to 1.21), gastrointestinal effects (RR 0.58, 95% CI 0.27 to 1.23), headache (RR 0.56, 95% CI 0.07 to 4.43) or dizziness (RR 0.88, 95% CI 0.09 to 8.60) (Analysis 5.2). Heterogeneity was not detected for the adverse effects outcome.
Ciprofloxacin versus comparator fluoroquinolone
Three trials (Bundrick 2003; Naber 2002; Zhang 2012) compared a cycle of treatment with ciprofloxacin (500 mg twice daily for four weeks) with a comparator second‐generation fluoroquinolone (Bundrick 2003 and Zhang 2012: levofloxacin 500 mg once daily for four weeks; Naber 2002: lomefloxacin 400 mg once daily for four weeks) (summary of findings Table 3).
-
When microbiological efficacy outcome data were pooled, the RR for pathogen eradication was 0.87 (95% CI 0.80 to 0.94) (Analysis 6.1, fixed‐effect model). Substantial heterogeneity was found between the studies (Chi² = 22.32 (P value < 0.0001); I² = 91%). When a random‐effects model was adopted to further analyze the pooled eradication data, the difference lost statistical significance (RR 0.91, 95% CI 0.70 to 1.18) (Analysis 6.2). When the Zhang study (identified as the likely source of heterogeneity by visual inspection of the forest plot) was excluded from the pooled analysis, the I2 value changed from 91% to 0% and the RR for a random‐effects model was 1.03 (95% CI 0.93 to 1.14) (Analysis 6.2). Sensitivity analysis was performed by calculating ORs in place of RRs for the microbiological efficacy outcome. The OR for pathogen eradication for a fixed‐effect model was 0.56 (95% CI 0.41 to 0.77) (I2 = 90%; forest plot not shown). The OR for pathogen eradication for a random‐effects model was 0.69 (95% CI 0.24 to 2.02) (I2 = 90%; forest plot not shown). When the Zhang study was excluded from the pooled studies, the I2 value changed from 90% to 0% and the OR for a random‐effects model was 1.15 (95% CI 0.74 to 1.80; forest plot not shown).
-
When clinical efficacy (cure or improvement) data were pooled, results did not differ significantly between treatment arms, both at the end of therapy (RR 0.90, 95% CI 0.75 to 1.08) and at follow‐up (RR 0.93, 95% CI 0.72 to 1.20). Also in this case significant heterogeneity was observed between the studies (end of therapy: Chi² = 19.30 (P value < 0.0001); I² = 90%; follow‐up: Chi² = 18.31 (P value = 0.0001); I² = 89%) (Analysis 6.3). Exclusion of the Zhang study reduced the I2 value to 0%. Sensitivity analysis was performed by calculating ORs in place of RRs for the clinical efficacy primary outcome. The OR for clinical efficacy at the end of therapy for a random‐effects model was 0.49 (95% CI 0.15 to 1.56) (I2 = 86%; forest plot not shown). The OR for clinical efficacy at follow‐up with a random‐effects model was 0.59 (95% CI 0.16 to 2.18) (I2 = 93%; forest plot not shown). Exclusion of the Zhang study reduced the I2 value to 0%.
-
The rate of adverse effects did not differ significantly between treatment groups in both studies (Analysis 6.4) and, when feasible, in pooled analyses. Heterogeneity was not detected for the adverse effects outcome.
Levofloxacin versus comparator fluoroquinolone
Three trials (Bundrick 2003; Giannarini 2007; Zhang 2012) compared a cycle of treatment with levofloxacin (500 mg once daily for four weeks) with a comparator fluoroquinolone (Bundrick 2003 and Zhang 2012: ciprofloxacin 500 mg twice daily for four weeks; Giannarini 2007: prulifloxacin 600 mg once daily for four weeks) (summary of findings Table 4).
-
When microbiological efficacy outcome data were pooled, a significant increase in RR for pathogen eradication was observed (RR 1.19, 95% CI 1.09 to 1.30, fixed‐effect model) (Analysis 7.1). Substantial heterogeneity was found between the studies (Chi² = 17.85 (P value = 0.0001); I² = 89%). When a random‐effects model was adopted to further analyze the pooled pathogen eradication data the difference lost statistical significance (RR 1.12, 95% CI 0.84 to 1.48) (Analysis 7.2). When the Zhang study (identified as the likely source of heterogeneity by visual inspection of the forest plot) was excluded from the pooled analysis, the I2 value changed from 89% to 0% and the RR for the random‐effects model was 0.98 (95% CI 0.87 to 1.10) (Analysis 7.2). Sensitivity analysis was performed by calculating ORs in place of RRs for the microbiological efficacy outcome. The OR for pathogen eradication with a fixed‐effect model was 1.93 (95% CI 1.39 to 2.68) (I2 = 89%; forest plot not shown). The OR for pathogen eradication with a random‐effects model was 1.54 (95% CI 0.52 to 4.52) (I2 = 89%; forest plot not shown). When the Zhang study was excluded from the pooled studies, the I2 value changed from 90% to 0% and the OR for the random‐effects model was 0.91 (95% CI 0.56 to 1.48; forest plot not shown).
-
The rate of adverse effects did not differ significantly between treatment groups (Analysis 7.3). Heterogeneity was not detected for the adverse effects outcome.
Fluoroquinolones versus other antibacterial agents
Prulifloxacin versus doxycycline
In one study involving 221 participants (Cai 2010), prulifloxacin (600 mg once daily for two weeks) was compared to doxycycline (100 mg twice daily for three weeks) in patients affected by chlamydial prostatitis.
-
Microbiological efficacy, evaluated as the absence of both chlamydial deoxyribonucleic acid (DNA) and anti‐Chlamydia immunoglobulin A (IgA) at the end of therapy, did not differ significantly between groups (RR 1.12, 95% CI 0.93 to 1.36) (Analysis 8.1).
-
For clinical efficacy, a significant difference in the total NIH‐CPSI scores was observed for the prulifloxacin and doxycycline comparison (SMD ‐0.66, 95% CI ‐0.94 to ‐0.39) (Analysis 8.2).
-
Clinical efficacy, defined as the fraction of asymptomatic patients at the end of therapy, did not differ significantly between treatment arms (RR 1.04, 95% CI 0.91 to 1.19) (Analysis 8.3).
-
The rate of adverse effects did not differ significantly between treatment groups (RR 1.17, 95% CI 0.32 to 4.24 (any adverse effects)) (Analysis 8.4).
Ofloxacin versus minocycline
In one study involving 18 participants (Ohkawa 1993), ofloxacin (200 mg thrice daily for two weeks) was compared to minocycline (100 mg twice daily for two weeks) in patients affected by ureaplasmal prostatitis.
-
Microbiological efficacy (pathogen eradication) did not differ significantly between groups (RR 1.00, 95% CI 0.78 to 1.29) (Analysis 9.1).
-
Clinical efficacy (cure or improvement), assessed at the end of therapy, did not differ significantly between groups (RR 0.87, 95% CI 0.59 to 1.26) (Analysis 9.2).
-
The trial authors reported that neither group was affected by adverse effects of therapy.
Ofloxacin versus carbenicillin
In one study involving 46 participants (Cox 1989), ofloxacin (300 mg twice daily for six weeks) was compared to carbenicillin (764 mg four times/day for six weeks).
-
Microbiological efficacy (pathogen eradication) did not differ significantly between groups (RR 1.04, 95% CI 0.76 to 1.42) (Analysis 10.1).
-
Clinical efficacy (cure or improvement), assessed at the end of treatment, did not differ significantly between groups (RR 1.06, 95% CI 0.85 to 1.32) (Analysis 10.2).
-
The rate of adverse effects did not differ significantly between the treatment groups (RR 0.73, 95% CI 0.31 to 1.71 (any adverse effects)) (Analysis 10.3).
Lomefloxacin versus trimethoprim‐sulfamethoxazole (co‐trimoxazole)
In one study involving 30 participants (Bustillo 1997), lomefloxacin (400 mg once daily for six weeks) was compared to co‐trimoxazole (160 + 800 mg twice daily for six weeks).
-
Microbiological efficacy (pathogen eradication) did not differ significantly between the groups at the end of therapy (RR 1.09, 95% CI 0.82 to 1.44) or at the end of a four‐month follow‐up (RR 1.09, 95% CI 0.82 to 1.44) (Analysis 11.1).
-
Clinical efficacy (cure or improvement) did not differ significantly between groups when assessed at the end of therapy (RR 1.00, 95% CI 0.87 to 1.15) or at the end of a four‐month follow‐up (RR 1.00, 95% CI 0.87 to 1.15) (Analysis 11.2).
-
The rate of adverse effects did not differ significantly between treatment groups (RR 0.43, 95% CI 0.04 to 4.25 (any adverse effects)) (Analysis 11.3).
Ciprofloxacin versus azithromycin
In one study involving 89 participants affected by chlamydial prostatitis (Skerk 2003), ciprofloxacin (500 mg twice daily for 20 days) was compared to azithromycin (500 mg once daily, thrice‐weekly (first three consecutive days of each week) for three weeks).
-
There was a significant increase in pathogen eradication in the azithromycin arm (RR 0.48, 95% CI 0.32 to 0.72) (Analysis 12.1).
-
There was a significant increase in clinical success (cure or improvement) in the azithromycin arm (RR 0.64, 95% CI 0.46 to 0.90) (Analysis 12.2).
-
The rate of adverse effects did not differ significantly between the treatment groups (RR 0.34, 95% CI 0.01 to 8.15 (any adverse effects)) (Analysis 12.3).
Comparisons between different non‐fluoroquinolone antibiotics
Minocycline versus cephalexin
In one study involving 27 participants (Paulson 1986), minocycline (100 mg twice daily for four weeks) was compared to cephalexin (500 mg four times/day for four weeks).
-
Microbiological efficacy (pathogen eradication and eradication plus superinfection) did not differ significantly between groups at the end of therapy (RR 1.70, 95% CI 0.54 to 5.34) (Analysis 13.1).
-
Microbiological recurrence rates did not differ significantly between groups at the end of therapy (RR 0.98, 95% CI 0.37 to 2.59) (Analysis 13.3).
-
Clinical efficacy (cure or improvement), assessed at the end of therapy, did not differ significantly between groups (RR 2.04, 95% CI 0.83 to 4.99) (Analysis 13.2).
-
Adverse effects of therapy were not reported.
Azithromycin versus clarithromycin
In one study involving 91 participants affected by chlamydial prostatitis (Skerk 2002), azithromycin (500 mg once daily, thrice weekly (first three consecutive days of each week) for three weeks) was compared to clarithromycin (500 mg twice daily for two weeks).
-
Microbiological efficacy (pathogen eradication at the test‐of‐cure (TOC) visit) did not differ significantly between groups (RR 1.01, 95% CI 0.82 to 1.23) (Analysis 14.1).
-
Clinical efficacy (cure rate) did not differ significantly between groups when assessed at the end of therapy (RR 0.98, 95% CI 0.75 to 1.28) (Analysis 14.2).
-
The rate of adverse effects did not differ significantly between the treatment groups (RR 1.96, 95% CI 0.18 to 20.83 (any adverse effects)) (Analysis 14.3).
Azithromycin versus doxycycline in chlamydial prostatitis
In one study involving 125 participants affected by chlamydial prostatitis (Skerk 2004a), azithromycin (1000 mg once weekly for four weeks) was compared to doxycycline (100 mg twice daily for four weeks).
-
Microbiological efficacy (pathogen eradication at the end of therapy) did not differ significantly between groups (RR 1.03, 95% CI 0.85 to 1.26) (Analysis 15.1).
-
Clinical efficacy assessed as inflammatory findings at the end of therapy (number of participants with white blood cell counts in EPS/VB3 < 10 per high power field) did not differ significantly between groups (RR 1.08, 95% CI 0.66 to 1.78) (Analysis 15.2).
-
Clinical efficacy (cure or improvement) did not differ significantly between groups when assessed at the end of therapy (RR 0.95, 95% CI 0.76 to 1.19) (Analysis 15.2).
-
The rate of adverse effects did not differ significantly between the treatment groups (RR 0.21, 95% CI 0.04 to 1.04 (any adverse effects)) (Analysis 15.3).
Azithromycin versus doxycycline in ureaplasmal prostatitis
In one study involving 63 participants affected by ureaplasmal prostatitis (Skerk 2006), azithromycin (500 mg once daily, thrice weekly (first three consecutive days of each week) for three weeks) was compared to doxycycline (100 mg twice daily for three weeks).
-
Microbiological efficacy (pathogen eradication at the end of therapy) did not differ significantly between groups (RR 1.05, 95% CI 0.80 to 1.39) (Analysis 16.1).
-
Clinical efficacy (cure) did not differ significantly between groups when assessed at the end of therapy (RR 1.01, 95% CI 0.72 to 1.42) (Analysis 16.2).
-
The rate of adverse effects did not differ significantly between the treatment groups (RR 0.09, 95% CI 0.01 to 1.53 (any adverse effects)) (Analysis 16.3).
Different duration of therapy courses for the same antibacterial agent
Azithromycin 4.5 g versus 6.0 g (total dose) in chlamydial prostatitis
In one study focusing on chlamydial prostatitis (Skerk 2004b), 89 participants were randomly divided into a treatment arm receiving a total dose of 4.5 g azithromycin (500 mg once daily, thrice weekly (first three consecutive days of each week) for three weeks) and a treatment arm receiving total 6.0 g azithromycin (500 mg once daily, thrice weekly (first three consecutive days of each week) for four weeks).
-
Microbiological efficacy (pathogen eradication at the end of therapy) did not differ significantly between groups (RR 0.99, 95% CI 0.81 to 1.21) (Analysis 17.1).
-
Clinical efficacy (cure) did not differ significantly between groups when assessed at the end of therapy (RR 0.96, 95% CI 0.74 to 1.26) (Analysis 17.2).
-
The rate of adverse effects did not differ significantly between treatment groups (RR 0.19, 95% CI 0.01 to 3.79 (any adverse effects)) (Analysis 17.3).
Co‐trimoxazole 480 mg twice daily for 12 weeks versus 10 days
In one study involving 38 participants affected by chronic bacterial prostatitis (Smith 1979), oral co‐trimoxazole (480 mg twice daily), administered for a period of 12 weeks, was compared to 480 mg oral co‐trimoxazole (400 mg sulfamethoxazole; 80 mg trimethoprim), administered twice daily for 10 days.
-
There was a significant increase in pathogen eradication in the 12‐week treatment arm (RR 3.00, 95% CI 1.01 to 8.95) (Analysis 18.1).
-
The rate of adverse effects did not differ significantly between the treatment groups (RR 0.53, 95% CI 0.05 to 5.31 (any adverse effects)) (Analysis 18.2).
Antibacterial agents combined with other medications or supplements
Fluoroquinolone plus phosphodiesterase‐5 inhibitors versus fluoroquinolone
In one study involving 103 participants divided into three treatment arms (Aliaev 2008), a combination of levofloxacin (500 mg once daily for four weeks) with vardenafil, administered at fixed daily doses (10 mg once daily) or on‐demand (a single 10 mg tablet), was compared with levofloxacin as single‐agent (500 mg once daily for four weeks). The two regimens of combined therapy were also directly compared.
Levofloxacin plus vardenafil at fixed daily dose versus levofloxacin
-
Microbiological efficacy (pathogen eradication at the end of therapy) did not differ significantly between groups (RR 1.04, 95% CI 0.90 to 1.19) (Analysis 19.1).
-
Clinical efficacy, assessed as NIH‐CPSI pain, voiding and QoL impact scores, did not differ significantly between groups when assessed at the end of therapy (pain score: SMD ‐0.13, 95% CI ‐0.62 to 0.35; voiding score: SMD ‐0.30, 95% CI ‐0.78 to 0.19; QoL impact score: SMD ‐0.24, 95% CI ‐0.72 to 0.25) (Analysis 19.2).
-
Clinical efficacy, defined as improvement of inflammatory findings (number of participants with leukocytosis in post‐massage urine specimens at the end of treatment), did not differ significantly between groups (RR 0.54, 95% CI 0.17 to 1.66) (Analysis 19.3).
-
Clinical efficacy, expressed as urinary peak flow rates (Qmax, mL/s), did not differ significantly between treatment groups (SMD 0.24, 95% CI ‐0.25 to 0.72) (Analysis 19.4).
Levofloxacin plus vardenafil on‐demand versus levofloxacin
-
Microbiological efficacy (pathogen eradication at the end of therapy) did not differ significantly between groups (RR 1.01, 95% CI 0.88 to 1.17) (Analysis 20.1).
-
NIH‐CPSI pain and voiding scores did not differ significantly between groups when assessed at the end of therapy (pain score: SMD ‐0.06, 95% CI ‐0.53 to 0.42; voiding score: SMD 0.27, 95% CI ‐0.20 to 0.75) (Analysis 20.2).
-
The scores of the NIH‐CPSI domain focusing on the impact of the disease on the QoL of participants differed between groups, in favor of treatment with levofloxacin alone (SMD 0.52, 95% CI 0.04 to 1.01) (Analysis 20.2).
-
Clinical efficacy, defined as improvement of inflammatory findings (number of participants with leukocytosis in post‐massage urine specimens at the end of treatment), did not differ significantly between groups (RR 0.74, 95% CI 0.28 to 1.98) (Analysis 20.3).
-
Clinical efficacy, expressed as urinary peak flow rates (Qmax, mL/s), did not differ significantly between treatment groups (SMD 0.10, 95% CI ‐0.37 to 0.57) (Analysis 20.4).
Levofloxacin plus vardenafil at fixed daily dose versus levofloxacin plus vardenafil on‐demand
-
Microbiological efficacy (pathogen eradication at the end of therapy) did not differ significantly between groups (RR 1.02, 95% CI 0.90 to 1.16) (Analysis 21.1).
-
The NIH‐CPSI pain score did not differ significantly between groups when assessed at the end of therapy (SMD ‐0.09, 95% CI ‐0.55 to 0.38) (Analysis 21.2).
-
The scores of the NIH‐CPSI domains focusing on voiding symptoms and on the impact of the disease on the QoL of participants differed between groups, in favor of the fixed‐dose scheme (voiding score: SMD ‐0.64, 95% CI ‐1.11 to ‐0.16; QoL impact score: SMD ‐0.69, 95% CI ‐1.17 to ‐0.21) (Analysis 21.2).
-
Clinical efficacy, expressed as improvement of inflammatory findings (number of participants with leukocytosis in post‐massage urine specimens at the end of treatment), did not differ significantly between groups (RR 0.73, 95% CI 0.22 to 2.35) (Analysis 21.3).
-
Clinical efficacy, expressed as urinary peak flow rates (Qmax, mL/s), did not differ significantly between treatment groups (SMD 0.15, 95% CI ‐0.31 to 0.62) (Analysis 21.4).
Fluoroquinolone plus herbal extracts or supplements versus fluoroquinolone
In one study, 154 participants were randomized to receive prulifloxacin (600 mg once daily for two weeks) combined with the products ProstaMEV and FlogMEV (Serenoa repens, oral, 160 mg once daily; Urtica dioica, oral, 120 mg once daily; Curcuma longa, oral, 200 mg once daily; quercetin, oral, 100 mg once daily for two weeks), or prulifloxacin alone (600 mg once daily for two weeks) (Cai 2009).
-
Total NIH‐CPSI scores were significantly different between groups, when assessed both at the end of therapy (SMD ‐2.56, 95% CI ‐3.04 to ‐2.08) and after a six‐month follow‐up period (SMD ‐3.78, 95% CI ‐4.36 to ‐3.20) (Analysis 22.1). The comparison between groups was in favor of the combined therapy.
-
IPSS scores were significantly different between groups, when assessed both at the end of therapy (SMD ‐2.21, 95% CI ‐2.66 to ‐1.75) and after a six‐month follow‐up period (SMD ‐2.50, 95% CI ‐2.98 to ‐2.03) (Analysis 22.2). The comparison between groups was in favor of the combined therapy.
-
The rate of adverse effects did not differ significantly between the treatment groups (RR 1.05, 95% CI 0.11 to 9.76) (Analysis 22.3).
Discussion
Summary of main results
Therapy of infection caused by traditional pathogens
Fluoroquinolones are universally recommended as first‐line agents for CBP. The results of four out of five studies directly comparing two different fluoroquinolones indicate substantial equivalence between levofloxacin and ciprofloxacin, prulifloxacin and levofloxacin, lomefloxacin and ofloxacin, and lomefloxacin and ciprofloxacin. Equivalence was shown both at the microbiological (eradication of diverse causative pathogens) and clinical levels, at the end of treatment and at follow‐up (Bundrick 2003; Giannarini 2007; Koff 1996; Naber 2002). The rates of adverse effects of therapy also appeared to be equivalent in the compared treatment arms.
Pooled analysis of lomefloxacin versus comparator fluoroquinolones confirmed such equivalence (RR for microbiological efficacy at follow‐up 0.96, 95% CI 0.80 to 1.16) (summary of findings Table 2).
In contrast to the Bundrick trial (Bundrick 2003), the study by Zhang and colleagues indicated increased microbiological eradication rates and increased rates of cured or improved participants in the levofloxacin arm, both at the end of treatment and at the end of a six‐month follow‐up period (Zhang 2012). When the Bundrick and Zhang pooled studies were analyzed by a random‐effects model, the difference between levofloxacin and ciprofloxacin was not significant (microbiological efficacy: RR 1.18, 95% CI 0.81 to 1.71; clinical efficacy at the end of therapy: RR 1.16, 95% CI 0.93 to 1.46) (summary of findings Table for the main comparison). The discrepancy between these studies influenced the outcomes of pooled analyses 1, 6 and 7. Summary of findings tables 1, 3 and 4 present in a synthetic form the outcome of such meta‐analyses (summary of findings Table for the main comparison; summary of findings Table 3; summary of findings Table 4). The possible reasons for this discrepancy are discussed in the 'Quality of the evidence' section below.
Lomefloxacin is not inferior to co‐trimoxazole at both the microbiological and clinical levels (Bustillo 1997). To be effective, the latter agent should be administered for extended periods of time (six to 12 weeks) (Smith 1979).
Beta‐lactams were shown in two low powered studies to be not inferior to fluoroquinolones (ofloxacin) or tetracyclines (minocycline) at the microbiological and clinical levels (Cox 1989; Paulson 1986).
Therapy of infection caused by obligate or facultative intracellular pathogens
Macrolides were shown to be more effective than fluoroquinolones in chlamydial prostatitis. Microbiological and clinical outcomes were superior for azithromycin when compared to ciprofloxacin (Skerk 2003). The rate of adverse effects of therapy did not differ between the treatment arms.
Different macrolides, like azithromycin and clarithromycin, showed equivalent activity against chlamydial CBP (Skerk 2002). Therapy with thrice weekly doses of azithromycin (500 mg once daily) may last as little as three weeks without apparent loss of microbiological or clinical efficacy compared to longer courses of treatment (Skerk 2004b). Macrolides were also equivalent to tetracyclines in both chlamydial and ureaplasmal prostatitis, both at the microbiological and clinical levels (Skerk 2004a; Skerk 2006).
Fluoroquinolones (prulifloxacin) were shown to be as effective as tetracyclines (doxycycline) in chlamydial prostatitis (Cai 2010). Although prulifloxacin was more effective in attenuating clinical symptoms at the test‐of‐cure (TOC) visit, equivalent numbers of participants were asymptomatic at the same time point. Similarly, fluoroquinolones (ofloxacin) and tetracyclines (minocycline) show comparable microbiological and clinical efficacy and an equivalent safety profile in ureaplasmal prostatitis (Ohkawa 1993).
In summary, macrolides appear to be the most effective agents against CBP caused by intracellular pathogens.
Combination therapy ‐ all pathogens
Combination of a fluoroquinolone with a phosphodiesterase‐5 inhibitor (levofloxacin plus vardenafil) neither improves microbiological eradication nor attenuates pain or voiding symptoms when compared to therapy with the fluoroquinolone alone. However, the impact of the disease on patients' QoL is significantly improved by the sole fluoroquinolone when compared to therapy with the fluoroquinolone plus phosphodiesterase‐5 inhibitor on‐demand, though the difference was not observed when the phosphodiesterase‐5 inhibitor was administered at a fixed daily dose (10 mg once daily) (Aliaev 2008).
Combination of a fluoroquinolone (prulifloxacin) with various herbal preparations may attenuate clinical symptoms without increasing the rate of adverse effects (Cai 2009).
Overall completeness and applicability of evidence
The evidence resulting from this systematic review is applicable to patients broadly fulfilling the specific inclusion and exclusion criteria of the study.
Patients should be diagnosed and classified according to the NIH (Schaeffer 2004) or Drach's criteria (Drach 1978).
The microbiological diagnosis should be based on correctly performed standard lower urinary tract segmented tests (for example, 4‐glass or 2‐glass tests) for the isolation of causative pathogens from expressed prostatic secretions or post‐massage urine. A diagnosis of CBP based on the sole sperm or midstream urine culture is doubtful and methodologically incorrect.
The antimicrobial agents described in this review must be administered at the correct doses, and the therapy should be long‐term, as demonstrated in all included studies. This is an essential requirement for correct and effective applicability of the evidence described in this review.
The massive worldwide onset of chemoresistance that occurred in the last two decades has likely hindered the relevance and applicability in contemporary practice of evidence derived from studies focusing on drugs like co‐trimoxazole or extended‐spectrum beta‐lactamase (ESBL) targeted beta‐lactam antibiotics. This should be taken into account in clinical decision‐making and therapy design processes.
Quality of the evidence
The overall quality of the evidence described in this review is affected by the methodological limitations of the included studies. In particular, the more recent and better‐designed trials on novel fluoroquinolones (levofloxacin, lomefloxacin, prulifloxacin) were de facto characterized by equivalency or non‐inferiority designs. It is well‐known that non‐inferiority studies have a number of inherent weaknesses compared to superiority studies (Njue 2011).
Seven out of 18 studies described antimicrobial treatment against CBP caused by a single pathogen: five studies were focusing on chlamydial infections (Cai 2010; Skerk 2002; Skerk 2003; Skerk 2004a; Skerk 2004b), and two studies included only participants with CBP caused by Ureaplasma urealyticum (Ohkawa 1993; Skerk 2006). The remaining 11 studies included participants with infection caused by any pathogen (Gram‐positive or Gram‐negative). Pooled pathogens may represent a limitation and a confounding factor for the resulting evidence since certain antimicrobials are more active against a particular family or group of pathogens (for example, first‐generation fluoroquinolones are less active against Gram‐positive bacteria than fourth‐generation agents).
Three out of 18 included studies were double‐blinded (Bundrick 2003; Giannarini 2007; Smith 1979), and one was single‐blinded (Paulson 1986). Four studies assessed clinical symptoms using an internationally validated scoring system (NIH‐CPSI) (Aliaev 2008; Cai 2009; Cai 2010; Giannarini 2007). The remaining studies adopted non‐validated qualitative evaluation systems (for example, 'cure' versus 'failure'). Only two studies reported in detail the randomization procedure or the system adopted for allocation concealment (Bundrick 2003; Giannarini 2007).
A clinical outcome was absent in two studies (Koff 1996; Smith 1979), and adverse effects of treatment were not reported in three studies (Aliaev 2008; Ohkawa 1993; Paulson 1986). Eradication data at the TOC visit (end of therapy) were not disclosed in one study (Koff 1996).
In general, most of the included studies were characterized by a very low sample size.
Meta‐analysis was performed to compare levofloxacin versus ciprofloxacin (two studies: Analysis 1.1; summary of findings Table for the main comparison), lomefloxacin versus a comparator fluoroquinolone (Analysis 5.1; summary of findings Table 2), ciprofloxacin versus a comparator fluoroquinolone (Analysis 6.1; summary of findings Table 3), and levofloxacin versus a comparator fluoroquinolone (Analysis 7.1; summary of findings Table 4). Three out of four pooled analyses (Analysis 1.1; Analysis 6.1; Analysis 7.1) showed very high heterogeneity of microbiological outcomes (I2 = 94, 91 and 89, respectively). The Zhang 2012 study was identified as the likely source of heterogeneity. This trial included a fraction of participants with CBP caused by ciprofloxacin‐resistant pathogens (about 40% of the isolated pathogens). In the same patient population, 21% of isolates were resistant to levofloxacin. Thus, each group randomized to ciprofloxacin or levofloxacin contained unbalanced fractions of ciprofloxacin or levofloxacin resistant cases, and the lower eradication rate achieved by ciprofloxacin in this study is the probable source of substantial heterogeneity in the meta‐analysis. When the Zhang study was excluded from meta‐analysis of microbiological efficacy, heterogeneity became zero in pooled analyses 6.2.2 and 7.2.2. In addition, microbiological efficacy lost significance in pooled analyses 1, 6 and 7 when the original fixed‐effect model was changed to a random‐effects model (Analysis 1.2; Analysis 6.2; Analysis 7.2).
Agreements and disagreements with other studies or reviews
A single systematic review focusing on both chronic bacterial and abacterial prostatitis was retrieved from the PubMed database (Erickson 2008). This systematic review included both randomized and observational trials and did not contain a meta‐analysis. The primary outcomes of this review were symptom improvement, urodynamics, QoL, rates of bacteriological cure and adverse effects of treatment. The quality of the evidence was rated according to the GRADE criteria (Guyatt 2008). EMBASE, CENTRAL and PubMed international databases were searched.
The Bundrick 2003, Giannarini 2007 and Naber 2002 comparisons between different fluoroquinolones were analyzed. The conclusions drawn in the Erickson paper and in the present review are similar: lomefloxacin or levofloxacin are as effective as ciprofloxacin at increasing bacteriological cure rates, and prulifloxacin and levofloxacin are equally effective at increasing microbiological eradication rates in men with chronic bacterial prostatitis (Erickson 2008).
The Erickson review included a randomized study by Hu and coworkers focusing on intraprostatic administration of aminoglycosides (Hu 2002). We excluded the study from the present review because the description of the microbiological diagnostic methods was considered insufficient.
The present review differed from Erickson 2008 concerning the evaluation of the quality of evidence according to the GRADE system. In general, the quality rating given to the included studies is lower in the present systematic review. We attribute this differential evaluation to the downgrading effect of selection, performance, detection, attrition and reporting biases, assessed with the Cochrane Collaboration risk of bias tool.
Beta‐lactams were shown in two studies to be not inferior to fluoroquinolones (ofloxacin) or tetracyclines (minocycline) at the microbiological and clinical levels (Cox 1989; Paulson 1986). This evidence appears to be in contrast with more recent findings demonstrating very limited distribution, and hence low activity, of beta‐lactam antibiotics into the prostatic tissue (Charalabopoulos 2003).
Flow chart of included and excluded studies.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Comparison 1 Different fluoroquinolones: levofloxacin versus ciprofloxacin, Outcome 1 Microbiological efficacy ‐ pathogen eradication (fixed‐effect model).
Comparison 1 Different fluoroquinolones: levofloxacin versus ciprofloxacin, Outcome 2 Microbiological efficacy ‐ pathogen eradication (random‐effects model).
Comparison 1 Different fluoroquinolones: levofloxacin versus ciprofloxacin, Outcome 3 Clinical efficacy.
Comparison 1 Different fluoroquinolones: levofloxacin versus ciprofloxacin, Outcome 4 Adverse effects of treatment.
Comparison 2 Different fluoroquinolones: prulifloxacin versus levofloxacin, Outcome 1 Microbiological efficacy ‐ pathogen eradication.
Comparison 2 Different fluoroquinolones: prulifloxacin versus levofloxacin, Outcome 2 Clinical efficacy ‐ NIH‐CPSI total score at the end of treatment.
Comparison 2 Different fluoroquinolones: prulifloxacin versus levofloxacin, Outcome 3 Adverse effects of treatment.
Comparison 3 Different fluoroquinolones: lomefloxacin versus ofloxacin, Outcome 1 Microbiological efficacy ‐ pathogen eradication at follow‐up (6 months).
Comparison 3 Different fluoroquinolones: lomefloxacin versus ofloxacin, Outcome 2 Adverse effects of treatment.
Comparison 4 Different fluoroquinolones: lomefloxacin versus ciprofloxacin, Outcome 1 Microbiological efficacy (intention‐to‐treat analysis).
Comparison 4 Different fluoroquinolones: lomefloxacin versus ciprofloxacin, Outcome 2 Microbiological efficacy (per‐protocol analysis).
Comparison 4 Different fluoroquinolones: lomefloxacin versus ciprofloxacin, Outcome 3 Clinical efficacy (intention‐to‐treat analysis).
Comparison 4 Different fluoroquinolones: lomefloxacin versus ciprofloxacin, Outcome 4 Clinical efficacy (per‐protocol analysis).
Comparison 4 Different fluoroquinolones: lomefloxacin versus ciprofloxacin, Outcome 5 Adverse effects of treatment.
Comparison 5 Different fluoroquinolones: lomefloxacin versus comparator fluoroquinolone, Outcome 1 Microbiological efficacy ‐ pathogen eradication at follow‐up (6 months).
Comparison 5 Different fluoroquinolones: lomefloxacin versus comparator fluoroquinolone, Outcome 2 Adverse effects of treatment.
Comparison 6 Different fluoroquinolones: ciprofloxacin versus comparator fluoroquinolone, Outcome 1 Microbiological efficacy ‐ pathogen eradication at the end of treatment (fixed‐effect model).
Comparison 6 Different fluoroquinolones: ciprofloxacin versus comparator fluoroquinolone, Outcome 2 Microbiological efficacy ‐ pathogen eradication at the end of treatment (random‐effects model).
Comparison 6 Different fluoroquinolones: ciprofloxacin versus comparator fluoroquinolone, Outcome 3 Clinical efficacy.
Comparison 6 Different fluoroquinolones: ciprofloxacin versus comparator fluoroquinolone, Outcome 4 Adverse effects of treatment.
Comparison 7 Different fluoroquinolones: levofloxacin versus comparator fluoroquinolone, Outcome 1 Microbiological efficacy ‐ pathogen eradication (fixed‐effect model).
Comparison 7 Different fluoroquinolones: levofloxacin versus comparator fluoroquinolone, Outcome 2 Microbiological efficacy ‐ pathogen eradication (random‐effects model).
Comparison 7 Different fluoroquinolones: levofloxacin versus comparator fluoroquinolone, Outcome 3 Adverse effects of treatment.
Comparison 8 Fluoroquinolone versus other antibacterial agent: prulifloxacin versus doxycycline in chlamydial prostatitis, Outcome 1 Microbiological efficacy ‐ absence of Chlamydia trachomatis DNA and IgA at the end of treatment.
Comparison 8 Fluoroquinolone versus other antibacterial agent: prulifloxacin versus doxycycline in chlamydial prostatitis, Outcome 2 Clinical efficacy ‐ NIH‐CPSI total score at the end of treatment.
Comparison 8 Fluoroquinolone versus other antibacterial agent: prulifloxacin versus doxycycline in chlamydial prostatitis, Outcome 3 Clinical efficacy ‐ number of asymptomatic participants at the end of therapy.
Comparison 8 Fluoroquinolone versus other antibacterial agent: prulifloxacin versus doxycycline in chlamydial prostatitis, Outcome 4 Adverse effects of treatment.
Comparison 9 Fluoroquinolone versus other antibacterial agent: ofloxacin versus minocycline in ureaplasmal prostatitis, Outcome 1 Microbiological efficacy ‐ pathogen eradication.
Comparison 9 Fluoroquinolone versus other antibacterial agent: ofloxacin versus minocycline in ureaplasmal prostatitis, Outcome 2 Clinical efficacy (cure or improvement) at the end of treatment.
Comparison 10 Fluoroquinolone versus other antibacterial agent: ofloxacin versus carbenicillin, Outcome 1 Microbiological efficacy ‐ pathogen eradication.
Comparison 10 Fluoroquinolone versus other antibacterial agent: ofloxacin versus carbenicillin, Outcome 2 Clinical efficacy (cure or improvement) at the end of treatment.
Comparison 10 Fluoroquinolone versus other antibacterial agent: ofloxacin versus carbenicillin, Outcome 3 Adverse effects of treatment.
Comparison 11 Fluoroquinolone versus other antibacterial agent: lomefloxacin versus co‐trimoxazole, Outcome 1 Microbiological efficacy.
Comparison 11 Fluoroquinolone versus other antibacterial agent: lomefloxacin versus co‐trimoxazole, Outcome 2 Clinical efficacy.
Comparison 11 Fluoroquinolone versus other antibacterial agent: lomefloxacin versus co‐trimoxazole, Outcome 3 Adverse effects of treatment.
Comparison 12 Fluoroquinolone versus other antibacterial agent: ciprofloxacin versus azithromycin in chlamydial prostatitis, Outcome 1 Microbiological efficacy (pathogen eradication) at the end of treatment.
Comparison 12 Fluoroquinolone versus other antibacterial agent: ciprofloxacin versus azithromycin in chlamydial prostatitis, Outcome 2 Clinical efficacy (cure or improvement) at the end of treatment.
Comparison 12 Fluoroquinolone versus other antibacterial agent: ciprofloxacin versus azithromycin in chlamydial prostatitis, Outcome 3 Adverse effects of treatment.
Comparison 13 Non‐fluoroquinolone antibacterial agents: minocycline versus cephalexin, Outcome 1 Microbiological efficacy (pathogen eradication and eradication plus superinfection) at the end of treatment.
Comparison 13 Non‐fluoroquinolone antibacterial agents: minocycline versus cephalexin, Outcome 2 Clinical efficacy (cure or improvement) at the end of treatment.
Comparison 13 Non‐fluoroquinolone antibacterial agents: minocycline versus cephalexin, Outcome 3 Microbiological recurrence.
Comparison 14 Non‐fluoroquinolone antibacterial agents: azithromycin versus clarithromycin in chlamydial prostatitis, Outcome 1 Microbiological efficacy (pathogen eradication) at the end of treatment.
Comparison 14 Non‐fluoroquinolone antibacterial agents: azithromycin versus clarithromycin in chlamydial prostatitis, Outcome 2 Clinical efficacy (cure) at the end of treatment.
Comparison 14 Non‐fluoroquinolone antibacterial agents: azithromycin versus clarithromycin in chlamydial prostatitis, Outcome 3 Adverse effects of treatment.
Comparison 15 Non‐fluoroquinolone antibacterial agents: azithromycin versus doxycycline in chlamydial prostatitis, Outcome 1 Microbiological efficacy (pathogen eradication) at the end of treatment.
Comparison 15 Non‐fluoroquinolone antibacterial agents: azithromycin versus doxycycline in chlamydial prostatitis, Outcome 2 Clinical efficacy.
Comparison 15 Non‐fluoroquinolone antibacterial agents: azithromycin versus doxycycline in chlamydial prostatitis, Outcome 3 Adverse effects of treatment.
Comparison 16 Non‐fluoroquinolone antibacterial agents: azithromycin versus doxycycline in ureaplasmal prostatitis, Outcome 1 Microbiological efficacy (pathogen eradication) at the end of treatment.
Comparison 16 Non‐fluoroquinolone antibacterial agents: azithromycin versus doxycycline in ureaplasmal prostatitis, Outcome 2 Clinical efficacy (cure) at the end of treatment.
Comparison 16 Non‐fluoroquinolone antibacterial agents: azithromycin versus doxycycline in ureaplasmal prostatitis, Outcome 3 Adverse effects of treatment.
Comparison 17 Different dosing regimens: azithromycin 4.5 g versus 6.0 g total doses in chlamydial prostatitis, Outcome 1 Microbiological efficacy (pathogen eradication) at the end of treatment.
Comparison 17 Different dosing regimens: azithromycin 4.5 g versus 6.0 g total doses in chlamydial prostatitis, Outcome 2 Clinical efficacy (cure) at the end of therapy.
Comparison 17 Different dosing regimens: azithromycin 4.5 g versus 6.0 g total doses in chlamydial prostatitis, Outcome 3 Adverse effects of treatment.
Comparison 18 Different therapy duration: co‐trimoxazole 480 mg twice daily for 12 weeks versus 10 days, Outcome 1 Microbiological efficacy (pathogen eradication) at the end of treatment.
Comparison 18 Different therapy duration: co‐trimoxazole 480 mg twice daily for 12 weeks versus 10 days, Outcome 2 Adverse effects of treatment.
Comparison 19 Fluoroquinolone combined with phosphodiesterase‐5 inhibitor versus fluoroquinolone: levofloxacin plus vardenafil 10 mg/day versus levofloxacin, Outcome 1 Microbiological efficacy (pathogen eradication) at the end of treatment.
Comparison 19 Fluoroquinolone combined with phosphodiesterase‐5 inhibitor versus fluoroquinolone: levofloxacin plus vardenafil 10 mg/day versus levofloxacin, Outcome 2 Clinical efficacy ‐ NIH‐CPSI score at the end of treatment.
Comparison 19 Fluoroquinolone combined with phosphodiesterase‐5 inhibitor versus fluoroquinolone: levofloxacin plus vardenafil 10 mg/day versus levofloxacin, Outcome 3 Clinical efficacy ‐ number of participants with leukocytosis in post‐massage urine specimens at the end of treatment.
Comparison 19 Fluoroquinolone combined with phosphodiesterase‐5 inhibitor versus fluoroquinolone: levofloxacin plus vardenafil 10 mg/day versus levofloxacin, Outcome 4 Clinical efficacy ‐ urine peak flow rate at the end of treatment (mL/s).
Comparison 20 Fluoroquinolone combined with phosphodiesterase‐5 inhibitor versus fluoroquinolone: levofloxacin plus vardenafil 10 mg on‐demand versus levofloxacin, Outcome 1 Microbiological efficacy (pathogen eradication) at the end of treatment.
Comparison 20 Fluoroquinolone combined with phosphodiesterase‐5 inhibitor versus fluoroquinolone: levofloxacin plus vardenafil 10 mg on‐demand versus levofloxacin, Outcome 2 Clinical efficacy ‐ NIH‐CPSI score at the end of treatment.
Comparison 20 Fluoroquinolone combined with phosphodiesterase‐5 inhibitor versus fluoroquinolone: levofloxacin plus vardenafil 10 mg on‐demand versus levofloxacin, Outcome 3 Clinical efficacy ‐ number of participants with leukocytosis in post‐massage urine specimens at the end of treatment.
Comparison 20 Fluoroquinolone combined with phosphodiesterase‐5 inhibitor versus fluoroquinolone: levofloxacin plus vardenafil 10 mg on‐demand versus levofloxacin, Outcome 4 Clinical efficacy ‐ urine peak flow rate at the end of treatment (mL/s).
Comparison 21 Fluoroquinolone combined with phosphodiesterase‐5 inhibitor: levofloxacin plus vardenafil 10 mg/day versus levofloxacin plus vardenafil 10 mg on‐demand, Outcome 1 Microbiological efficacy (pathogen eradication) at the end of treatment.
Comparison 21 Fluoroquinolone combined with phosphodiesterase‐5 inhibitor: levofloxacin plus vardenafil 10 mg/day versus levofloxacin plus vardenafil 10 mg on‐demand, Outcome 2 Clinical efficacy ‐ NIH‐CPSI score at the end of treatment.
Comparison 21 Fluoroquinolone combined with phosphodiesterase‐5 inhibitor: levofloxacin plus vardenafil 10 mg/day versus levofloxacin plus vardenafil 10 mg on‐demand, Outcome 3 Clinical efficacy ‐ number of participants with leukocytosis in post‐massage urine specimens at the end of treatment.
Comparison 21 Fluoroquinolone combined with phosphodiesterase‐5 inhibitor: levofloxacin plus vardenafil 10 mg/day versus levofloxacin plus vardenafil 10 mg on‐demand, Outcome 4 Clinical efficacy ‐ urine peak flow rate at the end of treatment (mL/s).
Comparison 22 Fluoroquinolone plus herbal extracts or supplements versus fluoroquinolone: prulifloxacin plus supplements versus prulifloxacin, Outcome 1 Clinical efficacy ‐ NIH‐CPSI total score.
Comparison 22 Fluoroquinolone plus herbal extracts or supplements versus fluoroquinolone: prulifloxacin plus supplements versus prulifloxacin, Outcome 2 Clinical efficacy ‐ IPSS score.
Comparison 22 Fluoroquinolone plus herbal extracts or supplements versus fluoroquinolone: prulifloxacin plus supplements versus prulifloxacin, Outcome 3 Adverse effects of treatment.
| Levofloxacin versus ciprofloxacin for chronic bacterial prostatitis | |||||
| Patient or population: patients with chronic bacterial prostatitis | |||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No. of participants | Quality of the evidence | |
| Assumed risk | Corresponding risk | ||||
| Ciprofloxacin | Levofloxacin | ||||
| Microbiological efficacy ‐ pathogen eradication | 667 per 1000 | 787 per 1000 | RR 1.18 | 669 | ⊕⊝⊝⊝ |
| Clinical efficacy ‐ cure or improvement at end of treatment | 722 per 1000 | 838 per 1000 | RR 1.16 | 669 | ⊕⊕⊝⊝ |
| Clinical efficacy ‐ cure or improvement at follow‐up (6 months) | 710 per 1000 | 823 per 1000 | RR 1.16 | 669 | ⊕⊝⊝⊝ |
| Adverse effects of treatment ‐ any adverse effects | 266 per 1000 | 229 per 1000 | RR 0.86 | 785 | ⊕⊕⊕⊝ |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence | |||||
| 1Bundrick 2003 ‐ high risk of reporting bias. | |||||
| Lomefloxacin versus comparator fluoroquinolone for chronic bacterial prostatitis | ||||||
| Patient or population: patients with chronic bacterial prostatitis | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No. of participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Comparator fluoroquinolone | Lomefloxacin | |||||
| Microbiological efficacy ‐ pathogen eradication at follow‐up (6 months) | 804 per 1000 | 771 per 1000 | RR 0.96 | 116 | ⊕⊕⊝⊝ | |
| Clinical efficacy ‐ cure or improvement at end of treatment | See comment | See comment | Not estimable | 0 (0) | See comment | No study reported or provided useable data for this outcome |
| Clinical efficacy ‐ cure or improvement at follow‐up (6 months) | See comment | See comment | Not estimable | 0 (0) | See comment | No study reported or provided useable data for this outcome |
| Adverse effects of treatment ‐ any adverse effects | 212 per 1000 | 135 per 1000 | RR 0.64 | 215 | ⊕⊕⊕⊝ | |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 The comparator fluoroquinolone was ofloxacin (Koff 1996) or ciprofloxacin (Naber 2002). | ||||||
| Ciprofloxacin versus comparator fluoroquinolone for chronic bacterial prostatitis | |||||
| Patient or population: patients with chronic bacterial prostatitis | |||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No. of participants | Quality of the evidence | |
| Assumed risk | Corresponding risk | ||||
| Comparator fluoroquinolone | Ciprofloxacin | ||||
| Microbiological efficacy ‐ pathogen eradication at end of treatment | 806 per 1000 | 733 per 1000 | RR 0.91 | 851 | ⊕⊝⊝⊝ |
| Clinical efficacy ‐ cure or improvement at end of treatment | 879 per 1000 | 791 per 1000 | RR 0.9 | 851 | ⊕⊝⊝⊝ |
| Clinical efficacy ‐ cure or improvement at follow‐up (6 months) | 808 per 1000 | 752 per 1000 | RR 0.93 | 851 | ⊕⊝⊝⊝ |
| Adverse effects of treatment ‐ any adverse effects | 212 per 1000 | 246 per 1000 | RR 1.16 | 967 | ⊕⊕⊕⊝ |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence | |||||
| 1 The comparator fluoroquinolone was levofloxacin (Bundrick 2003; Zhang 2012) or lomefloxacin (Naber 2002). | |||||
| Levofloxacin versus comparator fluoroquinolone for chronic bacterial prostatitis | ||||||
| Patient or population: patients with chronic bacterial prostatitis | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No. of participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Comparator fluoroquinolone | Levofloxacin | |||||
| Microbiological efficacy ‐ pathogen eradication | 674 per 1000 | 755 per 1000 | RR 1.12 | 758 | ⊕⊝⊝⊝ | |
| Clinical efficacy ‐ cure or improvement at end of treatment | See comment | See comment | Not estimable | 0 (0) | See comment | No study reported or provided useable data for this outcome |
| Clinical efficacy ‐ cure or improvement at follow‐up (6 months) | See comment | See comment | Not estimable | 0 (0) | See comment | No study reported or provided useable data for this outcome |
| Adverse effects of treatment ‐ any adverse effects | 258 per 1000 | 227 per 1000 | RR 0.88 | 874 | ⊕⊕⊕⊝ | |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 The comparator fluoroquinolone was ciprofloxacin (Bundrick 2003; Zhang 2012) or prulifloxacin (Giannarini 2007). | ||||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy ‐ pathogen eradication (fixed‐effect model) Show forest plot | 2 | 669 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.22 [1.11, 1.34] |
| 2 Microbiological efficacy ‐ pathogen eradication (random‐effects model) Show forest plot | 2 | 669 | Risk Ratio (M‐H, Random, 95% CI) | 1.18 [0.81, 1.71] |
| 3 Clinical efficacy Show forest plot | 2 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 3.1 Clinical efficacy (cure or improvement) at the end of treatment | 2 | 669 | Risk Ratio (M‐H, Random, 95% CI) | 1.16 [0.93, 1.46] |
| 3.2 Clinical efficacy (cure or improvement) at follow‐up (6 months) | 2 | 669 | Risk Ratio (M‐H, Random, 95% CI) | 1.16 [0.86, 1.55] |
| 4 Adverse effects of treatment Show forest plot | 2 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 4.1 Any adverse effects | 2 | 785 | Risk Ratio (M‐H, Random, 95% CI) | 0.86 [0.70, 1.06] |
| 4.2 Gastrointestinal adverse effects | 2 | 785 | Risk Ratio (M‐H, Random, 95% CI) | 0.86 [0.42, 1.77] |
| 4.3 Back pain | 1 | 377 | Risk Ratio (M‐H, Random, 95% CI) | 0.55 [0.13, 2.26] |
| 4.4 Headache | 2 | 785 | Risk Ratio (M‐H, Random, 95% CI) | 0.83 [0.39, 1.76] |
| 4.5 Dizziness | 2 | 785 | Risk Ratio (M‐H, Random, 95% CI) | 0.66 [0.02, 22.13] |
| 4.6 Arthralgia | 1 | 377 | Risk Ratio (M‐H, Random, 95% CI) | 1.46 [0.49, 4.39] |
| 4.7 Myalgia | 1 | 377 | Risk Ratio (M‐H, Random, 95% CI) | 1.07 [0.37, 3.11] |
| 4.8 Skeletal pain | 1 | 377 | Risk Ratio (M‐H, Random, 95% CI) | 0.55 [0.13, 2.26] |
| 4.9 Rhinitis | 1 | 377 | Risk Ratio (M‐H, Random, 95% CI) | 0.91 [0.27, 3.10] |
| 4.10 Upper respiratory tract infection | 1 | 377 | Risk Ratio (M‐H, Random, 95% CI) | 1.14 [0.31, 4.19] |
| 4.11 Dermal toxicity | 2 | 785 | Risk Ratio (M‐H, Random, 95% CI) | 0.47 [0.12, 1.90] |
| 4.12 Allergy to experimental agents | 1 | 408 | Risk Ratio (M‐H, Random, 95% CI) | 2.86 [0.12, 69.73] |
| 4.13 Leukopenia | 1 | 408 | Risk Ratio (M‐H, Random, 95% CI) | 0.32 [0.01, 7.75] |
| 4.14 Cough | 1 | 408 | Risk Ratio (M‐H, Random, 95% CI) | 2.86 [0.12, 69.73] |
| 4.15 Insomnia | 1 | 408 | Risk Ratio (M‐H, Random, 95% CI) | 0.48 [0.04, 5.21] |
| 4.16 Altered transaminase levels | 1 | 408 | Risk Ratio (M‐H, Random, 95% CI) | 0.48 [0.04, 5.21] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy ‐ pathogen eradication Show forest plot | 1 | 89 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.02 [0.79, 1.33] |
| 2 Clinical efficacy ‐ NIH‐CPSI total score at the end of treatment Show forest plot | 1 | 89 | Std. Mean Difference (IV, Random, 95% CI) | ‐0.03 [‐0.45, 0.39] |
| 3 Adverse effects of treatment Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 3.1 Any adverse effects | 1 | 89 | Risk Ratio (M‐H, Random, 95% CI) | 0.82 [0.36, 1.88] |
| 3.2 Gastrointestinal adverse effects | 1 | 89 | Risk Ratio (M‐H, Random, 95% CI) | 0.73 [0.25, 2.13] |
| 3.3 Dermal toxicity | 1 | 89 | Risk Ratio (M‐H, Random, 95% CI) | 7.16 [0.38, 134.62] |
| 3.4 Headache | 1 | 89 | Risk Ratio (M‐H, Random, 95% CI) | 0.15 [0.01, 2.75] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy ‐ pathogen eradication at follow‐up (6 months) Show forest plot | 1 | 33 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.11 [0.66, 1.88] |
| 2 Adverse effects of treatment Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 2.1 Any adverse effects | 1 | 33 | Risk Ratio (M‐H, Random, 95% CI) | 0.42 [0.16, 1.12] |
| 2.2 Gastrointestinal adverse effects | 1 | 33 | Risk Ratio (M‐H, Random, 95% CI) | 0.56 [0.19, 1.61] |
| 2.3 Headache | 1 | 33 | Risk Ratio (M‐H, Random, 95% CI) | 0.28 [0.01, 6.43] |
| 2.4 Dizziness | 1 | 33 | Risk Ratio (M‐H, Random, 95% CI) | 0.28 [0.01, 6.43] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy (intention‐to‐treat analysis) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 1.1 Microbiological efficacy (pathogen eradication) at the end of treatment | 1 | 182 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.96 [0.82, 1.11] |
| 1.2 Microbiological efficacy (pathogen eradication) at follow‐up (4 weeks) | 1 | 182 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.87 [0.72, 1.06] |
| 1.3 Microbiological efficacy (pathogen eradication) at follow‐up (3 months) | 1 | 182 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.90 [0.74, 1.09] |
| 1.4 Microbiological efficacy (pathogen eradication) at follow‐up (6 months) | 1 | 182 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.87 [0.67, 1.12] |
| 2 Microbiological efficacy (per‐protocol analysis) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 2.1 Microbiological efficacy (pathogen eradication and eradication plus superinfection) at the end of treatment | 1 | 87 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.98 [0.89, 1.09] |
| 2.2 Microbiological efficacy (pathogen eradication and eradication plus superinfection) at follow‐up (4 weeks) | 1 | 83 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.00 [0.94, 1.07] |
| 2.3 Microbiological efficacy (pathogen eradication and eradication plus superinfection) at follow‐up (3 months) | 1 | 77 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.03 [0.94, 1.12] |
| 2.4 Microbiological efficacy (pathogen eradication and eradication plus superinfection) at follow‐up (6 months) | 1 | 66 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.94 [0.80, 1.09] |
| 3 Clinical efficacy (intention‐to‐treat analysis) Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 3.1 Clinical efficacy (cure or improvement) at the end of treatment | 1 | 182 | Risk Ratio (M‐H, Random, 95% CI) | 1.01 [0.94, 1.09] |
| 3.2 Clinical efficacy (cure or improvement) at follow‐up (4 weeks) | 1 | 182 | Risk Ratio (M‐H, Random, 95% CI) | 0.91 [0.78, 1.05] |
| 3.3 Clinical efficacy (cure or improvement) at follow‐up (3 months) | 1 | 182 | Risk Ratio (M‐H, Random, 95% CI) | 0.97 [0.82, 1.15] |
| 3.4 Clinical efficacy (cure or improvement) at follow‐up (6 months) | 1 | 182 | Risk Ratio (M‐H, Random, 95% CI) | 0.91 [0.75, 1.11] |
| 4 Clinical efficacy (per‐protocol analysis) Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 4.1 Clinical efficacy (cure or improvement) at the end of treatment | 1 | 87 | Risk Ratio (M‐H, Random, 95% CI) | 0.96 [0.89, 1.03] |
| 4.2 Clinical efficacy (cure or improvement) at follow‐up (4 weeks) | 1 | 83 | Risk Ratio (M‐H, Random, 95% CI) | 1.00 [0.86, 1.18] |
| 4.3 Clinical efficacy (cure or improvement) at follow‐up (3 months) | 1 | 65 | Risk Ratio (M‐H, Random, 95% CI) | 1.07 [0.94, 1.21] |
| 4.4 Clinical efficacy (cure or improvement) at follow‐up (6 months) | 1 | 66 | Risk Ratio (M‐H, Random, 95% CI) | 0.88 [0.77, 1.01] |
| 5 Adverse effects of treatment Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 5.1 Any adverse effects | 1 | 182 | Risk Ratio (M‐H, Random, 95% CI) | 0.82 [0.40, 1.68] |
| 5.2 Gastrointestinal adverse effects | 1 | 182 | Risk Ratio (M‐H, Random, 95% CI) | 0.60 [0.20, 1.76] |
| 5.3 Headache | 1 | 182 | Risk Ratio (M‐H, Random, 95% CI) | 0.96 [0.06, 15.07] |
| 5.4 Dizziness | 1 | 182 | Risk Ratio (M‐H, Random, 95% CI) | 2.87 [0.12, 69.59] |
| 5.5 Dry mouth | 1 | 182 | Risk Ratio (M‐H, Random, 95% CI) | 4.79 [0.23, 98.35] |
| 5.6 Insomnia | 1 | 182 | Risk Ratio (M‐H, Random, 95% CI) | 0.32 [0.01, 7.73] |
| 5.7 Hyperglycemia | 1 | 182 | Risk Ratio (M‐H, Random, 95% CI) | 2.87 [0.12, 69.59] |
| 5.8 Dermal toxicity | 1 | 182 | Risk Ratio (M‐H, Random, 95% CI) | 2.87 [0.12, 69.59] |
| 5.9 Abnormal semen | 1 | 182 | Risk Ratio (M‐H, Random, 95% CI) | 2.87 [0.12, 69.59] |
| 5.10 Upper respiratory tract infection | 1 | 182 | Risk Ratio (M‐H, Random, 95% CI) | 0.11 [0.01, 1.95] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy ‐ pathogen eradication at follow‐up (6 months) Show forest plot | 2 | 116 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.96 [0.80, 1.16] |
| 2 Adverse effects of treatment Show forest plot | 2 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 2.1 Any adverse effects | 2 | 215 | Risk Ratio (M‐H, Random, 95% CI) | 0.64 [0.34, 1.21] |
| 2.2 Gastrointestinal adverse effects | 2 | 215 | Risk Ratio (M‐H, Random, 95% CI) | 0.58 [0.27, 1.23] |
| 2.3 Headache | 2 | 215 | Risk Ratio (M‐H, Random, 95% CI) | 0.56 [0.07, 4.43] |
| 2.4 Dizziness | 2 | 215 | Risk Ratio (M‐H, Random, 95% CI) | 0.88 [0.09, 8.60] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy ‐ pathogen eradication at the end of treatment (fixed‐effect model) Show forest plot | 3 | 851 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.87 [0.80, 0.94] |
| 2 Microbiological efficacy ‐ pathogen eradication at the end of treatment (random‐effects model) Show forest plot | 3 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 2.1 All studies | 3 | 851 | Risk Ratio (M‐H, Random, 95% CI) | 0.91 [0.70, 1.18] |
| 2.2 Sensitivity analysis, exclusion of Zhang 2012 | 2 | 443 | Risk Ratio (M‐H, Random, 95% CI) | 1.03 [0.93, 1.14] |
| 3 Clinical efficacy Show forest plot | 3 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 3.1 Clinical efficacy (cure or improvement) at the end of treatment | 3 | 851 | Risk Ratio (M‐H, Random, 95% CI) | 0.90 [0.75, 1.08] |
| 3.2 Clinical efficacy (cure or improvement) at follow‐up (6 months) | 3 | 851 | Risk Ratio (M‐H, Random, 95% CI) | 0.93 [0.72, 1.20] |
| 4 Adverse effects of treatment Show forest plot | 3 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 4.1 Any adverse effects | 3 | 967 | Risk Ratio (M‐H, Random, 95% CI) | 1.16 [0.95, 1.42] |
| 4.2 Gastrointestinal adverse effects | 3 | 967 | Risk Ratio (M‐H, Random, 95% CI) | 1.10 [0.76, 1.59] |
| 4.3 Headache | 3 | 967 | Risk Ratio (M‐H, Random, 95% CI) | 1.19 [0.58, 2.46] |
| 4.4 Dizziness | 3 | 967 | Risk Ratio (M‐H, Random, 95% CI) | 1.04 [0.09, 12.02] |
| 4.5 Dermal toxicity | 3 | 967 | Risk Ratio (M‐H, Random, 95% CI) | 1.59 [0.44, 5.75] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy ‐ pathogen eradication (fixed‐effect model) Show forest plot | 3 | 758 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.19 [1.09, 1.30] |
| 2 Microbiological efficacy ‐ pathogen eradication (random‐effects model) Show forest plot | 3 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 2.1 All studies | 3 | 758 | Risk Ratio (M‐H, Random, 95% CI) | 1.12 [0.84, 1.48] |
| 2.2 Sensitivity analysis, exclusion of Zhang 2012 | 2 | 350 | Risk Ratio (M‐H, Random, 95% CI) | 0.98 [0.87, 1.10] |
| 3 Adverse effects of treatment Show forest plot | 3 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 3.1 Any adverse effects | 3 | 874 | Risk Ratio (M‐H, Random, 95% CI) | 0.88 [0.72, 1.08] |
| 3.2 Gastrointestinal adverse effects | 3 | 874 | Risk Ratio (M‐H, Random, 95% CI) | 1.00 [0.69, 1.44] |
| 3.3 Dermal toxicity | 3 | 874 | Risk Ratio (M‐H, Random, 95% CI) | 0.37 [0.11, 1.32] |
| 3.4 Headache | 3 | 874 | Risk Ratio (M‐H, Random, 95% CI) | 1.00 [0.26, 3.80] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy ‐ absence of Chlamydia trachomatis DNA and IgA at the end of treatment Show forest plot | 1 | 38 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.12 [0.93, 1.36] |
| 2 Clinical efficacy ‐ NIH‐CPSI total score at the end of treatment Show forest plot | 1 | 211 | Std. Mean Difference (IV, Random, 95% CI) | ‐0.66 [‐0.94, ‐0.39] |
| 3 Clinical efficacy ‐ number of asymptomatic participants at the end of therapy Show forest plot | 1 | 211 | Risk Ratio (M‐H, Random, 95% CI) | 1.04 [0.91, 1.19] |
| 4 Adverse effects of treatment Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 4.1 Any adverse effects | 1 | 211 | Risk Ratio (M‐H, Random, 95% CI) | 1.17 [0.32, 4.24] |
| 4.2 Gastrointestinal adverse effects | 1 | 211 | Risk Ratio (M‐H, Random, 95% CI) | 0.70 [0.16, 3.06] |
| 4.3 Back pain | 1 | 211 | Risk Ratio (M‐H, Random, 95% CI) | 4.68 [0.23, 96.36] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy ‐ pathogen eradication Show forest plot | 1 | 14 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.0 [0.78, 1.29] |
| 2 Clinical efficacy (cure or improvement) at the end of treatment Show forest plot | 1 | 14 | Risk Ratio (M‐H, Random, 95% CI) | 0.87 [0.59, 1.26] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy ‐ pathogen eradication Show forest plot | 1 | 23 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.04 [0.76, 1.42] |
| 2 Clinical efficacy (cure or improvement) at the end of treatment Show forest plot | 1 | 23 | Risk Ratio (M‐H, Random, 95% CI) | 1.06 [0.85, 1.32] |
| 3 Adverse effects of treatment Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 3.1 Any adverse effects | 1 | 46 | Risk Ratio (M‐H, Random, 95% CI) | 0.73 [0.31, 1.71] |
| 3.2 Gastrointestinal adverse effects | 1 | 46 | Risk Ratio (M‐H, Random, 95% CI) | 0.47 [0.14, 1.59] |
| 3.3 Dermal toxicity | 1 | 46 | Risk Ratio (M‐H, Random, 95% CI) | 3.26 [0.14, 76.10] |
| 3.4 Nervous (sic) | 1 | 46 | Risk Ratio (M‐H, Random, 95% CI) | 5.43 [0.28, 107.33] |
| 3.5 Special senses toxicity | 1 | 46 | Risk Ratio (M‐H, Random, 95% CI) | 0.16 [0.01, 2.85] |
| 3.6 Respiratory toxicity | 1 | 46 | Risk Ratio (M‐H, Random, 95% CI) | 0.36 [0.02, 8.46] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 1.1 Microbiological success (pathogen eradication) at the end of treatment | 1 | 26 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.09 [0.82, 1.44] |
| 1.2 Microbiological success (pathogen eradication) at follow‐up (4 months) | 1 | 26 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.09 [0.82, 1.44] |
| 2 Clinical efficacy Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 2.1 Clinical efficacy (cure or improvement) at the end of treatment | 1 | 26 | Risk Ratio (M‐H, Random, 95% CI) | 1.0 [0.87, 1.15] |
| 2.2 Clinical efficacy (cure or improvement) at follow‐up (4 months) | 1 | 26 | Risk Ratio (M‐H, Random, 95% CI) | 1.0 [0.87, 1.15] |
| 3 Adverse effects of treatment Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 3.1 Any adverse effects | 1 | 28 | Risk Ratio (M‐H, Random, 95% CI) | 0.43 [0.04, 4.25] |
| 3.2 Gastrointestinal adverse effects | 1 | 28 | Risk Ratio (M‐H, Random, 95% CI) | 0.43 [0.04, 4.25] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy (pathogen eradication) at the end of treatment Show forest plot | 1 | 89 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.48 [0.32, 0.72] |
| 2 Clinical efficacy (cure or improvement) at the end of treatment Show forest plot | 1 | 89 | Risk Ratio (M‐H, Random, 95% CI) | 0.64 [0.46, 0.90] |
| 3 Adverse effects of treatment Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 3.1 Any adverse effects | 1 | 89 | Risk Ratio (M‐H, Random, 95% CI) | 0.34 [0.01, 8.15] |
| 3.2 Gastrointestinal adverse effects | 1 | 89 | Risk Ratio (M‐H, Random, 95% CI) | 0.34 [0.01, 8.15] |
| 3.3 Hepatic adverse effects (increased transaminases) | 1 | 89 | Risk Ratio (M‐H, Random, 95% CI) | 0.34 [0.01, 8.15] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy (pathogen eradication and eradication plus superinfection) at the end of treatment Show forest plot | 1 | 27 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.7 [0.54, 5.34] |
| 2 Clinical efficacy (cure or improvement) at the end of treatment Show forest plot | 1 | 27 | Risk Ratio (M‐H, Random, 95% CI) | 2.04 [0.83, 4.99] |
| 3 Microbiological recurrence Show forest plot | 1 | 20 | Risk Ratio (M‐H, Random, 95% CI) | 0.98 [0.37, 2.59] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy (pathogen eradication) at the end of treatment Show forest plot | 1 | 91 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.01 [0.82, 1.23] |
| 2 Clinical efficacy (cure) at the end of treatment Show forest plot | 1 | 91 | Risk Ratio (M‐H, Random, 95% CI) | 0.98 [0.75, 1.28] |
| 3 Adverse effects of treatment Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 3.1 Any adverse effects | 1 | 91 | Risk Ratio (M‐H, Random, 95% CI) | 1.96 [0.18, 20.83] |
| 3.2 Gastrointestinal adverse effects | 1 | 91 | Risk Ratio (M‐H, Random, 95% CI) | 1.96 [0.18, 20.83] |
| 3.3 Hepatic adverse effects (increased transaminases) | 1 | 91 | Risk Ratio (M‐H, Random, 95% CI) | 1.96 [0.18, 20.83] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy (pathogen eradication) at the end of treatment Show forest plot | 1 | 125 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.03 [0.85, 1.26] |
| 2 Clinical efficacy Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 2.1 Clinical efficacy ‐ presence of inflammatory findings (number of participants with white blood cell counts in EPS/VB3 < 10 per high power field) at the end of therapy | 1 | 125 | Risk Ratio (M‐H, Random, 95% CI) | 1.08 [0.66, 1.78] |
| 2.2 Clinical efficacy (cure or improvement) at the end of therapy | 1 | 125 | Risk Ratio (M‐H, Random, 95% CI) | 0.95 [0.76, 1.19] |
| 3 Adverse effects of treatment Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 3.1 Any adverse effects | 1 | 125 | Risk Ratio (M‐H, Random, 95% CI) | 0.21 [0.04, 1.04] |
| 3.2 Gastrointestinal adverse effects | 1 | 125 | Risk Ratio (M‐H, Random, 95% CI) | 0.21 [0.04, 1.04] |
| 3.3 Hepatic adverse effects (increased transaminases) | 1 | 125 | Risk Ratio (M‐H, Random, 95% CI) | 2.65 [0.13, 54.00] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy (pathogen eradication) at the end of treatment Show forest plot | 1 | 63 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.05 [0.80, 1.39] |
| 2 Clinical efficacy (cure) at the end of treatment Show forest plot | 1 | 63 | Risk Ratio (M‐H, Random, 95% CI) | 1.01 [0.72, 1.42] |
| 3 Adverse effects of treatment Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 3.1 Any adverse effects | 1 | 63 | Risk Ratio (M‐H, Random, 95% CI) | 0.09 [0.01, 1.53] |
| 3.2 Gastrointestinal adverse effects | 1 | 63 | Risk Ratio (M‐H, Random, 95% CI) | 0.09 [0.01, 1.53] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy (pathogen eradication) at the end of treatment Show forest plot | 1 | 89 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.99 [0.81, 1.21] |
| 2 Clinical efficacy (cure) at the end of therapy Show forest plot | 1 | 89 | Risk Ratio (M‐H, Random, 95% CI) | 0.96 [0.74, 1.26] |
| 3 Adverse effects of treatment Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 3.1 Any adverse effects | 1 | 89 | Risk Ratio (M‐H, Random, 95% CI) | 0.19 [0.01, 3.79] |
| 3.2 Hepatic adverse effects (increased transaminases) | 1 | 89 | Risk Ratio (M‐H, Random, 95% CI) | 0.19 [0.01, 3.79] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy (pathogen eradication) at the end of treatment Show forest plot | 1 | 30 | Risk Ratio (M‐H, Fixed, 95% CI) | 3.0 [1.01, 8.95] |
| 2 Adverse effects of treatment Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 2.1 Any adverse effects | 1 | 33 | Risk Ratio (M‐H, Random, 95% CI) | 0.53 [0.05, 5.31] |
| 2.2 Gastrointestinal/hepatic adverse effects | 1 | 33 | Risk Ratio (M‐H, Random, 95% CI) | 0.21 [0.01, 4.10] |
| 2.3 Drop in leukocyte counts | 1 | 33 | Risk Ratio (M‐H, Random, 95% CI) | 3.18 [0.14, 72.75] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy (pathogen eradication) at the end of treatment Show forest plot | 1 | 66 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.04 [0.90, 1.19] |
| 2 Clinical efficacy ‐ NIH‐CPSI score at the end of treatment Show forest plot | 1 | Std. Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 2.1 NIH‐CPSI pain score | 1 | 66 | Std. Mean Difference (IV, Random, 95% CI) | ‐0.13 [‐0.62, 0.35] |
| 2.2 NIH‐CPSI voiding symptom score | 1 | 66 | Std. Mean Difference (IV, Random, 95% CI) | ‐0.30 [‐0.78, 0.19] |
| 2.3 NIH‐CPSI quality of life impact score | 1 | 66 | Std. Mean Difference (IV, Random, 95% CI) | ‐0.24 [‐0.72, 0.25] |
| 3 Clinical efficacy ‐ number of participants with leukocytosis in post‐massage urine specimens at the end of treatment Show forest plot | 1 | 66 | Risk Ratio (M‐H, Random, 95% CI) | 0.54 [0.17, 1.66] |
| 4 Clinical efficacy ‐ urine peak flow rate at the end of treatment (mL/s) Show forest plot | 1 | 66 | Std. Mean Difference (IV, Random, 95% CI) | 0.24 [‐0.25, 0.72] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy (pathogen eradication) at the end of treatment Show forest plot | 1 | 69 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.01 [0.88, 1.17] |
| 2 Clinical efficacy ‐ NIH‐CPSI score at the end of treatment Show forest plot | 1 | Std. Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 2.1 NIH‐CPSI pain score | 1 | 69 | Std. Mean Difference (IV, Random, 95% CI) | ‐0.06 [‐0.53, 0.42] |
| 2.2 NIH‐CPSI voiding symptom score | 1 | 69 | Std. Mean Difference (IV, Random, 95% CI) | 0.27 [‐0.20, 0.75] |
| 2.3 NIH‐CPSI quality of life impact score | 1 | 69 | Std. Mean Difference (IV, Random, 95% CI) | 0.52 [0.04, 1.01] |
| 3 Clinical efficacy ‐ number of participants with leukocytosis in post‐massage urine specimens at the end of treatment Show forest plot | 1 | 69 | Risk Ratio (M‐H, Random, 95% CI) | 0.74 [0.28, 1.98] |
| 4 Clinical efficacy ‐ urine peak flow rate at the end of treatment (mL/s) Show forest plot | 1 | 69 | Std. Mean Difference (IV, Random, 95% CI) | 0.10 [‐0.37, 0.57] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Microbiological efficacy (pathogen eradication) at the end of treatment Show forest plot | 1 | 71 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.02 [0.90, 1.16] |
| 2 Clinical efficacy ‐ NIH‐CPSI score at the end of treatment Show forest plot | 1 | Std. Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 2.1 NIH‐CPSI pain score | 1 | 71 | Std. Mean Difference (IV, Random, 95% CI) | ‐0.09 [‐0.55, 0.38] |
| 2.2 NIH‐CPSI voiding symptom score | 1 | 71 | Std. Mean Difference (IV, Random, 95% CI) | ‐0.64 [‐1.11, ‐0.16] |
| 2.3 NIH‐CPSI quality of life impact score | 1 | 71 | Std. Mean Difference (IV, Random, 95% CI) | ‐0.69 [‐1.17, ‐0.21] |
| 3 Clinical efficacy ‐ number of participants with leukocytosis in post‐massage urine specimens at the end of treatment Show forest plot | 1 | 71 | Risk Ratio (M‐H, Random, 95% CI) | 0.73 [0.22, 2.35] |
| 4 Clinical efficacy ‐ urine peak flow rate at the end of treatment (mL/s) Show forest plot | 1 | 71 | Std. Mean Difference (IV, Random, 95% CI) | 0.15 [‐0.31, 0.62] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Clinical efficacy ‐ NIH‐CPSI total score Show forest plot | 1 | Std. Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 1.1 NIH‐CPSI total score at the end of treatment | 1 | 143 | Std. Mean Difference (IV, Random, 95% CI) | ‐2.56 [‐3.04, ‐2.08] |
| 1.2 NIH‐CPSI total score at follow‐up (6 months) | 1 | 143 | Std. Mean Difference (IV, Random, 95% CI) | ‐3.78 [‐4.36, ‐3.20] |
| 2 Clinical efficacy ‐ IPSS score Show forest plot | 1 | Std. Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 2.1 IPSS score at the end of treatment | 1 | 143 | Std. Mean Difference (IV, Random, 95% CI) | ‐2.21 [‐2.66, ‐1.75] |
| 2.2 IPSS score at follow‐up | 1 | 143 | Std. Mean Difference (IV, Random, 95% CI) | ‐2.50 [‐2.98, ‐2.03] |
| 3 Adverse effects of treatment Show forest plot | 1 | 143 | Risk Ratio (M‐H, Random, 95% CI) | 1.05 [0.11, 9.76] |
