INTRODUCTION
Lower urinary tract dysfunction (LUTD), which encompasses abnormalities of bladder storage and voiding, affects approximately 72%–76% of adults worldwide and has a substantial effect on quality of life and socioeconomic burden. In men, voiding symptoms such as a weak urinary stream, a sensation of incomplete emptying, straining, and intermittency are primarily associated with 2 pathophysiological mechanisms: bladder outlet obstruction (BOO) and/or detrusor underactivity (DU). Because these conditions require distinct therapeutic approaches, urodynamic studies (UDS) remain the gold standard for differentiating between them [
1-
6].
Obstruction indices, such as the bladder outlet obstruction index (BOOI) and urethral resistance algorithm (URA), and nomograms, including the Abrams-Griffiths (AG) and linear passive urethral resistance relation nomograms, simplify the interpretation of UDS. However, these tools rely on discrete measurements and do not account for the temporal dynamics of detrusor pressure or interindividual variability, which may limit their accuracy in complex cases, such as combined BOO and DU [
7].
Mathematical models of the lower urinary tract, such as the Perugia urodynamic method of analysis (PUMA) and the Valentini-Besson-Nelson (VBN) model, aim to quantify anatomical and physiological relationships. However, their clinical applicability is constrained by the need for high-resolution data acquisition, multidisciplinary expertise, and external validation [
8,
9]. In contrast, time-series analysis (TSA), which has been successfully applied in cardiology, such as for heart rate variability, and in neurology, offers a framework for evaluating physiological fluctuations in real time and identifying hidden patterns within urodynamic curves. For example, variability in detrusor pressure may reflect modulation by the autonomic nervous system (ANS), whose dysregulation has been associated with LUTD [
10-
12].
Recent developments, including wavelet-based algorithms for real-time estimation of detrusor pressure, suggest that signal-processing techniques can extract dynamic biomarkers from uroflowmetry curves, with reported diagnostic accuracies of up to 84% compared with 56% for standard urological diagnosis alone [
12].
Physiological control of the lower urinary tract depends on a hierarchical network that integrates the central nervous system, ANS, and peripheral nervous system. Dysfunction at any level, whether neurological or non-neurological, can result in a heterogeneous spectrum of urinary symptoms [
13]. Although approximately 40% of men with LUTD present with mixed profiles of BOO and DU, current diagnostic models often fail to capture this complexity, leading to the misclassification of some patients within “equivocal” zones [
7]. This diagnostic uncertainty may contribute to suboptimal treatment decisions, particularly in settings without access to videourodynamics or urethral pressure profilometry, which are costly and not widely available [
14].
Because conventional urodynamic parameters are based on discrete values and do not capture the dynamic variability of detrusor pressure during voiding, TSA may provide additional information for evaluating BOO.
Accordingly, the primary objective of this study was to assess whether detrusor pressure derived from TSA of pressure-flow studies can complement conventional urodynamic parameters and established nomograms in discriminating between men with and without BOO.
MATERIALS AND METHODS
This study was conducted after approval by the local research ethics committee. It was designed as an observational, analytical, cross-sectional, prospective study to evaluate the diagnostic utility of TSA of detrusor pressure in men with LUTD.
The study population consisted of male patients aged ≥18 years who were evaluated at the urodynamic service of our hospital between July 2021 and July 2022. All patients had a clinical diagnosis of voiding dysfunction (VD) and an indication for urodynamic evaluation.
Inclusion criteria comprised clinically significant lower urinary tract symptoms (LUTS), defined by at least 1 of the following: an International Prostate Symptom Score >8, prostate volume >30 g, postvoid residual urine volume >50 mL, or a maximum urinary flow rate (Qmax) <10 mL/sec on free uroflowmetry. Patients with active urinary tract infection, bladder lithiasis, bladder tumors, or a urodynamic diagnosis of overactive bladder were excluded. Patients without a demonstrable detrusor contraction during voiding or whose studies did not meet established technical quality standards for interpretation were also excluded [
15].
UDS were performed by a urologist specializing in functional urology and urodynamics, in accordance with International Continence Society (ICS) good urodynamic practice guidelines [
16], using the Ellipse Andromeda urodynamic system (ANDROMEDA medizinische Systeme GmbH, Germany).
Patient Classification
After urodynamic evaluation, patients were classified according to BOOI and bladder contractility index (BCI) definitions into the following groups: BOO, DU, combined BOO and DU, and no urodynamic abnormalities [
17].
In addition, a final clinical diagnosis was established by a urologist specializing in functional urology and urodynamics on the basis of the overall clinical and urodynamic assessment, and this diagnosis was used to guide therapeutic decision-making. Patients classified within the equivocal BOOI zone were also specifically identified.
Detrusor Pressure TSA
Detrusor pressure was analyzed as a time series during the pressure-flow study. Values were exported at a per-second resolution from the urodynamic acquisition system. The analysis was restricted to the voiding phase, defined as
the time interval during which urinary flow (Q) was present. Periods without measurable flow were excluded to ensure that the analysis reflected active detrusor contraction during micturition. To avoid inclusion of spurious or clinically irrelevant events, a minimum duration/volume threshold for urinary flow was applied, excluding isolated or transient flow signals that did not represent true voiding.
Within the selected segments, descriptive statistical metrics were calculated for detrusor pressure, including: mean detrusor pressure during voiding (mPeDet) and standard deviation of detrusor pressure during voiding (SDPeDet).
Additionally, mean urinary flow (mQura) was calculated over the same time intervals.
This approach enables characterization of detrusor contractile activity during voiding using simple, reproducible time-domain measures derived from routinely collected urodynamic data.
Statistical Analysis
Continuous variables were expressed as median and interquartile range, according to data distribution. Categorical variables were expressed as frequencies and percentages.
A linear regression analysis was performed to evaluate the association between mPeDet, SDPeDet, mQura, and conventional urodynamic indices such as the BOOI, to evaluate which of them presented the strongest correlation.
Receiver operating characteristic (ROC) curve analysis was then performed to assess the diagnostic accuracy of mPeDet for classifying BOO. The gold standard was the final clinical diagnosis of obstruction versus no obstruction, as established by an experienced urologist specializing in functional urology and urodynamics.
For each patient, obstructed or nonobstructed status was determined by 3 approaches: (1) urodynamic diagnosis based on BOOI, (2) final clinical diagnosis established by the treating urologist, and (3) classification derived from TSA. Agreement between the clinical classification and the mPeDet-based classification was evaluated using Cohen kappa coefficient, with values ≥0.70 considered indicative of acceptable agreement (P<0.05).
Exploratory and Stratified Analysis
In addition to the primary analysis, an exploratory stratified analysis was performed to assess the potential influence of neurological and rheumatological comorbidities on the diagnostic performance of the proposed urodynamic parameter. Patients were classified into 2 groups: those without neurological or rheumatological comorbidities and those with at least one of these conditions.
Between-group comparisons were performed using the nonparametric Mann-Whitney U-test because continuous variables were not normally distributed.
The area under the ROC curve (AUC) was calculated for the overall cohort and after stratification by the presence or absence of neurological or rheumatological comorbidities. In addition, an exploratory logistic regression model was constructed to evaluate the association between the presence of comorbidities and a diagnosis of BOO, with odds ratios (ORs) and 95% confidence intervals (CIs) estimated.
To further assess the robustness of the proposed parameter, a sensitivity analysis was performed excluding patients with a history of transurethral resections of the prostate (TURP). Diagnostic performance was re-evaluated in this subset using ROC curve analysis and corresponding AUC estimates.
All statistical analyses were performed using IBM SPSS Statistics ver. 24.0 (IBM Co., USA), and a P-value <0.05 was considered statistically significant.
RESULTS
A total of 60 male patients referred to our unit with a clinical diagnosis of VD were initially included in the study. Eighteen patients were excluded because they were unable to generate urinary flow during the pressure-flow phase of the urodynamic study, and 1 additional patient was excluded because of a urodynamic diagnosis of detrusor overactivity.
Among the remaining 41 patients, the median age was 59 years. A history of prior urological surgery was present in 37% of patients (n=15), including 1 cystoscopy, 1 ureterolithotripsy, 7 TURP, 1 urethrotomy, and 5 kidney transplants. Neurological comorbidities were identified in 20% of patients (n=8), and rheumatological comorbidities were present in 7% of patients (n=3) (
Table 1).
Based on the urodynamic diagnosis established using BCI and BOOI, patients were classified as follows: 29% (n=12) were diagnosed with BOO; 36% (n=15) with DU; 15% (n=6) fulfilled criteria for both BOO and DU; and 20% (n=8) did not meet urodynamic criteria for classification into any voiding dysfunction category using these indices (
Table 2).
According to BOOI values, 37% of patients (n=13) were classified within the equivocal range for BOO.
Based on the final clinical diagnosis, 49% of patients (n=20) were diagnosed with BOO, 29% (n=12) with DU, 10% (n=4) with combined BOO and DU, and 12% (n=5) with no VD (
Table 2).
Among the metrics obtained from TSA, mPeDet demonstrated the strongest linear correlation with BOOI, with a correlation coefficient of r=0.89 (P<0.001). On the basis of this association, mPeDet was selected as the primary parameter for further evaluation of diagnostic performance and was compared with the final urological diagnosis, which was considered the gold standard. ROC curve analysis demonstrated excellent diagnostic performance, with an AUC of 0.953, indicating high discriminative ability (P<0.001) (
Fig. 1).
An mPeDet threshold of 37 cm H₂O provided the optimal balance between sensitivity and specificity for diagnosing BOO, with a sensitivity of 88% and a specificity of 87%. Agreement between the obstruction classification based on this mPeDet cutoff and the final urological diagnosis was substantial, with a weighted Cohen kappa of 0.69 (P<0.001).
When the mPeDet cutoff of 37 cm H₂O was compared with obstruction classification according to BOOI criteria, agreement was moderate, with a Cohen kappa coefficient of 0.61 (P<0.001).
To further illustrate the clinical utility of mPeDet, a transition analysis was performed comparing BOOI-based categories with mPeDet-based classification. Among patients with BOOI<20, 9 of 10 were classified as nonobstructed and 1 as obstructed by mPeDet. Among those in the equivocal BOOI range (20–40), mPeDet reclassified 6 of 13 as nonobstructed and 7 of 13 as obstructed. Among patients with BOOI>40, 17 of 18 were classified as obstructed and 1 as nonobstructed by mPeDet (
Fig. 2).
Stratified Analysis
Among the 41 included patients, 30 had no neurological or rheumatological comorbidities, whereas 11 had at least 1 such condition. mPeDet did not differ significantly between patients with and without neurological or rheumatological comorbidities (P=0.441).
In the stratified analysis, the AUC was 0.98 (95% CI, 0.95– 1.00; P=0.001) in patients without neurological or rheumatological comorbidities, whereas in patients with these comorbidities the AUC decreased to 0.66 (95% CI, 0.359–0.975; P=0.48).
In contrast, in patients with these comorbidities, the AUC decreased to 0.66 (95% CI, 0.359–0.975; P=0.48). However, this estimate showed a wide confidence interval and was not statistically significant, reflecting substantial uncertainty due to the small sample size (n=11).
In univariate logistic regression analysis, the absence of neurological or rheumatological comorbidities was not independently associated with a diagnosis of BOO (OR, 0.25; 95% CI, 0.05–1.38; P=0.112).
DISCUSSION
Several formulas and nomograms, including the AG number [
18], URA, and passive urethral resistance relation, have been developed to diagnose and classify bladder outlet resistance [
19]. However, all of these approaches rely on single-point measurements obtained at specific moments during voiding. In 1997, the ICS published a standardized nomogram based on a modification of the AG nomogram, establishing cutoff values for BOO: <20 nonobstructed, 20–40 equivocal, and >40 obstructed [
20].
The present study evaluated detrusor pressure as a dynamic process by analyzing the entire voiding phase on a second-bysecond basis. This approach enabled derivation of a novel parameter, mPeDet, which showed a strong association with BOOI.
The optimal diagnostic cutoff value for mPeDet identified in this study was 37 cm H₂O. This analysis was intended to address one of the main limitations of BOOI: the presence of an equivocal diagnostic range. When mPeDet was compared with the final clinical diagnosis established by an experienced func-tional urologist, the weighted Cohen kappa was 0.69, indicating moderate (κ=0.69).
An important consideration is the incremental clinical value of mPeDet in relation to established indices such as BOOI. While mPeDet demonstrates strong correlation with BOOI and high diagnostic performance, its primary contribution may lie in its ability to refine classification in cases where conventional indices are inconclusive.
In particular, patients within the equivocal BOOI range (20– 40) represent a well-recognized diagnostic challenge. In our cohort, mPeDet reclassified patients in this group into both obstructed and nonobstructed categories, thereby reducing diagnostic uncertainty. This reclassification is illustrated in
Fig. 2 and highlights the potential role of mPeDet as a complementary parameter. Importantly, mPeDet is not intended to replace existing indices but to provide additional, integrative information derived from the entire voiding phase. This suggests that its clinical utility lies in supporting decision-making in borderline cases, rather than improving diagnostic performance in clearly defined groups.
The 37 cmH₂O cutoff identified in this study should be interpreted cautiously. Because it was derived from ROC analysis in a relatively small single-cohort sample, it is likely influenced by the demographic and clinical characteristics of the study population, including the distribution of BOO, comorbidity status, and prior surgical history. Therefore, this threshold should be considered preliminary and hypothesis-generating rather than definitive. Future studies should validate this cutoff in larger, independent, and more heterogeneous cohorts before broader clinical application is considered.
In this study, the term “time-series” refers to the analysis of continuously recorded detrusor pressure as a function of time during the voiding phase. Rather than relying on single-point measurements, this approach considers the temporal structure of the signal to derive clinically meaningful descriptors. Importantly, the proposed method does not involve complex modeling or advanced signal-processing techniques, but instead focuses on simple, reproducible time-domain measures derived from routinely acquired urodynamic data. Time-series data are commonly generated when physiological processes are monitored over time and are widely used in biomedical and public health research to examine how variables evolve dynamically and influence outcomes [
21].
In urology, however, relatively few studies have explored time-series approaches, and their clinical application remains limited. Mehnert et al. [
11] investigated ANS activity during filling cystometry in healthy women and identified a characteristic pattern of autonomic modulation, with a gradual increase in sympathetic activity preceding the report of a strong desire to void.
Analyses of urodynamic signals have demonstrated that detrusor activity may exhibit complex temporal behavior, including variability and nonlinear patterns, which can be associated with underlying functional states [
22]. Similarly, pattern-based approaches, such as those evaluating detrusor contraction morphology or flow curve characteristics, have shown potential in improving the identification of voiding dysfunction [
23]. More recently, advances in automated analysis, including artificial intelligence– based models, have further highlighted the value of dynamic signal interpretation in urology [
12]. These approaches have demonstrated improved diagnostic accuracy in selected contexts, particularly when compared to conventional clinicianbased interpretation.
In this context, the present study contributes a pragmatic and clinically interpretable alternative by focusing on time-domain analysis of detrusor pressure during voiding. Unlike more complex computational models, our approach relies on simple statistical descriptors derived from the entire voiding phase, facilitating reproducibility and potential integration into routine clinical practice. This approach aligns with broader trends in physiological signal analysis, where dynamic variables are increasingly evaluated using continuous measurements rather than isolated time points. Importantly, the simplicity of this method enhances its reproducibility and facilitates its potential adoption in routine clinical practice.
From a practical standpoint, the proposed method is relatively simple to implement. The current workflow involves exporting urodynamic time-series data from standard acquisition systems (e.g., Andromeda), followed by offline analysis restricted to the voiding phase, during which simple statistical measures such as mPeDet are calculated. The computational burden of the analysis is minimal, and processing can be completed within seconds using standard computing resources, supporting its feasibility in routine clinical settings without requiring specialized hardware.
Although the present implementation is performed offline, the simplicity of the analytical steps makes it suitable for integration into existing urodynamic software platforms. This would enable real-time or near–real-time calculation of mPeDet during pressure-flow studies, potentially enhancing clinical decision-making at the point of care.
Conventional urodynamic parameters used to assess BOO, such as pdetQmax and derived nomograms, rely on singlepoint measurements obtained at specific moments during voiding. This approach is vulnerable to intravoid variability, artifacts, and dependence on a representative void. In contrast, the mPeDet, derived from TSA, integrates information across the entire voiding phase, providing a more global and stable estimate of detrusor effort. This may explain its stronger association with clinical obstruction observed in this study. Importantly, mPeDet is not intended to replace existing parameters but to complement them by better reflecting the dynamic nature of micturition.
External validation remains essential to ensure generalizability. Future prospective and longitudinal studies should evaluate diagnostic performance across diverse urological settings, assess reproducibility, and explore correlations with clinically relevant outcomes, including response to surgical or pharmacological treatments.
Study limitations include the relatively small sample size and the inclusion of patients with neurological and rheumatological comorbidities, which may introduce heterogeneity in detrusor function. The reduced diagnostic performance observed in this subgroup (AUC=0.66) should be interpreted with caution. The wide CI (0.359–0.975) and lack of statistical significance indicate substantial uncertainty, primarily driven by the limited sample size (n=11). Accordingly, this subgroup analysis should be considered exploratory and not indicative of definitive differences in diagnostic performance.
Overall, the observed findings likely reflect a combination of statistical variability and potential physiological heterogeneity. Further studies with larger, well-characterized cohorts are required to clarify the performance of this parameter in patients with neurological and systemic comorbidities.
An important source of heterogeneity in our cohort is the inclusion of patients with prior prostate surgery, particularly TURP. Surgical modification of the bladder outlet may significantly alter voiding dynamics by reducing outlet resistance and inducing adaptive changes in detrusor contractility. These alterations can affect pressure-flow relationships, potentially modifying both the magnitude and temporal structure of detrusor pressure signals. Despite these considerations, our sensitivity analysis excluding patients with prior TURP demonstrated that the diagnostic performance of mPeDet remained robust (AUC, 0.93; 95% CI, 0.85–1.00; P=0.001), suggesting that the proposed parameter is not primarily driven by surgically altered outlet conditions.
Nevertheless, the inclusion of these patients reflects realworld clinical practice, where heterogeneous populations are frequently encountered. Future studies should further investigate the impact of prior surgical interventions, considering this variable either as an exclusion criterion in mechanistic studies or as a stratification factor in clinical validation cohorts.
An additional limitation of this study is the use of final clinical judgment as the reference standard for bladder outlet obstruction. This may lead to an overestimation of diagnostic performance, including measures such as AUC and agreement coefficients, since the index test and reference standard are not fully independent. However, this approach reflects real-world clinical practice, where diagnostic decisions are typically based on an integrated interpretation of urodynamic findings rather than a single numerical index.
Despite these limitations, the consistency of the observed associations and the complementary role of mPeDet in relation to established indices support the potential clinical relevance of the findings. Nevertheless, future studies should aim to validate these results using more independent or standardized reference criteria to minimize bias.
Overall, mPeDet represents a simple and clinically interpretable parameter that may complement conventional urodynamic indices, particularly in cases where diagnostic uncertainty persists.
In conclusion, in this study, the mean detrusor pressure obtained through time-series analysis during the pressure–flow phase showed a significant association with bladder outlet obstruction in men. This parameter demonstrated a strong correlation with the obstruction index. These findings suggest that analyzing detrusor pressure throughout the entire voiding phase allows a more global characterization of voiding behavior, providing complementary information to standard urodynamic assessment.