Diagnostic value of urodynamic bladder outlet obstruction to select patients for transurethral surgery of the prostate: Systematic review and meta-analysis

Purpose To investigate the diagnostic value of urodynamic bladder outlet obstruction (BOO) in the selection of patients for transurethral surgery of the prostate. Materials and methods We systematically searched online PubMed, Embase, and Cochrane Library databases from January 1989 to June 2014. Results A total of 19 articles met the eligibility criteria for this systematic review. The eligible studies included a total of 2321 patients with a median number of 92 patients per study (range: 12–437). Of the 19 studies, 15 conducted conventional transurethral prostatectomy (TURP), and 7 used other or multiple modalities. In urodynamic bladder outlet obstruction (BOO) positive patients, the pooled mean difference (MD) was significant for better improvement of the International Prostate Symptom Score (IPSS) (pooled MD, 3.48; 95% confidence interval [CI], 1.72–5.24; p < 0.01; studies, 16; participants, 1726), quality of life score (QoL) (pooled MD, 0.56; 95% CI, 0.14–1.02; p = 0.010; studies, 9; participants, 1052), maximal flow rate (Qmax) (pooled MD, 3.86; 95% CI, 2.17–5.54; p < 0.01; studies, 17; participants, 1852), and post-void residual volume (PVR) (pooled MD, 32.46; 95% CI, 23.34–41.58; p < 0.01; studies, 10; participants, 1219) compared with that in non-BOO patients. Some comparisons showed between-study heterogeneity despite the strict selection criteria of the included studies. However, there was no clear evidence of publication bias in this meta-analysis. Conclusions Our meta-analysis results showed a significant association between urodynamic BOO and better improvements in all treatment outcome parameters. Preoperative UDS may add insight into postoperative outcomes after surgical treatment of benign prostatic hyperplasia.

Introduction Traditionally, the primary goal of treatment of benign prostatic hyperplasia (BPH) has been to lessen the bothersome lower urinary tract symptoms (LUTS) caused by prostatic enlargement [1,2]. Surgery is the most invasive option for BPH treatment which can cause irreversible complications [3]. To ensure a better outcome, proper indicators for surgical intervention should be selected. The most recent international treatment guidelines commonly recommend that a surgical intervention should be considered in BPH patients with failure to treatment with oral medications or with complicated LUTS [1,2].
The mechanism for surgery is based on the classic bladder outlet obstruction (BOO) model. Enlarged prostate tissue causes obstruction and increases the urethral resistance to flow, and therefore requires higher intravesical pressure to void [2]. The urodynamic study (UDS) is the only gold standard for the diagnosis of BOO [4]; however, invasiveness, cost, and morbidity of UDS limit its clinical use [5]. In this regard, most guidelines recommend UDS for male LUTS evaluation only in specific situations such as, prior to surgery, previous unsuccessful treatment, functional cystometric capacity < 150mL, post-void residual urine (PVR) > 300mL, patient too young (< 50 years) or too old (> 80 years) for surgery [1], or maximal flow rate (Q max ) > 10mL/s (relative BOO) [2]. However, most of those recommendations are supported by very low level of evidences (LEs) (all LE = 3) [1]. To our knowledge, there have been no randomized studies regarding the usefulness of UDS for guiding clinical application in male LUTS. There are no published randomized controlled trials in men with LUTS that compare the standard investigations such as symptom score or uroflowmetry (UFM) with UDS [1]. Moreover, the utility of performing UDS before transurethral surgery has rarely been studied in a systemic fashion.
Because reports on the diagnostic value of urodynamic BOO for LUTS in men are few, a combination of these data to reach a reasonable conclusion is necessary. The objective of the present study was to conduct a systematic review and meta-analysis of published literature investigating the diagnostic value of urodynamic BOO in the selection of patients for transurethral surgery of the prostate and to provide a higher LE for guiding practical use of UDS in BPH patients.

Search strategy for relevant studies
The whole process for this systematic review and meta-analysis was conducted according to the study protocol approved by all authors and followed the up-dated versions of MOOSE and PRISMA recommendations (S1 and S2 Checklists) [6,7]. We systematically searched online PubMed, Embase, and Cochrane Library database up to June 2014. Our overall search strategies included terms for UDS (urodynamic, cystometry, and pressure flow study), BPH (benign prostatic hyperplasia, benign prostatic obstruction, and male LUTS), and transurethral surgery (transurethral resection, transurethral incision, vaporization, ablation, and enucleation). Detailed queries for search strategy are presented in S1 Table. Manual search of relevant studies were also performed referring to review articles or original research articles on similar subjects.

Selection criteria of eligible studies for meta-analysis
The inclusion criteria for our systematic review were as follows: (1) original research articles published in English; (2) studies that included patients with BPH alone; (3) studies that included subjects who underwent transurethral surgery for BPH; (4) studies that included cases preoperatively sub-grouped by the urodynamic criteria of BOO; (5) studies in which outcome parameters were objectively described using standard investigation tools such as the International Prostate Symptom Score (IPSS) or UFM parameters; (6) studies that investigated the association between the urodynamic BOO and improvement of treatment outcome; (7) studies with a definite sample size; and (8) studies that had standard deviation (SD), confidence interval (CI), or other distributional information of outcome parameters. When duplication of patient data was suspected, the most recently published or most informative single article was selected. If the population of the study underwent two or more surgical procedures [8,9], data were processed separately according to the type of surgery. Due to the unavailability of randomized studies, all non-randomized and retrospective studies were included in the systematic review and meta-analysis. Exclusion criteria were as follows: (1) the study could not satisfy the aforementioned inclusion criteria; (2) review articles or letters; (3) laboratory studies such as studies on ex-vivo and animal models; and (4) studies with data that were insufficient to estimate the mean and SD of improved outcomes.
To minimize the bias, abstract screening and full text assessment for eligibility were independently performed by all three authors (MK, CWJ, and SJO). All screened abstracts were classified into three categories: (1) not-eligible, (2) unclear, and (3) potentially-eligible. The full texts of "potentially-eligible" and "unclear" studies were obtained and assessed for eligibility. All disagreements among three reviewers were resolved by a consensus meeting.

Data extraction and quality assessments
The extracted data elements were as follows: (1) overall characteristics of eligible studies, which included the name of first author, publication year, country, recruitment period, study design, population size, type of surgical intervention, urodynamic standards and cut-off value to diagnosis BOO, and quality score for each study; (2) patient characteristics, which included analyzedpopulation size, mean or median age, time of treatment outcome evaluation, and compared-outcome parameters; and (3) mean improvement of IPSS (ΔIPSS), IPSS-quality of life score (QoL) (ΔIPSS-QoL), Q max (ΔQ max ), and PVR (ΔPVR) of each subgroup with their SD, after the surgical interventions. Study quality was assessed independently by all three authors using the MINORS (score range: 0-24) [10]. Any disagreement was resolved by discussion.

Statistical analysis
Primary analysis. The improvements of the main outcome parameters (ΔIPSS, ΔIPSS-QoL, ΔQ max , and ΔPVR) were compared between the preoperative urodynamic BOO positive and negative subgroups. Owing to the continuous parametric feature of outcomes, the pooled mean difference (MD) was used to summarize each outcome parameters across the eligible studies. A random-effect model was used to obtain the pooled MDs and their 95% CIs, because we hypothesized that the selected eligible studies contains inter-study heterogeneities (e.g., surgical modalities, period of postoperative outcome evaluation, and urodynamic cut-off values) as well as within-study heterogeneities (e.g., sampling variation). Mean value and SD of outcome parameters were needed for data integration. For each study, these values were estimated based on the data provided in the publications using previously suggested techniques [11,12]. An observed pooled MD > 0 indicated better improvement of BOO positive group compared to BOO negative, and would be considered statistically significant if the 95% CI did not overlap the pooled MD value of zero, with p < 0.05.

Subgroup (sensitivity) analysis.
Subsequently, we assessed subgroup analysis in patients who underwent conventional transurethral resection of the prostate (TURP) to evaluate the effect of type of surgery performed. Furthermore, subgroup analyses were also performed according to the two dominant criteria for BOO diagnosis (BOO index [BOOI] > 40 [13], and linear passive urethral resistance relation [lin PURR] grade ! 2, 3, or 4 [14]) to evaluate the effects of diagnostic criteria.
Assessment of heterogeneity. Heterogeneity was assessed using the heterogeneity х 2 test (Cochran's Q-test), a p value of > 0.05 indicated the absence of significant heterogeneity [15]. The I 2 statistic (Higgin H-test) was performed to visualize degree of heterogeneity [16].
Publication bias. Possibilities of publication bias were assessed using Funnel plots (Harbord test) [17].
Utilized tools. Review Manager (RevMan) version 5.3 (The Nordic Cochrane Center, The Cochrane Collaboration, Copenhagen, Denmark) was utilized for the meta-analysis.

Outcome of the selection process
A flow chart of the whole study selection process is shown in Fig 1. Our search strategy identified 3875 articles (PubMed, 1445 articles; Embase, 2137 articles; Cochrane Library database, 293 articles). Additionally, 133 articles were found by manual searching. After duplicates were removed, 2611 abstracts were independently screened by three authors. After abstract screening, 223 articles were entered to full text assessment for eligibility. After careful review of full articles, 204 were excluded for the following reasons: 23 studies were written in languages other than English, seven were review articles, one was a letter to editor, the full text of the manuscript could not obtained in 15 articles, 63 were out of scope, 34 covered related subjects but could not satisfy the inclusion criteria regarding methodology, 52 lacked eligible data, and nine studies were excluded due to the duplication of the study population. Eventually 19 studies were selected as eligible for the data synthesis [8,9,. Tables 1 and 2 show the general characteristics of eligible studies. The 19 eligible studies included a total of 2321 patients, with a median number of 92 patients per study (range: 12-487). None of the selected studies was a randomized prospective study. Seven of the 19 included studies were non-randomized prospective studies, and the remaining studies had a retrospective design. Patients received TURP in 15 of 19 studies [18,19,24-36], transurethral microwave thermotherapy (TUMT) in one study [20], interstitial laser coagulation (ILC) in one study [23], and multiple intervention modalities in two studies [8,9]. The definition of urodynamic BOO varied among studies. The urodynamic BOO was defined as BOOI > 40 cmH 2 [32], and other definitions in four studies [9,18,19,37]. The median quality score measured by MINORS recorded as 16 (range: 14-18). There was no significant correlation between population size and quality scores (p = 0.231, by Spearman's correlation analysis) ( Table 1). Median or mean ages of study populations were as shown in Table 2. Time of treatment outcome evaluation varied (range: 1.5-144 months). Mean and SD values of ΔIPSS, ΔIPSS-QoL, ΔQmax, and ΔPVR could not be obtained in all studies ( Table 2). If the mean and SD values could not be obtained, these values were estimated using other representative and distributional values. Details of the process applied are shown in the S2 Table. the respective 95% CI of ΔIPSS (Fig 2A), ΔIPSS-QoL (Fig 2B), ΔQ max (Fig 2C), and ΔPVR ( Fig  2D). In the comparisons of BOO positive and negative patients, all pooled MDs were significantly greater than zero: ΔIPSS (pooled MD, 3.48; 95% CI, 1.72-5.24; p < 0.01; studies, 16; participants, 1726; Fig 2A), ΔIPSS-QoL (pooled MD, 0.56; 95% CI, 0.14-1.02; p = 0.01; studies, 9; participants, 1052; Fig 2B) studies, 10; participants, 1219; Fig 2D). This means that BOO positive patients have better surgical outcomes in all parameters compared to BOO negative patients.

Subgroup analysis
Subsequently, the subgroup analyses using patients who underwent TURP were performed (

Assessment of heterogeneity and publication bias
Despite our attempt to limit between-study heterogeneity through strict inclusion and exclusion criteria, heterogeneity between treatment outcomes still remained (heterogeneity х 2 test: p < 0.05 in ΔIPSS, ΔIPSS-QoL, and ΔQ max comparisons, I 2 range: 12-88%; Fig 2). However, there was no clear evidence of funnel plot asymmetry for outcomes (Fig 3). Therefore, it can be concluded that there was no clear evidence of publication bias.

Urodynamic BOO and surgical outcome
There have been insufficient evidences on which to base clear statements about 'the right treatment', despite a large number of studies published over several decades [40]. The preoperative evaluation of BPH patients provided clear examples of practice based on poor evidence. Recent international guidelines recommend surgical intervention and the decision to perform surgery primarily relies on the physician who should decide the best initial treatment on a case-by-case basis according to clinical conditions [1,2]. Poor correlation between the degree of urodynamic BOO and degree of patient symptoms is suggested by some researchers [41,42]. Those findings have indirectly supported the fact that UDS is not particularly useful as a preoperative evaluation in transurethral surgery for BPH. However, if the basic principles of transurethral surgery are based on the classic BOO model, it can be expected that the degree of obstruction affects the treatment outcome. However, there has been not much evidence supporting this hypothesis [1]. The results of the current study provide a higher LE, which confirms the utility of performing UDS before transurethral surgery for BPH treatment.
During our survey of literature, we encountered various definitions of the urodynamic findings, especially for "BOO". BOO is basically defined as methods of analyzing the pressure-flow  plots. One of the major aims of pressure-flow study is to provide an objective diagnosis on weather the urethral resistance to flow is abnormally elevated [43]. For that purpose, methods have been developed to quantify pressure-flow plots in terms of one or more numerical parameters [13,14,[44][45][46][47][48][49]. However, the optimal method for BOO diagnosis has not been confirmed [43]. Nevertheless, it seems that there is some degree of reliability on inter-test agreement due to the underlying similarity for diagnosing BOO [50]. Moreover, in our current study, subgroup analyses using the two dominant definition criteria demonstrated consistency except for the ΔIPSS-QoL of BOOI subgroup (Table 4).

Limitations and strengths of the current study
Our study presents some limitations. First, none of the studies included in the current metaanalysis specified a prospective design. To the best of our knowledge, there has been no randomized study regarding the usefulness of UDS for guiding clinical management in male LUTS [1]. This may be due to the complexity of designing prospective study. It is difficult to draw any definitive conclusions when studies are not conducted prospectively. Thus, the results of the present study are important because, they can provide a higher LE regarding the diagnostic value of preoperative UDS in male LUTS patients who are being considered for transurethral surgery. The results of our study also can be an important reference for designing of further prospective studies in the future. Second, due to the unavailability of mean and variance (or SD) data in some studies, those values are estimated using other presented distributional parameters for the outcome synthesis (see S2 Table). This can lead to some errors used in the estimation processes. However, the data imputation techniques used in this study were shown to have a low possibility of statistical errors in a previous study [51]. For the clarification of these points, accumulation of more evidence is needed. Third, in our current meta-analysis, there was some heterogeneity in the included studies (Fig 2; especially in ΔIPSS, ΔIPSS-QoL, and ΔQ max comparisons). Heterogeneity can be caused by numerous factors such as inclusion criteria, type of surgery, sample size, period of postoperative outcome evaluation, urodynamic cut-off values, and adjustment for other co-factors. It is also very difficult to explain inter-study heterogeneity, due to the variability in clinical characteristics across patients within studies. To lessen the heterogeneity related bias, we adopted the random-effect model for data synthesis, which is known to be to draw more conservative results [12]. Moreover, the direct evidence due to publication bias was not shown (Fig 3).
Lastly, the BOO negative group also can experience symptom improvements from BPH surgery, although the degree of improvement in the BOO negative group is significantly less than that in the BOO positive group (Fig 2). Therefore, urodynamic BOO (BOOI > 40 or lin PURR grade ! 2, 3, or 4 in this study) might not be an absolute indication for surgical treatment in patients with PBH. This indicates that urodynamic BOO positive patients with BPH who are considering surgery would have better treatment outcomes than BOO negative Diagnostic value of BOO for transurethral surgery patients. However, being BOO positive (or negative) is not an absolute indication that the patient should (or should not) receive the surgery.
Despite some limitations, the findings from the present study suggest that preoperative urodynamic BOO has a diagnostic role in predicting treatment outcomes of surgery in male LUTS patients. The strengths of the current study are as follows: (1) broad, unbiased search of the literature; (2) strict criteria for study selection; and (3) application of standardized methods for systematic review [6,7]. Moreover, due to the relatively large number of eligible studies, subgroup (sensitivity) analysis could be performed according to the type of surgery (TURP) and the definition of BOO. Our subgroup analysis demonstrated consistency with the main results (Tables 3 and 4).

Conclusions
In conclusion, our meta-analysis results demonstrated significant association between preoperative BOO positive patients and better improvement of surgical outcome parameters including IPSS, IPSS-QoL, Q max , and PVR. On these grounds, preoperative urodynamic BOO may have a diagnostic role in predicting treatment outcomes after surgery in male LUTS patients. However, the diagnostic value of UDS for preoperative evaluation also needs to be confirmed in prospective controlled trials before any definitive conclusions can be made.