Signs and Symptoms of Detrusor Underactivity: An Analysis of Clinical Presentation and Urodynamic Tests From a Large Group of Patients Undergoing Pressure Flow Studies

Data from patients who underwent PFS, studied in a single specialist centre between 1985 and 2012, were recorded in a database that used the same variable fields throughout the 28-yr period.

Data gathering included patient interview to obtain symptoms and medical history, bladder diary data, physical examination, urodynamic studies, and diagnostic conclusions. PFS were carried out according to International Continence Society guidelines current at the time of testing. Free flow uroflowmetry was performed before each PFS. Postvoid residual urine volume was based on the volume obtained with catheterisation before filling commenced. The data from each PFS were screened for artefacts and manually entered into the database, thus avoiding automated data extraction errors. Prior to analysing the data, impossible values were removed in order to reduce corruption of data by manual entry errors. Several categorical (yes/no) variables used in the analysis were derived from a combination of database entry fields. For example, additional variables for straining and for decreased sensation were derived by combining the number of patients who reported these as symptoms with the number of patients for whom these were noted during PFS.

Patients without full voiding data, with neurological diseases affecting the lower urinary tract such as multiple sclerosis, paraplegia, or Parkinson's disease, and/or with a urodynamic diagnosis of detrusor overactivity were excluded as these require special consideration [5]. This resulted in 9928 eligible patient records (men: 1639; women: 8289) without confounding causes of vesico-urethral dysfunction (Fig. 1).

In order to classify patients with pure DU, BOO, or normal PFS, very strict criteria were used to avoid overlap. The criterion values were based on expert opinion and are shown in Table 1, which are in line with other studies cited by Osman et al [2]. A normal group was composed of patients with PFS judged to be normal, taking no medication related to bladder or urethra, and (for women) no clinical obstruction. Men who had both a low bladder contractility index and a high BOO index, suggesting simultaneous DU and BOO, were excluded from the analysis. Women patients with clinical obstruction, defined as urethral/bladder neck obstruction and/or large cystocoele or prolapse through the introitus, were also excluded from the DU and normal groups. Using these criteria, 1788 patient records (men: 507; women: 1281) were classified to DU, BOO, or normal PFS groups and used in the analysis (Fig. 1).

Age for men and women DU patients was statistically significantly higher compared with patients with normal PFS (median values: men 63 yr vs 55 yr; women 59 yr vs 44 yr). Age for women with DU was statistically significantly higher compared with women with BOO (median, 59 yr vs 44 yr). There were no significant differences in height, weight, or body mass index between the groups.

Table 2 and Table 3 show the odds ratios and corresponding 95% confidence intervals for categorical variables that showed a statistically significant difference between DU patients and at least one other group (p < 0.05). Table 4 gives a summary of medians and interquartile ranges for numerical variables that had statistically significant differences between DU patients and at least one other group. Supplementary Table 6 details all variables included in the analysis, so it may be deduced which variables were not statistically significantly different between groups by their absence from other tables.

The primary group comparisons of interest were DU versus BOO in men (the most difficult clinical differential diagnosis) and DU versus normal PFS in women. These comparisons are therefore described below and are all adjusted for age.

Refer to Supplementary Figure 2 (men) and Figure 3 (women) of forest plots and Table 5 showing symptoms with statistically significant differences for patients with DU compared with those with normal PFS or BOO.

This study shows that cross-sectional data can be used to associate signs and symptoms with urodynamically-defined DU.

Two earlier studies failed to detect clear differences in clinical presentation of DU patients compared with other patients, and concluded that few symptoms were helpful [6] and [7], whereas this study and another recent report in women by Rademakers et al [8] used strict criteria and a wide range of symptoms.

The present study corroborated Rademakers et al's [8] findings of an increased prevalence of incomplete bladder emptying, hesitancy, and a weak stream in women with DU compared with women with normal PFS. Some variables from the medical history that were significantly associated with DU may reflect some patients’ natural history. For example, previous surgery could affect pelvic innervation and decrease bladder function and sensation. Also, the association of DU with prior BOO surgery may be due to preoperatively present DU, contributing to the overall clinical picture. Any retrospective look at such patients would find some patients with persisting symptoms due to unmasked DU.

The strengths of this analysis are that the urodynamic technique and structure of the interview were similar throughout the period and that data covering a large number of patients were used. A limitation is that the database is not a reflection of the general urological patient population, nor of the normal population in general, since all patients were referred for specialist evaluation of functional urological problems and many had received prior diagnosis and/or intervention. This was a retrospective, post hoc analysis of an existing database. A nonvalidated, although constant, set of questions were used and many data points were derived from clinician recording of patient responses to questions, and the potential for bias in the reported rates of underlying symptoms within each group cannot be discounted. The percentage of patients showing certain derived variables (see Materials and methods section) may have differed if patients had been directly asked if they experienced the symptom. Additionally, there are inherent limitations caused by the testing of multiple groups and endpoints. Notwithstanding these limitations, the analysis shows that there appears to be differences in signs and symptoms between DU patients in comparison with BOO patients and patients with normal PFS, which can be used to create instruments used to evaluate the results of treatments for UAB/DU.

Whilst patients in the normal PFS group were required to have a bladder voiding efficiency (BVE = volume voided/[volume voided + post void residual volume] × 100%) of 100%, a cut-off of 90% was used to distinguish between DU and BOO. Some studies have used BVE <60% as indicative of DU. The results did not differ greatly when comparing the symptom patterns of DU patients using cut-offs of 60%, 70%, 80%, and 90%. We therefore feel that it is legitimate to use BVE <90% as the criterion for DU, and BVE ≥90% for BOO patients in order to clearly differentiate the groups. Although there would be obstructed patients with a BVE <90%, to have incorporated these patients in the analysis would potentially have masked true differences in symptom patterns due to patients with both DU and BOO.

The observations in this study suggest that further work to develop a specific symptom questionnaire to assess DU severity, possibly coupled with noninvasive tests, could be useful for diagnosis, assessment, and evaluation of treatment outcomes. The importance of the development of noninvasive methods to characterise DU has been highlighted by several authors [2], [4], [8], [9], and [10]. Further analyses of this database, including flow and voiding parameters that may be relevant to the definition of DU [3], are planned.

The present study has demonstrated the utility of database analysis to aid the development of a symptom-based definition of a traditionally invasively diagnosed urological condition. The identification and diagnosis of UAB patients has currently been hampered by a poor understanding of the clinical presentation of DU and the necessity of invasive PFS. The present study has shown that there are signs and symptoms that can distinguish men and women DU patients from patients with normal PFS, and further distinguish between DU and BOO. This analysis is a first step to better understand the clinical presentation of DU patients, is consistent with the recently published UAB working definition [1], and justifies developing and testing a diagnostic algorithm based on the signs and symptoms of DU.

Author contributions: Andrew Gammie had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Gammie, Kaper, Dorrepaal, Kos, Abrams.

Acquisition of data: Gammie.

Analysis and interpretation of data: Gammie, Kaper, Dorrepaal, Kos, Abrams.

Drafting of the manuscript: Gammie, Abrams.

Critical revision of the manuscript for important intellectual content Gammie, Kaper, Dorrepaal, Kos, Abrams.

Statistical analysis: Kaper.

Obtaining funding: None.

Administrative, technical, or material support: None.

Supervision: None.

Other: None.

Financial disclosures: Andrew Gammie certifies that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: Kaper, Dorrepaal, and Kos are employees of Astellas Pharma Europe.

Funding/Support and role of the sponsor: Gammie and Abrams were supported for this analysis by a grant from Astellas Pharma Europe.

Acknowledgments: Analysis work was funded by Astellas Pharma Europe, who also contributed to the design of the study, analysis of the data, and preparation and review of the manuscript.

The clinical diagnosis of detrusor underactivity (DU) is hampered by the need for invasive pressure flow studies (PFS) and a lack of knowledge of the associated signs and symptoms. This has contributed to a lack of awareness of DU and its clinical correlate, underactive bladder (UAB) [1]. In consequence, this condition has been neglected compared with other causes of lower urinary tract symptoms. A recent review [2] concluded that DU “is surrounded by ambiguity” and recognises the limitations of the current definition. The International Continence Society defines DU as “a contraction of reduced strength and/or duration, resulting in prolonged bladder emptying and/or a failure to achieve complete bladder emptying within a normal time span” [3]. This, however, does not define “prolonged bladder emptying” or “normal time span”. Various methods have been proposed to determine contraction strength [2]; however, none of these take into account the duration of contraction – a key factor in the definition [3].

Despite this imprecision, estimates suggest that DU is a prevalent condition, ranging from 9% to 23% in men <50 yr, increasing to as much as 48% in men >70 yr [2]. Elderly women show a DU prevalence ranging from 12% to 45% [2]. An analysis of the signs and symptoms associated with DU could potentially facilitate the diagnosis of patients with UAB, improve our knowledge of the epidemiology, indicate possible noninvasive diagnostic approaches, and facilitate the development and evaluation of treatment outcomes of new therapies for UAB [4].

The aim of this study was to investigate the signs and symptoms recorded in a large database of patients referred for urological evaluation who met strictly defined PFS criteria for DU, bladder outlet obstruction (BOO) or normal, in order to identify the clinical features associated with DU.

Data from patients who underwent PFS, studied in a single specialist centre between 1985 and 2012, were recorded in a database that used the same variable fields throughout the 28-yr period.

Data gathering included patient interview to obtain symptoms and medical history, bladder diary data, physical examination, urodynamic studies, and diagnostic conclusions. PFS were carried out according to International Continence Society guidelines current at the time of testing. Free flow uroflowmetry was performed before each PFS. Postvoid residual urine volume was based on the volume obtained with catheterisation before filling commenced. The data from each PFS were screened for artefacts and manually entered into the database, thus avoiding automated data extraction errors. Prior to analysing the data, impossible values were removed in order to reduce corruption of data by manual entry errors. Several categorical (yes/no) variables used in the analysis were derived from a combination of database entry fields. For example, additional variables for straining and for decreased sensation were derived by combining the number of patients who reported these as symptoms with the number of patients for whom these were noted during PFS.

Patients without full voiding data, with neurological diseases affecting the lower urinary tract such as multiple sclerosis, paraplegia, or Parkinson's disease, and/or with a urodynamic diagnosis of detrusor overactivity were excluded as these require special consideration [5]. This resulted in 9928 eligible patient records (men: 1639; women: 8289) without confounding causes of vesico-urethral dysfunction (Fig. 1).

In order to classify patients with pure DU, BOO, or normal PFS, very strict criteria were used to avoid overlap. The criterion values were based on expert opinion and are shown in Table 1, which are in line with other studies cited by Osman et al [2]. A normal group was composed of patients with PFS judged to be normal, taking no medication related to bladder or urethra, and (for women) no clinical obstruction. Men who had both a low bladder contractility index and a high BOO index, suggesting simultaneous DU and BOO, were excluded from the analysis. Women patients with clinical obstruction, defined as urethral/bladder neck obstruction and/or large cystocoele or prolapse through the introitus, were also excluded from the DU and normal groups. Using these criteria, 1788 patient records (men: 507; women: 1281) were classified to DU, BOO, or normal PFS groups and used in the analysis (Fig. 1).

Age for men and women DU patients was statistically significantly higher compared with patients with normal PFS (median values: men 63 yr vs 55 yr; women 59 yr vs 44 yr). Age for women with DU was statistically significantly higher compared with women with BOO (median, 59 yr vs 44 yr). There were no significant differences in height, weight, or body mass index between the groups.

Table 2 and Table 3 show the odds ratios and corresponding 95% confidence intervals for categorical variables that showed a statistically significant difference between DU patients and at least one other group (p < 0.05). Table 4 gives a summary of medians and interquartile ranges for numerical variables that had statistically significant differences between DU patients and at least one other group. Supplementary Table 6 details all variables included in the analysis, so it may be deduced which variables were not statistically significantly different between groups by their absence from other tables.

The primary group comparisons of interest were DU versus BOO in men (the most difficult clinical differential diagnosis) and DU versus normal PFS in women. These comparisons are therefore described below and are all adjusted for age.

Refer to Supplementary Figure 2 (men) and Figure 3 (women) of forest plots and Table 5 showing symptoms with statistically significant differences for patients with DU compared with those with normal PFS or BOO.

This study shows that cross-sectional data can be used to associate signs and symptoms with urodynamically-defined DU.

Two earlier studies failed to detect clear differences in clinical presentation of DU patients compared with other patients, and concluded that few symptoms were helpful [6] and [7], whereas this study and another recent report in women by Rademakers et al [8] used strict criteria and a wide range of symptoms.

The present study corroborated Rademakers et al's [8] findings of an increased prevalence of incomplete bladder emptying, hesitancy, and a weak stream in women with DU compared with women with normal PFS. Some variables from the medical history that were significantly associated with DU may reflect some patients’ natural history. For example, previous surgery could affect pelvic innervation and decrease bladder function and sensation. Also, the association of DU with prior BOO surgery may be due to preoperatively present DU, contributing to the overall clinical picture. Any retrospective look at such patients would find some patients with persisting symptoms due to unmasked DU.

The strengths of this analysis are that the urodynamic technique and structure of the interview were similar throughout the period and that data covering a large number of patients were used. A limitation is that the database is not a reflection of the general urological patient population, nor of the normal population in general, since all patients were referred for specialist evaluation of functional urological problems and many had received prior diagnosis and/or intervention. This was a retrospective, post hoc analysis of an existing database. A nonvalidated, although constant, set of questions were used and many data points were derived from clinician recording of patient responses to questions, and the potential for bias in the reported rates of underlying symptoms within each group cannot be discounted. The percentage of patients showing certain derived variables (see Materials and methods section) may have differed if patients had been directly asked if they experienced the symptom. Additionally, there are inherent limitations caused by the testing of multiple groups and endpoints. Notwithstanding these limitations, the analysis shows that there appears to be differences in signs and symptoms between DU patients in comparison with BOO patients and patients with normal PFS, which can be used to create instruments used to evaluate the results of treatments for UAB/DU.

Whilst patients in the normal PFS group were required to have a bladder voiding efficiency (BVE = volume voided/[volume voided + post void residual volume] × 100%) of 100%, a cut-off of 90% was used to distinguish between DU and BOO. Some studies have used BVE <60% as indicative of DU. The results did not differ greatly when comparing the symptom patterns of DU patients using cut-offs of 60%, 70%, 80%, and 90%. We therefore feel that it is legitimate to use BVE <90% as the criterion for DU, and BVE ≥90% for BOO patients in order to clearly differentiate the groups. Although there would be obstructed patients with a BVE <90%, to have incorporated these patients in the analysis would potentially have masked true differences in symptom patterns due to patients with both DU and BOO.

The observations in this study suggest that further work to develop a specific symptom questionnaire to assess DU severity, possibly coupled with noninvasive tests, could be useful for diagnosis, assessment, and evaluation of treatment outcomes. The importance of the development of noninvasive methods to characterise DU has been highlighted by several authors [2], [4], [8], [9], and [10]. Further analyses of this database, including flow and voiding parameters that may be relevant to the definition of DU [3], are planned.

The present study has demonstrated the utility of database analysis to aid the development of a symptom-based definition of a traditionally invasively diagnosed urological condition. The identification and diagnosis of UAB patients has currently been hampered by a poor understanding of the clinical presentation of DU and the necessity of invasive PFS. The present study has shown that there are signs and symptoms that can distinguish men and women DU patients from patients with normal PFS, and further distinguish between DU and BOO. This analysis is a first step to better understand the clinical presentation of DU patients, is consistent with the recently published UAB working definition [1], and justifies developing and testing a diagnostic algorithm based on the signs and symptoms of DU.

Author contributions: Andrew Gammie had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Gammie, Kaper, Dorrepaal, Kos, Abrams.

Acquisition of data: Gammie.

Analysis and interpretation of data: Gammie, Kaper, Dorrepaal, Kos, Abrams.

Drafting of the manuscript: Gammie, Abrams.

Critical revision of the manuscript for important intellectual content Gammie, Kaper, Dorrepaal, Kos, Abrams.

Statistical analysis: Kaper.

Obtaining funding: None.

Administrative, technical, or material support: None.

Supervision: None.

Other: None.

Financial disclosures: Andrew Gammie certifies that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: Kaper, Dorrepaal, and Kos are employees of Astellas Pharma Europe.

Funding/Support and role of the sponsor: Gammie and Abrams were supported for this analysis by a grant from Astellas Pharma Europe.

Acknowledgments: Analysis work was funded by Astellas Pharma Europe, who also contributed to the design of the study, analysis of the data, and preparation and review of the manuscript.

The clinical diagnosis of detrusor underactivity (DU) is hampered by the need for invasive pressure flow studies (PFS) and a lack of knowledge of the associated signs and symptoms. This has contributed to a lack of awareness of DU and its clinical correlate, underactive bladder (UAB) [1]. In consequence, this condition has been neglected compared with other causes of lower urinary tract symptoms. A recent review [2] concluded that DU “is surrounded by ambiguity” and recognises the limitations of the current definition. The International Continence Society defines DU as “a contraction of reduced strength and/or duration, resulting in prolonged bladder emptying and/or a failure to achieve complete bladder emptying within a normal time span” [3]. This, however, does not define “prolonged bladder emptying” or “normal time span”. Various methods have been proposed to determine contraction strength [2]; however, none of these take into account the duration of contraction – a key factor in the definition [3].

Despite this imprecision, estimates suggest that DU is a prevalent condition, ranging from 9% to 23% in men <50 yr, increasing to as much as 48% in men >70 yr [2]. Elderly women show a DU prevalence ranging from 12% to 45% [2]. An analysis of the signs and symptoms associated with DU could potentially facilitate the diagnosis of patients with UAB, improve our knowledge of the epidemiology, indicate possible noninvasive diagnostic approaches, and facilitate the development and evaluation of treatment outcomes of new therapies for UAB [4].

The aim of this study was to investigate the signs and symptoms recorded in a large database of patients referred for urological evaluation who met strictly defined PFS criteria for DU, bladder outlet obstruction (BOO) or normal, in order to identify the clinical features associated with DU.

Data from patients who underwent PFS, studied in a single specialist centre between 1985 and 2012, were recorded in a database that used the same variable fields throughout the 28-yr period.

Data gathering included patient interview to obtain symptoms and medical history, bladder diary data, physical examination, urodynamic studies, and diagnostic conclusions. PFS were carried out according to International Continence Society guidelines current at the time of testing. Free flow uroflowmetry was performed before each PFS. Postvoid residual urine volume was based on the volume obtained with catheterisation before filling commenced. The data from each PFS were screened for artefacts and manually entered into the database, thus avoiding automated data extraction errors. Prior to analysing the data, impossible values were removed in order to reduce corruption of data by manual entry errors. Several categorical (yes/no) variables used in the analysis were derived from a combination of database entry fields. For example, additional variables for straining and for decreased sensation were derived by combining the number of patients who reported these as symptoms with the number of patients for whom these were noted during PFS.

Patients without full voiding data, with neurological diseases affecting the lower urinary tract such as multiple sclerosis, paraplegia, or Parkinson's disease, and/or with a urodynamic diagnosis of detrusor overactivity were excluded as these require special consideration [5]. This resulted in 9928 eligible patient records (men: 1639; women: 8289) without confounding causes of vesico-urethral dysfunction (Fig. 1).

In order to classify patients with pure DU, BOO, or normal PFS, very strict criteria were used to avoid overlap. The criterion values were based on expert opinion and are shown in Table 1, which are in line with other studies cited by Osman et al [2]. A normal group was composed of patients with PFS judged to be normal, taking no medication related to bladder or urethra, and (for women) no clinical obstruction. Men who had both a low bladder contractility index and a high BOO index, suggesting simultaneous DU and BOO, were excluded from the analysis. Women patients with clinical obstruction, defined as urethral/bladder neck obstruction and/or large cystocoele or prolapse through the introitus, were also excluded from the DU and normal groups. Using these criteria, 1788 patient records (men: 507; women: 1281) were classified to DU, BOO, or normal PFS groups and used in the analysis (Fig. 1).

Age for men and women DU patients was statistically significantly higher compared with patients with normal PFS (median values: men 63 yr vs 55 yr; women 59 yr vs 44 yr). Age for women with DU was statistically significantly higher compared with women with BOO (median, 59 yr vs 44 yr). There were no significant differences in height, weight, or body mass index between the groups.

Table 2 and Table 3 show the odds ratios and corresponding 95% confidence intervals for categorical variables that showed a statistically significant difference between DU patients and at least one other group (p < 0.05). Table 4 gives a summary of medians and interquartile ranges for numerical variables that had statistically significant differences between DU patients and at least one other group. Supplementary Table 6 details all variables included in the analysis, so it may be deduced which variables were not statistically significantly different between groups by their absence from other tables.

The primary group comparisons of interest were DU versus BOO in men (the most difficult clinical differential diagnosis) and DU versus normal PFS in women. These comparisons are therefore described below and are all adjusted for age.

Refer to Supplementary Figure 2 (men) and Figure 3 (women) of forest plots and Table 5 showing symptoms with statistically significant differences for patients with DU compared with those with normal PFS or BOO.

This study shows that cross-sectional data can be used to associate signs and symptoms with urodynamically-defined DU.

Two earlier studies failed to detect clear differences in clinical presentation of DU patients compared with other patients, and concluded that few symptoms were helpful [6] and [7], whereas this study and another recent report in women by Rademakers et al [8] used strict criteria and a wide range of symptoms.

The present study corroborated Rademakers et al's [8] findings of an increased prevalence of incomplete bladder emptying, hesitancy, and a weak stream in women with DU compared with women with normal PFS. Some variables from the medical history that were significantly associated with DU may reflect some patients’ natural history. For example, previous surgery could affect pelvic innervation and decrease bladder function and sensation. Also, the association of DU with prior BOO surgery may be due to preoperatively present DU, contributing to the overall clinical picture. Any retrospective look at such patients would find some patients with persisting symptoms due to unmasked DU.

The strengths of this analysis are that the urodynamic technique and structure of the interview were similar throughout the period and that data covering a large number of patients were used. A limitation is that the database is not a reflection of the general urological patient population, nor of the normal population in general, since all patients were referred for specialist evaluation of functional urological problems and many had received prior diagnosis and/or intervention. This was a retrospective, post hoc analysis of an existing database. A nonvalidated, although constant, set of questions were used and many data points were derived from clinician recording of patient responses to questions, and the potential for bias in the reported rates of underlying symptoms within each group cannot be discounted. The percentage of patients showing certain derived variables (see Materials and methods section) may have differed if patients had been directly asked if they experienced the symptom. Additionally, there are inherent limitations caused by the testing of multiple groups and endpoints. Notwithstanding these limitations, the analysis shows that there appears to be differences in signs and symptoms between DU patients in comparison with BOO patients and patients with normal PFS, which can be used to create instruments used to evaluate the results of treatments for UAB/DU.

Whilst patients in the normal PFS group were required to have a bladder voiding efficiency (BVE = volume voided/[volume voided + post void residual volume] × 100%) of 100%, a cut-off of 90% was used to distinguish between DU and BOO. Some studies have used BVE <60% as indicative of DU. The results did not differ greatly when comparing the symptom patterns of DU patients using cut-offs of 60%, 70%, 80%, and 90%. We therefore feel that it is legitimate to use BVE <90% as the criterion for DU, and BVE ≥90% for BOO patients in order to clearly differentiate the groups. Although there would be obstructed patients with a BVE <90%, to have incorporated these patients in the analysis would potentially have masked true differences in symptom patterns due to patients with both DU and BOO.

The observations in this study suggest that further work to develop a specific symptom questionnaire to assess DU severity, possibly coupled with noninvasive tests, could be useful for diagnosis, assessment, and evaluation of treatment outcomes. The importance of the development of noninvasive methods to characterise DU has been highlighted by several authors [2], [4], [8], [9], and [10]. Further analyses of this database, including flow and voiding parameters that may be relevant to the definition of DU [3], are planned.

The present study has demonstrated the utility of database analysis to aid the development of a symptom-based definition of a traditionally invasively diagnosed urological condition. The identification and diagnosis of UAB patients has currently been hampered by a poor understanding of the clinical presentation of DU and the necessity of invasive PFS. The present study has shown that there are signs and symptoms that can distinguish men and women DU patients from patients with normal PFS, and further distinguish between DU and BOO. This analysis is a first step to better understand the clinical presentation of DU patients, is consistent with the recently published UAB working definition [1], and justifies developing and testing a diagnostic algorithm based on the signs and symptoms of DU.

Author contributions: Andrew Gammie had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Gammie, Kaper, Dorrepaal, Kos, Abrams.

Acquisition of data: Gammie.

Analysis and interpretation of data: Gammie, Kaper, Dorrepaal, Kos, Abrams.

Drafting of the manuscript: Gammie, Abrams.

Critical revision of the manuscript for important intellectual content Gammie, Kaper, Dorrepaal, Kos, Abrams.

Statistical analysis: Kaper.

Obtaining funding: None.

Administrative, technical, or material support: None.

Supervision: None.

Other: None.

Financial disclosures: Andrew Gammie certifies that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: Kaper, Dorrepaal, and Kos are employees of Astellas Pharma Europe.

Funding/Support and role of the sponsor: Gammie and Abrams were supported for this analysis by a grant from Astellas Pharma Europe.

Acknowledgments: Analysis work was funded by Astellas Pharma Europe, who also contributed to the design of the study, analysis of the data, and preparation and review of the manuscript.

The clinical diagnosis of detrusor underactivity (DU) is hampered by the need for invasive pressure flow studies (PFS) and a lack of knowledge of the associated signs and symptoms. This has contributed to a lack of awareness of DU and its clinical correlate, underactive bladder (UAB) [1]. In consequence, this condition has been neglected compared with other causes of lower urinary tract symptoms. A recent review [2] concluded that DU “is surrounded by ambiguity” and recognises the limitations of the current definition. The International Continence Society defines DU as “a contraction of reduced strength and/or duration, resulting in prolonged bladder emptying and/or a failure to achieve complete bladder emptying within a normal time span” [3]. This, however, does not define “prolonged bladder emptying” or “normal time span”. Various methods have been proposed to determine contraction strength [2]; however, none of these take into account the duration of contraction – a key factor in the definition [3].

Despite this imprecision, estimates suggest that DU is a prevalent condition, ranging from 9% to 23% in men <50 yr, increasing to as much as 48% in men >70 yr [2]. Elderly women show a DU prevalence ranging from 12% to 45% [2]. An analysis of the signs and symptoms associated with DU could potentially facilitate the diagnosis of patients with UAB, improve our knowledge of the epidemiology, indicate possible noninvasive diagnostic approaches, and facilitate the development and evaluation of treatment outcomes of new therapies for UAB [4].

The aim of this study was to investigate the signs and symptoms recorded in a large database of patients referred for urological evaluation who met strictly defined PFS criteria for DU, bladder outlet obstruction (BOO) or normal, in order to identify the clinical features associated with DU.

Data from patients who underwent PFS, studied in a single specialist centre between 1985 and 2012, were recorded in a database that used the same variable fields throughout the 28-yr period.

Data gathering included patient interview to obtain symptoms and medical history, bladder diary data, physical examination, urodynamic studies, and diagnostic conclusions. PFS were carried out according to International Continence Society guidelines current at the time of testing. Free flow uroflowmetry was performed before each PFS. Postvoid residual urine volume was based on the volume obtained with catheterisation before filling commenced. The data from each PFS were screened for artefacts and manually entered into the database, thus avoiding automated data extraction errors. Prior to analysing the data, impossible values were removed in order to reduce corruption of data by manual entry errors. Several categorical (yes/no) variables used in the analysis were derived from a combination of database entry fields. For example, additional variables for straining and for decreased sensation were derived by combining the number of patients who reported these as symptoms with the number of patients for whom these were noted during PFS.

Patients without full voiding data, with neurological diseases affecting the lower urinary tract such as multiple sclerosis, paraplegia, or Parkinson's disease, and/or with a urodynamic diagnosis of detrusor overactivity were excluded as these require special consideration [5]. This resulted in 9928 eligible patient records (men: 1639; women: 8289) without confounding causes of vesico-urethral dysfunction (Fig. 1).

In order to classify patients with pure DU, BOO, or normal PFS, very strict criteria were used to avoid overlap. The criterion values were based on expert opinion and are shown in Table 1, which are in line with other studies cited by Osman et al [2]. A normal group was composed of patients with PFS judged to be normal, taking no medication related to bladder or urethra, and (for women) no clinical obstruction. Men who had both a low bladder contractility index and a high BOO index, suggesting simultaneous DU and BOO, were excluded from the analysis. Women patients with clinical obstruction, defined as urethral/bladder neck obstruction and/or large cystocoele or prolapse through the introitus, were also excluded from the DU and normal groups. Using these criteria, 1788 patient records (men: 507; women: 1281) were classified to DU, BOO, or normal PFS groups and used in the analysis (Fig. 1).

Age for men and women DU patients was statistically significantly higher compared with patients with normal PFS (median values: men 63 yr vs 55 yr; women 59 yr vs 44 yr). Age for women with DU was statistically significantly higher compared with women with BOO (median, 59 yr vs 44 yr). There were no significant differences in height, weight, or body mass index between the groups.

Table 2 and Table 3 show the odds ratios and corresponding 95% confidence intervals for categorical variables that showed a statistically significant difference between DU patients and at least one other group (p < 0.05). Table 4 gives a summary of medians and interquartile ranges for numerical variables that had statistically significant differences between DU patients and at least one other group. Supplementary Table 6 details all variables included in the analysis, so it may be deduced which variables were not statistically significantly different between groups by their absence from other tables.